Lazard Venture & Growth 2023 Outlook – Six sector themes for investors to be aware of
2022 was a tale of two halves in global venture and growth markets. 1H22 appeared to benefit from the spill-over from prior year activity, before the change in global liquidity and financing conditions drove a second half slowdown and a valuation reset. US and European fundraising ended the year down around 30% YoY. Private capital flows rotated between sectors and congregated around specific industry verticals. Late-stage growth companies were most affected.
We may expect a similar segmentation in 2023, with deal volumes potentially lighter in the first half, and fundraising concentrated within areas of structural change and higher growth visibility. However, in 2H23 we are more constructive, and believe deal activity could pick up as companies seek equity financing following the prevalence of bridge financing options throughout 2022 such as internally led rounds, debt financing, and a reduction in operating costs. Many investors are well capitalized and may look to take advantage of the lower valuation environment.
In Lazard VGB’s 2023 outlook, we cover six areas of potential opportunity across our sector coverage: InfraTech, Enterprise Software, FinTech, Healthcare, DeepTech, and Consumer. As a global team, the Lazard Venture & Growth Banking team cover both the European and US markets. In the report, we first provide an overview of 10 predictions for 2023, and then we map six disruptive trends which may benefit from investment activity.
InfraTech: voluntary carbon markets
Enterprise Software: the modern data stack
FinTech: embedded finance
Healthcare: value-based care-enabling technology
DeepTech: sorting through the AI landscape
This document has been prepared by Lazard & Co., Limited ("Lazard") solely for general information purposes and is based on publicly available information which has not been independently verified by Lazard. The information contained herein is preliminary and should not be relied upon for any purpose. No liability whatsoever is accepted, and neither Lazard nor any member of the Lazard Group (being Lazard Ltd and its direct and indirect subsidiary and associated undertakings) nor any of their respective directors, partners, officers, employees, representatives or other agents is, or will be, making any warranty, representation or undertaking (expressed or implied) concerning the accuracy or truthfulness of any of the information, ideas, forecasts, projections or of any of the views or opinions contained in this document or any other written or oral statement provided in connection herewith or for any errors, omissions or misstatements contained herein or for any reliance that any party may seek to place upon any such information. Nothing contained in this document constitutes, or should be relied upon as, (i) the giving of financial, investment or other advice by, or the issuance of research by, Lazard, or (ii) a promise or representation as to any matter whether as to the past or the future. Lazard undertakes no obligation to provide the recipient with access to any additional information or to update or correct any information contained herein. Interested parties should conduct their own investigation and analysis of the matters and companies referenced herein. Nothing contained in this document constitutes, or should be deemed to constitute, an offer or solicitation for the purchase or sale of any security. You undertake to keep this document confidential and to not distribute it to any third party, or excerpt from or reproduce this document (in whole or in part), without the prior written consent of Lazard. Lazard, which is a regulated financial adviser, only acts for those entities whom it has identified as its client in a signed engagement letter and no-one else and will not be responsible to anyone other than such client for providing the protections afforded to clients of Lazard nor for providing advice. Recipients are recommended to seek their own financial and other advice and should entirely rely solely on their own judgment, review and analysis of this document. Lazard or other members of the Lazard Group (i) may have acted in the past, or act currently or in the future as adviser to some of the companies referenced herein, (ii) may receive fees in connection with any such advisory engagements, (iii) may at any time be in contact with such companies in order to solicit them to enter into advisory engagements, and/or (iv) may from time to time have made, and may in the future make, investments in such companies. By accepting this document, recipients agree to be bound by the terms and conditions set out above.
In 2022, core inflation remained elevated throughout the year, central banks increased rates, and the US two-year yield rose almost 400bps. The US dollar strengthened by c.10% against major currencies and “risk-off” sentiment saw equities sell off through most of the year. Pockets of financial markets which were most elevated in 2021 were more impacted, such as technology stocks and venture capital. Equity and bond markets posted negative returns – with global bonds delivering their worst annual performance in over a century, and the Nasdaq down over 30%. On reflection, 2021 may have potentially pulled forward future returns or booked gains which now may not be realized.
At the start of 2023, there may be some reasons to be cautiously optimistic. While some reports suggest that consumer confidence has reached its lowest level in 50 years and rolling country recessions are the baseline expectation for economists across the Street, there may be a few green shoots. Headline inflation has started to fall, with the US inflation rate forecasted to exit 2023 at around 2-3% (from 6.5% in December 2022). The Federal Reserve is expected to slow the rate of interest-rate increases. The equity markets have started 2023 with some resilience. And, in our opinion, most pertinently, exciting companies are continuing to disrupt verticals in the venture and growth (V&G) markets.
Before diving into our sector analysis, let’s first generally evaluate some popular predictions for 2023 we have identified.
1. “Inflation to subside by YE23”: Breakevens currently suggest US headline inflation will exit 2023 at nearer 2% from current levels in the 6-7% range. While the impact of inflation may drive risk sentiment, we believe disruptive tech led businesses will be central to productivity gains and economic growth, with many areas of the economy still left to disrupt.
2. “Companies to continue focus on path to profitability”: The duration of public and private market investors seems to have converged, with significant erosion to the terminal value of many growth companies. The key theme of 2021 was perhaps ‘growth at all costs’; 2022 was reduce burn at all costs’; 2023 should hopefully be a better balance of risk, investment in growth, and building visibility of durable long-term margins.
3. “Best VC vintages follow recessions”: The reset in private valuations is likely to benefit future fund returns. To us, that suggests two things: i) the reduction in valuation levels could support future returns; and ii) funding should accrue to higher-quality companies.
4. “Down rounds expected to increase”: Companies were able to avoid marking their equity to market in 2022 by i) entering the year well-capitalized, ii) cutting operating costs, iii) engaging in short-term bridge financing, or iv) extending prior round terms with downside protection. In 2023, down rounds may increase as options i), ii) and iii) become potentially less viable, and the likely distorted incentives of iv) become more apparent.
5. “More industry consolidation and strategic activity”: When capital is freely available, companies are incentivized to seek growth independently. We may expect category leaders to continue to attract through-the-cycle funding, with mid-tier players possibly i) forced into defensive mergers, ii) acquired by industry leaders, or iii) targeted by strategic investors.
6. “Negative impact of goods disinflation, with services inflation ‘stickier’”: The greatest challenge for core inflation may now be the stickiness of services prices, which continue to rebound despite the disinflation in goods. In the near term, the shift to B2B service-based companies may potentially be preferred compared to firms with direct consumer exposure. That said, some of the more disruptive consumer propositions have been born in downturns - though in our view it is likely to be another year or two before these reach the growth stage.
7. “Private markets to be resilient given companies have been started by serial entrepreneurs”: The maturity in European private markets over the past cycle means founders are on average more experienced and could be better equipped to manage economic volatility. While 2021/22 vintage returns are likely to look challenged, the longer-run returns of the asset class could remain strong.
8. “‘Dry powder’ will support V&G activity”: We expect dry powder to provide some valuation support for companies with certain core attributes: category leaders, robust business models, revenue visibility, strong unit economics, and large addressable markets. However, we also believe investors’ quality thresholds may remain elevated, with certain verticals attracting the lion’s share of funding.
9. “Compression of private capital in late-stage growth will harm valuations”: The closure of the IPO market in 2H22 led to a high compression of capital in later-stage growth companies. Retaining profitable growth ahead of a future listing is paramount, with tuck-in M&A and pre-IPO growth rounds likely to become more common through 2023. There may also be significant attrition with some stars permanently dimmed.
10. “Employee layoffs will boost new company formation”: Layoffs from technology companies could catalyze a further rebound in new business formations. A career choice shift from knowledge industries to entrepreneurship looks to us to be structural not cyclical.
Section 1: InfraTech − carbon management
Offsetting emissions may be a contributor to meeting net-zero goals....
When we dissect the general ledger for climate change, on the liability side, organizations produce greenhouse gases (GHGs) by moving or making goods. On the asset side, companies reduce, substitute, and offset these emissions. Under net-zero targeting, companies must eliminate their net climate-liability position.
Companies have two options to reach net zero: 1) abate GHG emissions or 2) offset emissions through the carbon credit markets. However, the lack of current technological availability and the challenges of abating scope 3 emissions has shifted a large volume of corporates into offsets. Against this backdrop, the voluntary carbon market (VCM) formation is attracting capital across various aspects of the value chain.
In 2023, four areas are likely to attract focus, in our view:
- first, project developers which supply carbon credits from either removal or avoidance projects;
- second, independent platforms which provide post-certification verification;
- third, carbon measurement tools for end buyers; and
- last, VCM infrastructure platforms.
