Staying Optimistic about B2B SaaS

B2B SaaS has experienced a huge period of growth, with more than 180 European SaaS startups becoming unicorns between 2016-2022. Many investors were happy to accept low levels of profitability, following the “Rule of 40”, which started to become less relevant once capital became more expensive. As a result, the Bessemer Cloud Index has trailed the S&P500 in the last 12 months and top quartile FTM revenue multiples have fallen from more than 20x to closer to 8x.

This has resulted in a flurry of M&A activity from software PE funds, thanks to more attractive entry points, but 50x exit multiples are likely to be rare in the next couple of years. SaaS startups that can grow with high gross margins, sales efficiency and wide moats are going to be the most highly prized in this environment.

Despite this, we’re still excited to invest in B2B SaaS startups, with a focus on those that can leverage emerging technologies as well as those that can digitise “boring” industries.

Creative software

2022 has been the year of somewhat surprisingly bold (due to its size) acquisition of Figma by Adobe. The event is yet another sign for the end of the “single-player” era and dominance of the “multiplayer mode” in creative software. Real time collaboration is the central point of any creative process and Figma is set to largely benefit from putting Figma as the collaborative hub of its vast product offering (50 live apps in its ecosystem, at the time of writing). Network effects are set to become much more powerful in this new setting.

Democratization of creation. Similarly to Instagram leading everyone to access professional quality filters and to become a (more or less) pro photographer, increasing accessibility and ease of use of creative tools will broaden the user base of graphic design or audio/video edition software tools. For instance, marketing and brand managers will be much more inclined to perform specific creative tasks themselves, instead of outsourcing them to an agency. This will enlarge addressable markets of creative tools (given higher ease of use for less sophisticated users) and will push creative professionals to more specialized roles.

Shifting to visual content creation processes has also been one of the predominant trends in 2022. Visual AI models such as DALL-E 2, Midjourney and Stable Diffusion are leading the way and are on track to significantly shorten the creative process between describing and executing specific design. As the marginal cost of creating additional art approaches zero (majority of this cost is, afterall, the human touch), skillsets of creative professionals will adapt to a new reality of mass supply of content, hyper-personalization and increasing consumption cycles.

Niche and vertically-specialized categories, in areas such as audio/video editing, are gaining momentum. Podcast creators, for instance, with use of tools such as Podcastle, can easily improve their work by removing silences and ‘emm’ noises in a matter of seconds. Youtubers can leverage tools like Elevenlabs to reach new audiences with seamless voice translations to other languages.

Artificial Intelligence

This post wasn’t written by ChatGPT, but the recent advances in Large Language Models (LLMs) and Latent Diffusion Models (LDMs) have led to a huge leap in how we think about generating content. The rapid success of products like Github Co-pilot or Jasper show the speed that new products and services can reach scale with these tools, but it’s likely that the really exciting products haven’t been launched yet.

Copilot for knowledge work where the product allows the knowledge worker to spend more time on complex work instead of the drudgery of creating PowerPoint decks or drafting contracts. According to some estimates, the potential improvement in knowledge worker’s output is estimated at 140% until 2030 due AI-based work augmentation.

It’s easy to imagine replacing an email to a lawyer with “write me an NDA including these key terms” or “simulate the liquidation waterfall for an exit of €250M based on this shareholders agreement”. Similarly a copilot for medical professionals could speed up triage of patients and be able to recommend options for treatment, allowing generalists to perform at a similar level as specialist physicians.

On a unit basis, the largest productivity gains coming from AI-assisted tools will be visible among office and administrative roles. We’ve seen some early examples with products such as ChatBA and we expect to see plenty more come forward in the next few months with the hype around the space. Perhaps the days of a junior management consultant are numbered!

Building the software translation layer. We currently spend a huge amount of effort and expense building software so applications can speak to each other. It’s possible to envisage a model using AI that can build and test these integrations in an ad-hoc manner, lowering the barriers to entry for using new software and removing the need for tools like Zapier. Products such as Levity are a good first step, but haven’t yet crossed the chasm to providing end-to-end automation.

Multiple content business models are at risk of being disrupted. Rise of GPT-3 and other powerful models led to companies such as Notion and Microsoft to add this capability directly into their software and services. This move allows them to stay in the layer of customer interaction and therefore be able to ‘own’ (and monetize) the user. The alternative is to become scraped by another algorithm and be blended into someone else’s interface.

Check out our previous post for our outlook on 2023 as well as a focus piece on consumer and marketplaces.