…and is being enabled by the formation of voluntary carbon markets
The carbon markets have two components: i) mandatory markets; and ii) voluntary markets. According to RBC, mandatory (or compliance markets) cover only 23% of GHG emissions under current legislation. The residual falls into the VCM, where companies decide to offset their emissions, including Scope 1, 2, and 3 emissions, through the use of credits. Accordingly, it seems there is a need for a VCM that is large, transparent, verifiable, and environmentally robust.
The VCM reached $2bn of traded volume in 2021...
Carbon markets aim to transform emissions into a commodity, which is then priced and subsequently traded in a liquid market. The ultimate VCM market size will be a function of demand and corporate participation. The annual traded volume on the VCM market reached nearly $2bn in 2021, with demand of around 360 MtCO2. The VCM remains nascent compared to the mandatory market, which traded over $100bn in 2021.
Lazard’s Climate Centre estimate the VCM may reach $50bn in 2030, with the Taskforce on Scaling VCM (TSVCM) expecting demand to reach 1 to 2 GtCO2. Other sources such as Credit Suisse expect the VCM to reach between $50bn to $100bn by 2030, based on an average carbon price of $50 to $100 per ton at a demand of 1 GtCO2. In short, many industry observers are bullish on the VCM.
...with venture-backed companies increasingly being established across the value chain
When we break down the VCM value chain, we identify some key stakeholders: the financial institution; project developer; brokers; buyers; and regulators.
Figure 1: Voluntary carbon markets value chain
Source: Lazard VGB Insights
Capital availability is high to fund projects with reduced technical risk and attractive economics....
Financial institutions provide funding support to project developers who initiate carbon credits from projects. Maturity and liquidity development in the VCM will require continued growth in the availability of capital and loan supply.
Project developers (or issuers) initiate either avoidance or removal projects. Revenue from carbon credits used to fund projects has been concentrated among the top 20 developers, who have received around 40% of the credits.
There are two main types of projects. Avoidance projects include areas such as renewables, afforestation, carbon capture, bioenergy, and cookstove, whereas removal projects include direct air capture or reforestation, for example.
...with new technology key to continue to grow the market
Projects can be segmented a level further by i) technology and ii) nature – together technology and nature-based avoidance credits have accounted for over 95% of credit volume since 2010 according to the Berkeley Carbon Trading Projects. Removal is at an earlier stage of development.
Figure 2: VCM – distribution of avoidance and removal projects
Source: Berkeley Carbon Trading Project
- Technology-based avoidance includes projects which do not involve storing carbon, i.e. renewable-energy generation. According to Credit Suisse, these issuances have accounted for over half of carbon credits since 2010.
- Technology-based removal includes projects which extract carbon from the atmosphere for storage purposes.
- Nature-based avoidance includes avoided deforestation and removing carbon emissions that could have been avoided. Avoidance projects have accounted for over 40% of credit volumes in the last 10 years.
- Nature-based removal projects covers areas such as reforestation.
Carbon capture and removal has seen significant funding
Two areas of recent disruption have been within carbon capture (capturing CO2 from industrial processes as it is emitted) and carbon removal (capturing CO2 already in the atmosphere). These technologies are currently relatively expensive but represent a huge potential opportunity to contribute to greenhouse gas reduction. A number of companies have raised equity and debt capital to continue development, and to commercialize the technology.
Figure 3: Carbon capture and removal – company profiles
Source: Lazard VGB Insights, Pitchbook Data, Inc., company websites
Carbon credit rating agencies aim to improve regulatory environment
Regulators include crediting mechanisms which are needed to verify and list carbon credits on a registry for a particular project. There are a number of registries such as Verified Carbon Standard (or Verra) and Climate Action Reserve – registries can be domestic, international, and independent institutions.
There are two core issues: i) limited transparency with the potential for greenwashing; and ii) the potential for credit duplication (double counting) – especially post certification. Consequently, regulation such as Article 6 of the Paris Agreement aims to ensure only one country counts each credit. In the private markets, independent carbon-credit coverage has been an area of disruption, with the emergence of rating agencies which aim to provide oversight throughout the carbon credit life cycle.
BeZero Carbon is one example, raising $50m at series B in November 2022, with a view to building an independent rating system using proprietary methodologies across the full spectrum of projects – similar to Moody’s, S&P, or Fitch in the bond markets. Sylvera provides a similar service but is more focused on forestry. It raised $34m in a series A round led by Index Ventures in January 2022. Should demand continue to grow, post-certification oversight and uniformity will likely be a key enabler of VCM liquidity.
Carbon measurement tools have driven investor activity
End buyers are corporates or institutions which are seeking to offset their carbon emissions to the extent they cannot be abated. There are two identified challenges for corporates. First, scope 3 emissions account for over 80% of the total. Scope 3 relates to emissions produced by assets which are not owned and are therefore more readily offset. The second challenge is measurement error, with the accuracy gap running as high as 30% to 40%, according to BCG.
To address these issues, three main types of business models have emerged. Currently, measurement techniques are top down given the inability to capture emissions on a bottom-up basis currently. The business models are as follows:
- Carbon management tools which provide real-time environmental data insights;
- Carbon accounting platforms which measure the climate impact of an organization's activities based on certain standards; and
- ESG rating companies which score individual companies’ ESG credentials.
This part of the value chain has received sizeable funding in recent years. Capital has been deployed both in Europe and the US, and we expect continued flows across this segment, as scope 3 targets and tighter industry regulation threaten to widen the climate-risk premia.
Figure 4: Carbon measurement business models
Source: Lazard VGB Insights, Pitchbook Data, Inc., company websites
- Carbon management platforms like Carbon Direct, Sweep, and Doconomy all attracted growth capital in 2022 – these firms seek to help corporates manage their carbon footprint against defined climate goals. Doconomy more specifically, aims to quantify climate impact at the transaction level.
- Carbon accounting firms have similarly been a focus for investors. In Europe, Stockholm-based Normative raised a $30m series B round in July 2022. US-based Watershed – which offers a full-stack platform (measuring, reporting, and removing) − raised a $95m series B round in June 2022. Watershed assists in the production of carbon reporting (which is audited by third parties) and also integrates with direct investment in removal projects and carbon offsets.
- ESG-rating companies typically service sell-side and buy-side institutions, scoring underlying companies against specific criteria to help quantify ESG risks. Rating firms have developed from manual qualitative processes into AI integrated software models – ClarityAI is an example of a disruptive company within this vertical.
Improving VCM liquidity is an additional area of focus
Brokers: Most transactions occur over the counter (OTC) between project developer and the end offset buyer. In other cases, brokers and retail traders buy carbon credits from the producer, construct portfolios, and transfer them to the end buyer (companies) – these purchases can be tailored to the buyer’s offset requirements.
Some venture-backed companies have emerged within the trading infrastructure, aiming to reduce friction. San-Francisco-based Patch, which raised $55m at series B in September 2022, has developed an API integration which aims to connect carbon-credit suppliers (developers) with end buyers. The platform is B2B and B2B2C, seeking to help companies manage their own carbon credits or embed offsetting access to consumers.
Exchanges: Given the challenging liquidity requirements, a growing volume of trading is occurring on exchange. The creation of an agency market, in theory, should create standardized products (based on type, location, and vintage) and improved liquidity. Venture-backed Xpansiv (based in San Francisco) is the largest carbon exchange globally, facilitating around a third of market volume in 2021. Flowcarbon (based in New York − $32m series A in May 2022) also aims to help facilitate improved VCM liquidity, through bringing tokenized carbon credits on-chain.
Figure 5: Voluntary carbon market – company profiles
Source: Lazard VGB Insights, Pitchbook Data, Inc., company websites
Section 2: Enterprise software – the modern data stack
Infrastructure and data tools may continue to be attractive near-term plays...
Forester is projecting the global enterprise-software market to grow by 12% this year, with four verticals being prominent: 1) security, 2) IT management, 3) database management, and 4) infrastructure. These foundational pieces of IT operations are seen by many as safe havens for investors during times of market turbulence and generally may have more attractive efficiency metrics, as shown through indexing of top public infrastructure and application-software companies conducted by Redpoint Ventures:
Figure 6: Key metrics for public data software companies
*Data as of May 2022, infrastructure index includes top select security, data, DevOps companies
Source: Redpoint Ventures
We believe that data-infrastructure tools, often packaged with consumption/usage-based pricing models, can be an attractive area for software investors in a down market. One reason for this is because these companies can potentially command higher ACVs than most application software companies by having the ability to scale data volumes with minimal additional cost inputs. These solutions also are among the stickiest within a customer’s tech stack as significant quantities of data flow in and out of the platforms and become part of the ‘plumbing’ of both IT-user and business-user workflows.
A recent survey conducted by growth investor ICONIQ Growth found that the top categories in which startup and growth-stage founders expect to see increased spending in 2023 include key areas of infrastructure, such as product development (specifically BI/analytics tools and DevOps), and data and security (specifically data warehousing and application monitoring). Similarly, as shown in results from Battery Ventures’ recent CXO survey, companies are prioritizing security and data as top areas for budget growth in 2023 despite broader expense cuts across the tech stack:
Figure 7: Ranking of cloud software areas CXOs are prioritizing over the next 12 months
Rankings 1 (top priority) to 7 (least priority)
Source: Battery Ventures
...with implementation of modern data stacks a continued business need...
Additionally, many corporations are actively investing in the implementation of modern data stacks (MDS) to attempt to maximize developer efficiency, reduce costs, and minimize reliance on complex manual configurations. In Forrester’s recent enterprise software market analysis, the database segment is projected as one of the fastest growth areas (13% YoY) in 2023, trailing only cybersecurity and AI platform software. Analysts cited demand for real-time analytics as a key driver in the growth of the data software market.
Spurred by pervasive cloud computing adoption, MDS enable teams to replicate both unstructured and structured datasets in warehouses, and to transform and analyze that information in real-time to deliver key business insights. The components of MDS shown below are designed to offer three primary benefits: 1) automation; 2) predictive analytics and insights; and 3) flexibility/scalability.
Figure 8: MDS overview
Source: Portable, Lazard VGB Insights
Within all data-related investments, enterprises are investing most heavily in the foundational components of the MDS as revealed below in Battery Ventures’ survey results:
Figure 9: Surveyed company priorities within data software
Rankings 1 (top priority) to 10 (least priority)
Source: Battery Ventures
...requiring enterprises to choose quality breed components…
Data software architecture is a complex system comprised of many interdependent components along the process from ingestion to transformation to analysis. It seems that few companies truly capture end-to-end services, creating opportunities for investors within each segment of the modern data-architecture structure. Outlining MongoDB’s view of the modern architecture structure in the visual summary below, there are six layers within this architecture: data sourcing, acquisition, application, security, analytics, governance:
Figure 10: Modern data architecture
Source: MongoDB, Lazard VGB Insights
Our 2023 areas of focus within the broader data landscape center around several key themes:
1. The convergence of lakes and warehouses and the broader optimization of pipelines:
As many companies migrate their traditional on-premises systems to the cloud and SaaS solutions, top data storage competitors are moving toward converged analytics platforms. In theory, this type of solution could enable companies to focus less on where data is stored, and to focus more on the quality and outcomes of their datasets.
The separation of lakes and warehouses has historically created inefficiencies for enterprises, requiring IT teams to devote resources to both 1) users capable of evaluating the raw, unstructured data sitting in the lakes (typically data scientists), and 2) users who are closer to the business and focused on analyzing cleaned, refined data in warehouses to generate insights.
The merging of the two could create a range of benefits for data and business teams, as summarized below from those identified in a Google Cloud white paper back in 2021, when this trend was just beginning to materialize:
- Time to market: Data can be ingested and acted upon immediately, whether it is sourced from batch or real-time sources.
- Reduced risk: Existing tools and applications continue to work without needing to be rewritten. This reduces the risk and costs associated with change.
- Predictive analytics: The move away from the traditional view of data-marts and data mining to real-time decision-making, using fresh data, enhances business value.
- Data sharing: A converged environment becomes the one-stop shop for all, regardless of the type of user (i.e. data analyst / data engineer / data scientist). All can access the same managed environment and at different stages when it is required.
- Multi-modal support: Semi-structured and structured data are key differentiators within data warehouses and data lakes. Semi-structured data has some organizational properties such as metadata to make it easier to organize, but the data still does not conform to a strict schema. In the converged world, this is accommodated with extended semi-structured data support.
- Breakdown silos and ETL pipelines: Traditionally data capture, ingest, storage, processing, and serving are managed by different tools and environments, usually orchestrated by ETL tools. Now, most of the traditional ETL tasks can be done by ELT.
- Schema and governance: Business requirements and challenges evolve constantly. As the need for data changes, applying data-quality rules becomes more challenging and requires schema enforcement and evolution, which is easier in a converged environment.
We have seen examples of several emerging and well-profiled entrants into the market capitalizing on the convergence trend around optimizing extraction and integration through the data pipelines. These cloud-native companies seek to enable both the extraction of data from a broad variety of environments (ETL/ELT) and delivery of data from warehouses to business applications for end-use (reverse ETL).
Figure 11: Data pipeline company profiles
Source: Lazard VGB Insights, Pitchbook Data, Inc.
2. Demand for real-time intelligence:
Companies seeking to maximize SaaS tool productivity are likely seeking analytical/BI solutions that enable all employees – whether tech users (such as developers and data scientists) or typical business users – to be able to access and extract insights from structured and unstructured datasets in real-time with minimal manual intervention needed.
While the segment of the analytics market comprised of horizontal BI and notebook solutions has become somewhat saturated with market leaders probably limiting the scalability of new entrants, we are positive about 1) self-service analytics tools (both in data warehouses and outside them); 2) vertical, full-stack packages coming to market addressing specific business use cases; and 3) decision intelligence tools that help companies visualize and interpret trends from a broad range of sources.
According to a 2022 Komprise report, 65% of surveyed IT leaders said they are investing in unstructured data analytics solutions to empower end-users who are consuming data flowing from an increasing variety of sources. This reflects the growing market sentiment that investing in data science teams to manage data lakes and query data in its native form is a necessary expense to maximize the outputs of analytics functions and increase valuable use cases for large datasets.
Figure 12: Data type overview
Heading into 2023, key analytics/BI trends we are monitoring include the following:
- Democratizing data across the enterprise: With a growing willingness to embrace automation tools, we believe that enterprises are likely to continue to look for ways to get actionable data into the hands of regular business users, rather than just data science teams. McKinsey research shows that companies that make data accessible to their entire workforce are 40x more likely to report that analytics functions have a positive impact on revenue. This is being reinforced by the growth of the data-as-a-service market (estimated to now exceed $10bn), which could enable enterprises to streamline information flows and establish effective governance policies.
- Augmented analytics – incorporating AI/ML into BI processes: Leveraging data fabrics and focusing on metadata analysis may enable organizations to draw insights from a broader variety of sources in real-time by creating relationships between different systems. Gartner predicts that through 2024, organizations that adopt aggressive metadata analysis across their complete data management environment will decrease time to delivery of new data assets to users by as much as 70%.
- Early signs of consolidation: Gartner predicts that by 2025, 90% of new data and analytics deployments will be made through an established data ecosystem, causing consolidation across the data and analytics market. We believe industry-agnostic BI tools struggling to compete with legacy players could be attractive value-driven acquisition targets over the next 12 – 24 months.
- Predictive analytics a growing priority: In our view, organizations are likely to increasingly experiment with predictive insights to include use cases such as fraud detection, business operations optimization, and general risk reduction. Industries we have seen begin to adopt these tools include heavily-regulated sectors such as banking and financial services, utilities, and the public sector, as well as retail and consumer-focused companies looking to better understand their customer base.
Sources: McKinsey, Gartner, SAS
Below we highlight a number of emerging growth companies innovating in the analytics and transformation segment:
Figure 13: Analytics and transformation company profiles
Sources: Lazard VGB Insights, Pitchbook Data, Inc.
3. Governance and security in-motion:
In our view, the data security market remains poised for healthy growth after more than 4,100 publicly disclosed data breaches occurred in 2022, equating to approximately 22 billion records being exposed. Magna Research estimates the broader global data security market to average 12% annual growth over the next several years, driven most by data encryption, tokenization, and masking tools.
Figure 14: Global data-security-software market value by component, 2020-2029
Source: Magna Research
Figure 15: Data governance lifecycle
Source: Dataiku, Lazard VGB Insights, Dataversity
Figure 16: Emerging data governance and privacy players
Source: Lazard VGB Insights, Pitchbook Data, Inc.
4. New innovations in database management and performance:
Open source and horizontally scalable platforms have spurred strong growth…
Previously considered by some to be a slow-to-evolve area of software that was less user-friendly to both technical and business users, database management systems (DBMS) have undergone significant innovation with the near-universal adoption of cloud migration in the enterprise. DBMS are used to find and store information that can be securely modified based on an organization’s needs, and coupled with other business intelligence and computing tools, help them use organized, defined data to gain improved and effective decision-making, agility, and scalability.
Gartner estimates that the DBMS market will reach $150bn by 2030, and it has been one of the fastest-growing segments in enterprise software for the past several years, now accounting for over 13% of the overall software market. While relational DBMS (RDBMS) still represent the most popular segment of the market, growth rates are in decline. RDBMS differentiate from traditional DBMS in the following ways:
Figure 17: DBMS vs RDBMS
Source: Snowflake, UnumDB Blog
The rise of open-source and scale-out DBMS and No-SQL DBMS is challenging incumbent market leaders such as Oracle, Microsoft, and IBM, and creating a plethora of opportunities for investors. Additionally, managed databases (dbPaas) have driven 80%+ of the segment’s overall growth in each of the past two years, according to Gartner.
Figure 18a: DBMS overview by type
Source: Zuar, Analytics Vidhya, Lazard VGB Insights
Figure 18b: Transition from legacy vendors to high-growth engines
Source: Lazard VGB Insights
…particularly over the past two years as new market leaders have emerged…
Putting numbers behind the substantial momentum in this space, of a select 20 DBMS VC-backed companies who had raised over $50m as of the end of 2021 (shown below), there are some notable trends that we extrapolated based on initial analysis conducted by Ashot Vardanian of Unum:
- The sample companies have cumulatively raised over $9bn throughout their lifetime
- 2021 Records: 13 out of the 20 companies raised capital in 2021, totaling $4.5bn. Prior to 2021, these names had raised $7.0bn collectively, meaning 65% of their total capital raised occurred within the year
- 2022 Activity: These companies raised $1.3bn in 2022, with large rounds being heavily weighted towards 1H 2022 in line with broader market trends. This was still sizeable given the runway these companies established through $100m+ rounds raised in 2021. Additionally, one company IPO’d (MariaDB). New names emerged including Voltron Data ($110m through two rounds), Neon ($30m Series A), ReadySet ($29m Series A), and Convex ($26m Series A)
- Notable Investors: Investors in these rounds included: Redpoint Ventures, Andreessen Horowitz, Altimeter Capital, Benchmark, Coatue, Index Partners, Insight Partners, and Tiger Global
- Sub-Verticals of Interest: Select areas where investors are deploying into include:
- Graphs: Neo4J, Tigergraph
- Time-Series: Clickhouse
- In-Memory: Redis
- Postgres-compatible: Yugabyte, Timescale
- Analytics: Starburst, SingleStore
Figure 18c: Annual capital invested and funding round count of select 20 DBMS companies
$ in millions
Source: Pitchbook Data, Inc., UnumDB
In the following table, we summarize emerging database innovators that offer solutions focused on performance optimization and are capitalizing on DBMS market growth.
Figure 19: Select emerging database optimization players
Source: Lazard VGB Insights, Pitchbook Data, Inc.
4. Exploration with MLOps
Lastly, we expect to see more corporations exploring ML-based data tools that can be integrated into their production environments. Known broadly as MLOps, the industry is in the early stages of adoption for infrastructure solutions that enable model production at scale. We believe that several drivers, as identified in a recent Bessemer Venture Partners deep-dive and in analysis from Absolutdata, may support business cases for new ML infrastructure investments in 2023 and beyond, as follows:
- Growing sophistication with cloud-based architectures: Many enterprises are becoming more sophisticated with cloud-native IT implementations and governance. Additionally, IT executives are investing in data teams to optimize productivity while handling ever-growing data sources. As data tooling standardizes further, ML projects should become more feasible and investable.
- Network effects within ML: Over the last two years, the widespread success of foundational models trained on expansive datasets has unlocked high performance across ML modalities. The emergence of foundational models creates network effects, with one model able to flow into other derived models to fine-tune performance.
- Nuanced data training capabilities: ML practitioners are employing a diverse array of strategies to increase the amount of training data available. For example, self-supervised learning, which avoids human tagging by using the data and its underlying structure as labels, has seen significant progress in the natural language and vision domains. Simulation is also enabling more businesses to train models with minimal costs.
- Range of cost savings and efficiencies: When implemented with proper strategy and resources, MLOps tools can potentially offset the upfront costs required by providing a range of benefits including:
- enabling companies to use AI to efficiently address business concerns;
- allowing teams to create reproducible, reconfigurable, and reusable models to reduce development time;
- enhancing development and experimentation with new models, which leads to higher-quality results;
- using automation to support continuous delivery and monitoring;
- simplifying the hand-off from development to production, which enables better deployment; and
- offering continuous testing and monitoring to provide ongoing feedback, which is used to tune models for better results.
Below we present a four-quadrant market map, consisting of the following categories:
- Cloud Platforms: a group of services that provide storage and compute capabilities in an Internet-based data center
- Machine Learning Platforms: provides an end-to-end experience that affects the majority of a researcher’s ML workflow
- Specialized ML Tools: includes a variety of custom tools that affect how a researcher performs a particular ML task
- Standard SWE Tools: the standard software toolkit engineers use, even outside the context of ML
Figure 20: Abbreviated MLOps market map
Source: LJ Miranda, Lazard VGB Insights
Private markets appear robust for data software
In 2022, the market for data software continued to attract capital despite macroeconomic headwinds – five data-software companies attracted nearly $500m in funding in Q122 alone. While pace slowed throughout the year, growth investors seem to remain willing to pay premium valuations for differentiated data, infrastructure, management, and analytics solutions.
Figure 21: Selection of 2022 global data software V&G fundings
$ in millions
Source: Pitchbook Data, Inc., Lazard VGB Insights
Figure 22: Select additional data software prospects
Source: Pitchbook Data, Inc., Lazard VGB Insights
Section 3: Fintech − embedded finance
Every company may become a fintech company...
“Software is eating the world” is among the most commonly used phrases in markets given the wide-spread industry disruption which followed. Over the next decade, fintech could be poised to eat commerce, as A16Z partner Angela Strange recently stated: “Every company will be a fintech company”.
Many small and medium-sized businesses have benefitted from the democratization of technology across core business functions including accounting, payroll, HR, advertising, and logistics. Embedded finance is another: disintermediating incumbents, resetting distribution channels, and integrating banking products into non-financial companies. Recent adoption has likely benefitted from the shift in consumer spend to online, as the development of the orchestration layer creates potential new profit pools for both technology providers and merchants.
In our view, embedded finance could be central to private capital flows within the V&G markets. The market dynamics, value chain, and private-market activity are outlined below.
...driving the US embedded finance market towards $50bn+ by 2025
In 2021, embedded finance revenues exceeded $20bn in the US alone, according to McKinsey. Revenue was generated on transaction volume of $2.6trn, suggesting a take rate of around 80bps which varies by product. According to a report by Future Markets Insights, North America represents around 35% of the embedded-finance market, suggesting the global market may have reached almost $60bn in 2021.
The US market alone is expected to at least double in the next five years to $50bn – while Lightyear Capital expects the global embedded finance market to be worth up to $230bn by 2025. A large portion of this growth is expected to accrue to the top line of private fintech companies.
Figure 23: US embedded finance market size ($bn)
Source: McKinsey, Lazard VGB Insights
Tech-enabled providers have been a key area of disruption for growth companies
Simplifying the value chain is important to mapping the V&G landscape, and we would highlight a few players within embedded finance: the financial institution, technology enabler, merchant, and consumer.
- Regulated institution: A regulated license holder which sits at the top of the stack, offering services such as balance-sheet capital and license provisioning
- Tech enabler: A financial-software provider delivering an integration of financial services into the customer journey via APIs
- Merchants: These companies utilize customer insights and their distribution capabilities to provide financial products at the point of sale across industries
- Consumer: Customers are consuming financial products at the point of delivery for goods and services. Consumer education and younger digital natives are helping increase attachment rates.
In the figure below, we map the embedded-finance value chain. We differentiate from banking-as-a-service which reflects the more full-stack distribution of banking products.
Figure 24: Embedded finance – value chain
Source: Lazard VGB Insights
Payments and buy-now pay-later (BNPL) have been the main embedded products to date
There are four main types of standalone products:
- Payments: These contribute the bulk of volume to date given the wide usage of payment enablers like Stripe. Payments can be broken down into B2B or direct to consumer.
- Lending: BNPL is included in this product segment, reflecting the extension of credit at the point of sale. Revenue-based financing and factoring are being increasingly integrated.
- Deposits and issuing (or banking): This type of product includes more traditional savings and credit cards.
- Insurance: includes the attachment of single or multi-policy cover at the point of sale.
Based on 2021 data within a recent Bain report, payments accounted for over 90% of embedded transaction volume and 50% of revenues but derived the lowest take rate of around 60bps. Lending and banking delivered higher take rates, likely commensurate with the higher level of risk, relative maturity, and lower competition. Embedded payments and BNPL providers have raised the largest funding rounds in recent years, reflecting the high contribution to overall volume.
Figure 25: Implied take rates by embedded product
Source: Bain, Lazard VGB Insights
Secular drivers are supporting merchant adoption at the point of sale
In our view, a number of tailwinds exist for embedded finance adoption, potentially balanced between i) demand from consumers and merchants, and ii) increased supply driven by regulation.
- E-commerce penetration: Around 50% of US card spending now occurs online (33% globally), according to industry reports. Online card penetration is likely to be higher among SMEs, with relatively higher contribution from certain industries.
- Customer preferences: The integration of finance within the customer journey could provide a natural adjacency for many merchants, as consumers become increasingly familiar with digitally embedded products. Reduced friction at the point of sale is also likely to increase attach rates.
- Improved unit economics: Against a challenging macro backdrop, merchants have often faced both revenue growth and profitability headwinds. Embedded finance could present an attractive opportunity for diversifying revenue streams as well as margin accretion. Access to customer traffic means merchants are positioned to extract a share of the economics.
- Vertical SaaS: Many software firms with industry focus have been able to increase CLV and lower churn through embedding payments into the core value proposition. Merchants benefit proportionally through simplified processes and cost savings.
- Regulatory unlock: Regulation such as the Consumer Financial Protection Bureau’s (CFPB) announcement to democratize access to consumer data in October 2022 should improve the interaction between third-party applications and could help unlock increased supply of embedded finance providers.
Venture activity is spread across embedded products
Mapping the private market ecosystem for embedded finance is challenging, given market and geographical fragmentation. However, based on our analysis and interactions with companies and investors, we have mapped out some early-stage companies below:
Figure 26: Embedded finance – market map
Source: Lazard VGB Insights
A number of these companies raised rounds in 2022. We highlight some interesting deals below.
Figure 27: Embedded finance – company profiles
Source: Lazard VGB Insights, Pitchbook Data, Inc., company websites
Four trends expected in embedded finance
We now highlight four interesting themes which we believe may potentially play out in private markets:
- Increased supply within certain embedded verticals: Payments and BNPL have seen the greatest traction with consumers and investors to date. Looking ahead, products which have seen less relative activity and offer higher take rates may be targeted - such as lending.
- Broadening growth beyond the core product: Horizontal product expansion may help preserve technology providers’ differential advantage, and address product areas of greater resilience from take-rate erosion.
- Upstream integration: While not without risk, the backward integration of technology providers into balance-sheet lenders could help capture a larger share of the economics. This strategy may be dependent on the ability to obtain a banking license and presents exposure to credit risk.
- Embedding products within SaaS: The increasing penetration of financial products embedded within verticalized SaaS offerings could continue. This helps ensure software platforms are business critical and allows SaaS providers to leverage more diversified revenue streams to cross-subsidize core platform functions, pricing below competitors and reducing churn. Toast (which provides payment systems for the restaurant sector) is one good example of this strategy in practice.
Section 4: Digital health − value-based care enablement technology
Digital health funding eased in 2022...
Even as one of the more defensive sectors in tech, with a relatively stable, fixed spend base, US digital health investment dropped to pre-pandemic levels of $15.3bn in 2022, after having doubled to approximately $30bn in 2021. In the US, deals exceeding $100m showed a significant drop-off in the last 12 months, likely reflecting broader cross-industry trends of investors reducing check volumes for late-stage growth raises during a period of macro volatility.
As noted in Rock Health’s recent 2022 review, this trend marked a significant reversal from 2019-2021, when capital concentration was more prevalent in digital health than most other sectors and the top-five fundraises in the US totaled, on average, over 30% of total invested capital each quarter. The last three months of 2022 saw just $2.7bn invested across 146 deals (which includes a $330m raise for home care start-up Dispatch Health), less than half of 2021’s Q4 total of $7.4bn.
Much of the drop-off can likely be attributed to a waning of pandemic-inspired demand for virtual care solutions, which grew from 11% adoption of US consumers in 2019 to over 50% in 2021, according to McKinsey. Such quick growth in the market created overly saturated sub-verticals that could potentially spur consolidation in 2023 and beyond. Most investors expect a flat-to-moderate growth year in 2023 in the sector, with predicted investments totaling $15 – 25bn, according to a recent GSR Ventures survey of top digital health investors.
Figure 28: US annual digital health investment activity
Sources: DigitalHealth, Pitchbook Data, Inc., Lazard VGB Insights
Market focus in healthcare transitioning to Value-Based Care (VBC)...
Looking forward, the tone of conversations at the 2023 JP Morgan Healthcare Conference among investors, corporate executives, and entrepreneurs alike was cautiously optimistic. Select key themes which have been identified from the conference include:
- Value-based primary care remains focus, while specialty is the future: Although specialty and other alternative forms of care command a large proportion of spend in the health-care industry, start-ups and V&G investors seem focused on reimagining primary care as a first target in the transition to value-based care. Broader market players, including retailers such as CVS Health and Optum, have been exploring and completing acquisitions to expand hybrid care models and enable greater VBC access for patients. Many in the industry expect JVs and partnerships between VBC-enablers and health systems to accelerate. Below is a depiction of how specialty areas currently rate in terms of VBC adoption:
Figure 29: VBC adoption by specialty area
- Commercial, Medicare Advantage, and Medicaid critical to long-term growth: To put numbers on progress to date, Medicare Advantage (35%) and Traditional Medicare (24%) reimbursements continue to lead the path to VBC adoption, as measured by total percentage of payments flowing through alternative models (APMs) in 2021, according to LAN's annual APM study. Medicaid—which covers one in every four U.S. citizens—is seen as an immediate growth area for VBC adoption, based on moderate YoY increases seen in each of the past two years. For both Medicaid and commercial reimbursements, more than 50% of payments still flow through the traditional fee-for-service (FFS) model. In total, ~20% of all U.S. healthcare payments flowed through risk-based payment models in 2021, with another 40% tied to quality and value. However, nationally, only ~10% of health systems’ revenue is tied to VBC models, and measured by enterprise value, VBC companies account for an estimated $60bn of the $1.3trn healthcare market (<5%), which seems to affirm that we are still in the early stages of an industry transition that many prognosticators consider may be inevitable. VBC contracts cover ~25% of total insured U.S. lives today, based on industry estimates.
Figure 30: Projected lives in VBC models by reimbursement type (millions of lives)
- Financial and operational improvements expected: Many health-system operators expect revenues to rebound after a sluggish 2022 and believe endemic staffing shortages could begin to abate and return to a stable capacity by the end of 1H23.
- Health systems expansion: Health systems seem to be increasingly active in evaluating potential acquisitions of provider groups and expanding into specialty service offerings outside of hospitals seeking to expand access to a larger patient population, reinforcing care delivery in support of the VBC transition.
- Valuations in a new reality: Certain healthcare services companies that recently were valued at 20x ARR are now being valued at closer to 3x ARR. Chief executive of General Catalyst, Hemant Tenaja, indicated that start-ups need to be creative with figuring out how to grow their valuations because these multiples are unlikely to return to pandemic levels.
Sources: Pitchbook Data, Inc., Axios
...with a possible re-calibration of the sharing of risk between payers and providers...
We believe one area for growth investors to potentially focus on in 2023 is identifying differentiated tech platforms that enable the broader industry transition to VBC payment models. Put simply, VBC models enable health-care providers to be paid on the basis of patient outcomes rather than volumes. Back in 2016, Bessemer Venture Partners highlighted the $1tn opportunity in seeing an industry-wide shift from the historical FFS to VBC payment models. Recent McKinsey analysis estimates that in a bullish scenario fueled by increased private-market funding, this goal could be attainable as early as 2027.
At the heart of the VBC model is the sharing of risk among payers and providers, leading to more collaboration and improved care quality to improve patient outcomes. The range of VBC payment models, along with associated risks, compares to FFS as described below:
Figure 31: Range of VBC Payment Models
Source: BHM Healthcare Solutions
...but the shift may be hard to navigate with substantial challenges to reaping the benefits...
The VBC payment model should provide clear value to each of the three major health-care stakeholder groups. However, operationalizing and implementing the transformation away from FFS operations in an often outdated, fragmented industry has proven more challenging than many VBC advocates anticipated. In theory, VBC seeks to align incentives across the payer-provider-patient ecosystem. However, many providers and payers with near-term concerns, including complex contracting requirements and impacts to patient volumes, revenue, and implementation costs, seem to remain reluctant to change. Below we indicate some of the potential benefits and challenges stakeholders are likely weighing in the VBC transition:
Figure 32: Potential impacts of VBC adoption to healthcare stakeholders
Source: Veradigm, Lazard VGB Insights, McKinsey
The transition is benefiting from multiple tailwinds...
Despite delays, recent regulatory and market indicators have become possible tailwinds toward broader VBC acceptance:
- Regulatory: The Centers for Medicare & Medicaid Services (the largest US payer) is embracing more VBC initiatives and has issued an ambitious goal to shift 100% of Medicare beneficiaries into an accountable-care (ACO) relationship by 2030. Of the 1,760 active ACO contracts in the US, private-sector arrangements continue to represent the largest payer type.
- Private markets: According to McKinsey analysis, private-capital inflows to VBC companies increased more than 4x from 2019 to 2021, while new-hospital construction—a proxy for investment in legacy-care delivery models—remained flat. Additionally, as many investors are considering sustainable profits over expedited top-line growth, FFS models will likely become less appealing.
- A recent GSR Ventures survey found that digital health investors see the greatest opportunity for startups in 2023 to focus on value-based care models and the consumerization of healthcare businesses, followed by interoperability, connectivity, and reimbursement for digital health solutions.
- Patient coverage: McKinsey estimates the number of patients treated by physicians within the VBC landscape could roughly double in the next five years, growing approximately 15% per year.
- Stakeholder prioritization: Recent KLAS research and surveys of top health-care stakeholders revealed that 71% of primary-care providers believe their capabilities required to deliver on value-based care will improve over the next five years, but practices will have to be intentional in their plans to adopt VBC models.
- Payer/provider convergence: While having gained most traction in primary-care settings to date, payer and provider convergence seems to have also facilitated the emergence of enabling data tools to help stakeholders realize VBC. Payers can play an important role in helping providers make the shift to VBC by simplifying and aligning quality and process metrics so the providers can be rewarded for clinical and operational changes.
Source: McKinsey, Oliver Wyman, PE Hub, Buyouts Insider, GSR Ventures
... disruptive technology may drive the quickest ROI key to the next leg of growth...
Given the perceived regulatory and broader market support for the VBC transition, we believe now could be the time for investors to look at technologies that facilitate the transition. Below, we segment the market of VBC-enabling tech, and identify some areas we consider necessary for the transition to move forward.
Figure 33: Select components of VBC-enabling tech ecosystem with high ROI
Source: Lazard VGB Insights
We potentially see positive opportunities for investors in those enabling technologies that help providers, payers, and patients realize the quickest forms of ROI from adopting the VBC models. We have identified several areas below that are important to the success of the VBC transition. While some of these dependencies have already garnered significant private-investor backing, others require capital to transform:
- Virtual and hybrid/alternative care: While this category has seen the most investment dollars to date in the VBC-enablement ecosystem, hybrid and alternative-care companies (specialty chronic and at-home services) beyond telehealth remain an attractive opportunity.
- Remote patient monitoring: Patient health data is collected and shared electronically with providers to enable early interventions.
- Integrated care platforms: These optimize information exchange across EHRs and other provider and payer systems.
- Population health tracking: Population health analysis leverages risk-stratification models to help providers and payers spot high-risk patients within VBC populations.
- Referral management: This involves screening patients and coordinating the referral process across various medical and social organizations.
- Home and community-based care programs: As hospital care becomes increasingly expensive, payers and providers are looking to move care to lower-cost settings, such as the home (community-based care).
- Risk-adjustment coding: This is one important process in designing and setting budgets for VBC arrangements. Payers use risk adjustment to compensate providers fairly for taking on patients with different risk characteristics.
- Contract management: Designing, implementing, and tracking a VBC contract can be a complex process. Companies in this section offer software-analytics platforms to help payers and providers develop and manage VBC contracts.
- Utilization management: This is a cost-containment strategy that involves the evaluation of the medical necessity and efficiency of health-care services, procedures, and facilities. Companies in this category develop software solutions to reduce unnecessary medical services and administrative burdens
Source: Health Tech Stack
Figure 34: VGB analysis of levers key to industry-wide VBC success and current state of adoption
Source: Lazard VGB Insights
Diving deeper into some of the categories above, we have created an abbreviated market map of some emerging and later-stage growth companies which seem to be in support of the VBC transition. While many of these companies offer platforms that cut across multiple use cases, we have broadly bucketed select early- and late-stage leaders that we believe could have credentials to attract potential V&G investor interest in 2023 and beyond.
Figure 35: Market map – some emerging and mature growth VBC-enablers
Source: Health Tech Stack, Vital Signs, Lazard VGB Insights, Pitchbook Data, Inc.
Section 5: DeepTech – sorting through the AI landscape
AI demand increases despite broader software investment slowdown…
While generative AI has generated significant buzz in the V&G space in early 2023, and there are signs of market adoption (i.e. Microsoft’s recent $10bn dollar investment in OpenAI), in our opinion there may be other deep-learning and AI applications with enterprise use cases emerging as potentially more actionable investment opportunities for cautious V&G investors. As shown below, automation of IT and business operations, along with security and threat detection capabilities, remain core areas of focus for executives looking to develop or acquire AI-driven solutions.
Figure 36: % of surveyed respondents using AI for functional use cases
While it’s hard to define the market size for AI software as most general-purpose SaaS apps coming to market are powered in some fashion by AI, we estimate the global market to total between $60bn - $120bn ("core"/platform AI solutions only at the lower-end, all AI-enabled applications included at the higher-end). Additionally, industries including retail and financial services are the most active investors and producers of new AI applications. E-commerce tailwinds seem to have driven intelligent product recommendation capabilities to be differentiators for online retailers. In financial services, many institutions appear to be increasingly turning to AI to create personalized and automated banking experiences. Unsurprisingly, the education and public service sectors have been less rapid adopters, though we believe interest in AI innovation remains high in these industries and could be potential long-term growth catalysts.
Figure 37: % of surveyed respondents using AI by industry
- ROI by function has evolved: Across functions, service optimization, and customer analytics are key areas where enterprises are adopting AI tools, and this has remained consistent over the past four years. However, ROI by function has evolved – in 2018, AI in manufacturing and risk generated the most ROI, and today, as identified by McKinsey, marketing and sales, product development, and strategy and corporate finance are providing companies the best mix of improved revenues. Supply chain management stands out for significant cost reductions
Figure 38: YoY cost decreases and revenue increases by AI function (% of McKinsey respondents 2021 – 2022)
We have identified several areas that could drive potential corporate AI investment over the next 12 – 24 months:
- Security and risk management priorities: As enterprises look to reduce vendor load across the application and infrastructure layers in a challenging economy, those platforms with AI-enabled capabilities that can automate key parts of the security lifecycle are likely to be more attractive investments.
- Data management, analysis, and governance: As budgets for data science talent may decline due to economic constraints, businesses will need to become more familiar with AI/ML solutions that enable them to extract insights from large structured and unstructured datasets, automate data-driven processes, and generate predictive analytics.
- ESG reporting: Environmental and sustainability practices are at the forefront of many enterprises’ resource allocation strategies heading into 2023. Collecting, analyzing, and reporting on ESG performance from a variety of data sources will require automated solutions at scale.
- Natural language processing (NLP) market remains robust: As noted in Silicon Republic, NLP is one of the few fields in AI which is not limited by data – deep learning systems can merge languages, images, and real-life object detection to build models without first needing to aggregate large volumes of data from customers (i.e. as is the case with self-driving car software). The global NLP market was valued at $15bn as of 2022, and is projected to grow at a CAGR of 25%+ over the next five years.
- RPA use cases continue to expand: RPA is increasingly being viewed by some as a pivotal tool for businesses to at least partially-automate a range of functions including: IT Ops processes, customer care experiences, supply chain/inventory management, and asset utilization/performance.
- Moving towards an integrated AI future: Larger corporates (i.e. Fortune 500 companies) who have previously experimented with different siloed AI and ML applications could begin to adopt a more holistic strategy in which data obtained from across business functions can be used to strengthen ML models that touch multiple applications.
Sources: Lazard VGB Insights, VentureBeat, Silicon Republic, IBM, MarketsandMarkets Research
Figure 39: Abbreviated enterprise AI market map by functions and verticals
Source: AI Multiple, Lazard VGB Insights
Generative AI – hyper-growth demand may allow early winners to emerge, with a more fragmented landscape taking shape over time…
In our view, much of the current ‘noise’ surrounding generative AI is likely due to the fact that foundational models are enabling a volume of generative AI tools to be created with minimal resources required. As many in the V&G industry have highlighted recently, the ability to build a generative AI business without needing to invest significant dollars developing sophisticated models resembles the transition to the cloud that occurred over the past decade. Access to private APIs and open-source models reduces the upfront costs previously associated with building an AI business. Just as cloud adoption accelerated development speed and SaaS app creation, language models created by pioneers like OpenAI are advancing rapid deployment of new AI applications. Additionally, continuous training has enabled generative language models to increase in size by 10x annually on average in recent years. The fact that most generative AI apps today exist as plug-ins enhances distribution and attracts a broader user base from which models can be trained.
Private investment activity in this space has seen hyper-growth since last year, in which over $2.6bn was raised by generative AI companies. Year-to-date in 2023, we have already seen Microsoft announce a $10bn strategic investment into OpenAI as part of a broader exclusive commercial arrangement, as well as reported large fundings set to follow for Anthropic and Character AI.
Figure 40: Generative AI annual investment activity
Source: CB Insights, Pitchbook Data, Inc.
With the developer-friendly foundations of this technology mentioned above, growth investor Base10 Partners argues that companies in this category will reach $100m to $1bn in revenues over the next five-plus years, following a similar trajectory to SaaS leaders at the height of the cloud transition. Bessemer Venture Partners has predicted that we will go from less than 1% of online content being generated by AI models today, to over 50% in the next ten years. While we think it is too early to see how monetization schemes and content uptake for specific use cases could materialize, we expect over the next 12-15 months that the industry could see some businesses grab early, commanding market share and investor dollars while early entrants remain in an experimentation phase.
Over time, we believe investors’ potentially growing appetite for high-growth AI markets could support a broader competitive landscape with vertical-focused AI companies carving out their own niches. As far as timing of development and adoption of generative AI technologies is concerned, in our view the platform layer (for example, OpenAI, Stability AI, and Hugging Face) may continue to be the near-term focus of large check-writers because it helps create the foundation upon which applications will be built and scaled, and in some sense, could be seen is a mission-critical element of the category’s future growth.
Meanwhile, applications could be likely to trail infrastructure tools in adoption as enterprises attempt to limit their vendor load and scrutinize implementations of new technologies with unproven economics in a challenging economic environment. We summarize the foundations of the AI/ML stack below:
Figure 41: Illustrative generative AI architecture map
Source: Lazard VGB Insights, Scale Venture Partners, Sequoia Capital
The global generative AI software market is estimated to be valued between ~$6bn as of 2022, and industry experts project high-velocity compound annual growth of ~30% over the next decade. The end-customers today are largely in the creative/media and entertainment space (see market share breakdown below), however we might expect this disparity to reduce over the next 12+ months as text, audio, and visual applications come to market to address pain points for both soft- and heavy-asset industries (i.e. manufacturing, healthcare, financial services).
Figure 42: Generative AI 2022 market share by industry segment
Source: Precedence Research
A critical question investors will likely be asking regarding the trajectory of generative AI adoption in 2023 is how well will model-training capture learnings from a fast-growing global user base with varying use cases to prove tangible ROI? We see this as a difficult question to answer that could take years to formalize, with certain use cases proving success quicker than others (particularly in creative domains). In the meantime, in our view experienced DeepTech investors could be likely to hedge their bets on new generative AI use cases as the major cloud leaders (Microsoft, Google, Amazon) possibly invest more heavily into generative AI infrastructure tools developed in-house and by emerging startups.
- Text: This is the most advanced domain. Models are still adjusting to natural language nuances; however, current versions are mostly effective for short-form content with simplistic messaging.
- Code: Using AI to write code will have significant impacts on developer productivity and efficiency; currently, GitHub has become the training ground for developers implementing this use case. It could also benefit non-developers through more democratic code-sharing.
- Images: Use of image models to edit and modify generated images is a recent trend that is gaining traction.
- Speech: Speech-to-text automation has been prevalent in enterprises via contact centers and other back-office functions. More nuanced use cases, such as for film and podcasts, will require further refinements.
- Video and 3D: This is attracting attention for its potential ability to transform large markets in the consumer space such as gaming and AR/VR, and in the enterprise space, for things like construction, architecture, and product design.
- Other domains: Use cases have been seen in alternative proteins, audio/music, and other R&D functions.
Application layer (select use cases)
- Copywriting: A good early test-case for language models for use cases such as personalized web and email content to assist sales and marketing strategies and customer support. The short-form and stylized nature of the verbiage combined with the time and cost. pressures on these teams may drive demand for automated and augmented solutions. Additionally, marketing creatives were among the first to embrace social media and ‘low-code/no-code’ design tools for enterprise use.
- Vertical specific writing assistants: Sequoia’s belief that there is a strong opportunity to build better writing assistant applications for specific end-markets/use cases, such as legal contract writing, screenwriting, and healthcare documentation. Product differentiation here will probably depend on how well models are fine-tuned and designed for specific industries.
- Code generation: Some current applications can create significant efficiencies for developers to make them more productive: GitHub Copilot is now generating nearly 40% of code in the projects where it is installed, and significant market opportunity exists to expand code generation capabilities to consumers.
- Art generation: A growing proportion of art history and pop culture is now encoded in these large models, allowing anyone to explore themes and styles freely.
- Customer support: Given the recent rise of customer-success functions in the enterprise as a revenue-generating necessity, call centers have implemented synthetic voice and messaging solutions to increase service performance, quality, and timeliness.
- Gaming: One immediate focus is on speeding up production pipelines, however, the future state encompasses a potentially vast set of use cases including enablement of concept art and other content for game engines and in-game features, creating benefits for both developers and players.
- Media/advertising: This involves automation of agency work and optimization of ad copy and creative for consumers. Opportunities exist here for multi-modal generation that pairs sales messages with complementary visuals.
- Design: Prototyping digital and physical products is a highly manual and costly process. We have already seen high-fidelity renderings from rough sketches and prompt, and as 3-D models become available in the future, the generative design process will likely extend up through manufacturing and production (i.e. text to object).
- Social-media and digital communities: Many consumers are seeking new ways of expressing themselves, and generative tools could be a next evolution of this trend. New applications like Midjourney are creating new social experiences as consumers seek new outlets for public-facing content.
Source: Research from Sequoia Capital, Bessemer Venture Partners, Lazard VGB Insights
Figure 43: Select emerging companies in generative AI
Source: Base10 Partners, Lazard VGB Insights, Pitchbook Data, Inc.
Section 6: Consumer − TravelTech
Given macro headwinds the opportunity set in consumer may be weighted to earlier stage
The rotational forces within the consumer sector have meant finding durable earnings streams is often challenging for V&G investors. The delta in duration between public and private investors has converged, with near-term revenue visibility a priority for many investors when allocating capital. Based on our estimates, private funding levels across the US & European consumer sector fell by 60% in 2022, with late-stage companies seeing the greatest erosion to valuation.
A higher proportion of private capital is now being allocated to earlier-stage companies and appears to be congregated around certain thematic trends such as the circular economy and alternative proteins. We believe that another area of relative resilience has been within the experience economy, as digitally native consumers allocate the incremental dollar to personalized experiences, rather than goods.
Post pandemic TravelTech looks like an interesting growth vertical
In this section, we outline the broader theme of TravelTech and identify some pockets of potential opportunity for investors. Travel was severely impacted by Covid, but in our view the ongoing recovery, accelerated digital shift and high levels of fragmentation across the value chain suggests technological disruption in the sector could continue.
We have identified four areas of interest:
- Hotel management software
- Alternative accommodation
- Online booking platforms
- Corporate travel management
Let’s first set the scene, then travel across each of these areas below.
The global travel market was worth around $1.7trn in 2019
According to Citi, the global travel market was worth about $1.7trn in 2019. This splits down into a few buckets: hotels, airline, alternative accommodation (alts), experiences, and rail/car rental.
Figure 44: Global travel market – split by activity
Source: Citi Research, Lazard VGB Insights
Covid disrupted the travel market through 2020 and 2021, before a strong rebound last year. Heading into 2023, consumers seem to be continuing to prioritize travel despite the challenging economic outlook. Using data from a recent OC&C study, consumers are shifting their spend away from goods and towards leisure experiences such as travel, restaurants and cinema - likely due to pent up demand.
Figure 45: 2H22 – net consumer-spend outlook – expected to spend more vs. less
Source: OC&C, Lazard VGB Insights
Post Covid, online travel agents (OTAs) have accelerated their dominance
Booking volumes have accelerated their online shift post Covid, with OTAs taking share from offline tour operators, many of which have gone out of business. The top three OTAs command a dominant market share of booking volume and have continued to grow their penetration of the overall travel market. For AirBnB, Booking and Expedia, GMV is expected to exceed 2019 levels by around 24% in 2023E based on current street estimates.
Figure 46 OTA GMV versus 2019 levels
Source: FactSet, Lazard VGB Insights
Independent hotels are reliant on OTAs for customer traffic
Independent hotels are therefore reliant on the OTAs for indirect traffic and incremental occupancy. This has seen the top three OTAs extract a healthy share of the economics.
Figure 47: Online travel agencies – 2023E take rates
Source: Barclays, Lazard VGB Insights
Alternative accommodation platforms are also booming
Alternative accommodation is also running well ahead of 2019, with growth likely to remain as experiential and flexible travel continues. Hotels have returned to pre-Covid levels, with the lagged recovery in air travel suggesting domestic travelers may have led the rebound.
Figure 48: 2021-2024E – leisure bookings compared with 2019 levels by category
Source: Barclays, Lazard VGB Insights
1 – Hotel-management software offers improved economics for independents
Around 85% of hotels are independently owned, leaving a fragmented supply side often reliant on the OTAs for booking volumes. Many independent hotels lack pricing power, scale, and the ability to generate organic traffic. Hotel-management software provides strategic tools which can drive customer traffic and help recover economic ownership. In our view, there are potentially three main benefits to software adoption:
- first, network effects through leveraging data insights across platform customers;
- second, the ability to simplify in-house operations and consolidate the cost base; and
- third, the opportunity to deliver more personalized features and upsell incremental revenue opportunities.
There are four main use cases for hotel-management software: in-house operations, customer experience, revenue management, and marketing. Increasingly, providers are expanding beyond a core use case and targeting a full stack offering.
Figure 49: Hotel-management software-market breakdown
Source: Lazard VGB Insights
Using illustrative economics, we see why hotels might benefit from utilizing software tools to acquire direct traffic. Incremental occupancy delivers more marginal profit per room.
Figure 50: Illustrative economics – marginal profit per room
Source: Lazard VGB Insights
A number of vertical software providers have raised private funding rounds in recent years. Moreover, despite Covid, Siteminder – based in Australia − went public in November 2021. We set out below a few notable raises in Europe and the US over the past two years. Cloudbeds, based in San Diego raised the largest round to date, with $150m in late 2021. In Europe, Mews raised $185m in December 2022 against a backdrop of challenging market conditions. Mews supplements operational property management with embedded payments, creating higher switching costs and strategic value for customers.
Figure 51: Hotel-management software – company profiles
Source: Lazard VGB Insights, Pitchbook Data, Inc., company websites
2 - Alternative accommodation may be a secular growth trend
AirBnB trades at a higher multiple versus peers, likely due to its higher relative growth rate and Street expectations for continued outsized growth in alternative accommodation. If this thesis is correct, demand for short-term rentals and lodging accommodation may continue, catalyzing more activity within the segment. For us, there are two interesting areas.
- Short-term rental or lodging hosts may become more sophisticated as they scale, with software adoption expected to grow as hosts aim to drive occupancy and streamline costs. A number of companies have raised rounds in relation to this model in recent periods − Guesty is the largest, raising $170m in August 2022.
- The second area of interest relates to marketplaces. While AirBnB retains a high market share in alts (around 20%), alternative platforms may emerge with a more specific focus. Outdoorsy and Holidu are two companies we have identified which have attracted capital, each raising $100m rounds in June 2021 and October 2022 respectively. Texas-based Outdoorsy provides a marketplace for recreational-vehicle adventures, while German-based Holidu offers a rental booking platform, with an integrated host-software component, Bookiply.
Figure 52: Alternative accommodation – company profiles
Source: Lazard VGB Insights, Pitchbook Data, Inc., company websites
3 – Online tour operators simplify the frictions of complex travel
Travel spend has been more resilient following the pandemic compared to other leisure categories. However, this could change in 2023, with consumers under pressure from erosion to their purchasing power and low consumer confidence. In fact, it is estimated that consumer confidence has reached its lowest level for over 50 years.
Figure 53: Monthly consumer confidence – US and EA19
Source: World Bank, Lazard VGB Insights
The segmentation of the travel industry may provide some optimism. The figure below is taken from a recent OC&C report, which highlights the recovery of various sub-segments. We can make a few observations: i) traditional travel agencies ceded share to online; ii) complex, premium, multi-day tour packages look to have recovered well; iii) the OTAs have been important to the post-Covid rebound; and iv) vacation rentals (or alts) have seen strong demand.
Figure 54: Travel industry by sub-segment – current state of recovery vs. structural change in demand
Source: OC&C, Lazard VGB Insights
Heading into 2023, the figure above may be helpful to identify areas of relative resilience, with competitive moats, customer demographics, travel preferences, and geographical coverage particularly important.
- Deeper competitive moats: Construction of complex, multi-day travel appears to be comparatively well positioned. Aggregation of a fragmented supply base of hotels and experiences across a broad range of locations leaves these online tour operators potentially better insulated from OTA disruption.
- Demographics: Companies who target higher income and older demographics may also see relative resilience through a period of real-wage erosion and higher unemployment.
- Consumer preferences: Personalized travel and end-to-end digitization throughout the customer journey are two structural trends which continue post Covid. Integration of meta-search engines alongside booking platforms could help create community and loyalty.
- Geographical exposure: When we plot foreign-travel expenditure as a share of core PCE, the increase in foreign-travel intensity in the US pre Covid is interesting − it is around 30bps to 40bps higher in 2019 compared with 2015 after remaining rangebound since the 1990s. Targeting US consumers could present opportunities.
Figure 55: Travel PCE as a share of personal-consumption expenditures – 1990 to 2019
Source: US Census Bureau, Lazard VGB Insights
In V&G markets, a number of online tour operators have raised private capital in recent periods. These are split by geography but also by channel. B2C targets travelers directly and B2B provides booking infrastructure to existing tour operators (often local players).
In Europe, Paris-based Evaneos raised capital in 2018 and 2022, with a view to offering personalized and sustainable travel options to consumers. Exoticca – based in Barcelona – raised $30m in mid-2021 offering travelers buy-and-build optionality for complex travel, with a high exposure to US consumers.
4 – Corporate Travel slow to recover but SME market continues to offer opportunities
Our last post-Covid trend has been the recovery in corporate travel. Total transaction volumes fell by over 50% in 2020, and have been comparatively slow to recover. In fact, Morgan Stanley do not expect corporate travel to exceed 2019 levels until 2026.
Figure 56: Expected Global Business Travel Spend Recovery
Source: Global Business Travel Association (GBTA), Morgan Stanley Research
When we segment the $1.4 trillion business travel market, the opportunity for private companies within corporate travel software becomes clearer. According to Amex GBT, almost $950 billion is borne by SMEs, or companies who spend below $20 million on corporate air travel annually. Only around a third of SME travel spend is held under management, with market share fragmented. This is compared to enterprise travel spend of around $60 billion where Amex GBT commands 40% of annual volume.
Below we outline some of the potential benefits for SMEs adopting travel management software:
- End-to-end solution: platforms typically operate across use cases such as online booking, travel management, invoicing, and customer support;
- Efficiency gains: customers benefit from a simplified booking process, with improved workforce productivity and lower overheads;
- API integration: travel platforms are increasingly integrated across enterprise workflows through API connectivity with adjacent expense management tools such as HR and payroll.
There are also a few attractive characteristics to the SME platform business model:
- Large and growing market: the SME unmanaged segment is worth over $650 billion with an ongoing shift from unmanaged to managed corporate travel. Moreover, the overall SME business travel segment seems to be recovering ahead of enterprise in the post Covid recovery.
- Improved unit economics: the lower cost to serve SMEs can result in higher margins versus enterprise, according to Amex GBT. Amex GBT, for example, derives a share of total transaction volume of around 7-8%, suggesting take rates for the SME segment may exceed 10%.
- Two-sided monetization: both corporates and suppliers can be monetized by travel software platforms. Corporates via management and transaction fees, and suppliers through incentives and platform fees.
A number of corporate travel platforms have raised capital in recent years. Tripactions in the US, and Travelperk based in Spain have each raised large funding rounds to support their growth. Both firms have also been acquisitive to expand their global footprint and consolidate market share. For example, Tripactions acquired Comtravo in February last year to expand their reach in Scandinavia.
At an earlier-stage in the US market, Spotnana raised an $80m series B round in July 2022. Spotnana’s platform-as-a-service aims to help incumbent corporate travel firms digitize, and spend management companies integrate travel within broader enterprise spend. The company also provides core travel management software, with a more traditional end-to-end travel solution to corporates.
Corporate travel accounts on average for around 10% of company expenditure, and therefore capturing an increased share of company spend will help expand the profit pool. Looking ahead, we have identified three potential areas of expansion:
- First, growing customers through organic or inorganic entry into new geographies and sectors;
- Second, via access to ancillary revenue streams such as carbon management; and
- Third, through upward vertical integration into broader spend management.