GenAI startup valuations begin to peak back down

Article source:There's a new Newin

Image credit: Generated by Boundless AI

Valuations of AI startups may have fallen from their peaks in recent months. Three months ago, media outlet Information calculated such valuation multiples for eight companies that primarily offer consumer or enterprise services related to LLM.

Investors valued these companies at 83 times ARR (annual revenue from sales) or 12 times MRR (monthly revenue from sales) at the time of investment, with three AI startups averaging 50% revenue multiples of the original eight companies, a decline attributed to revenue growth.

When a startup starts generating revenue, their future sales multiples can slip quickly from the high multiples they had early on, when they were generating little revenue. According to Meritech Capital, the average future revenue for publicly traded software companies is at 7x.

The three AI startups that have raised money in recent months include Perplexity AI, which saw its revenue multiple drop to 67 from 150 in its last round of funding. investors may be wondering how Perplexity plans to challenge the search engine's incumbent giants while still using data from Google Search and Microsoft Bing. Even Perplexity's CEO Aravind Srinivas has previously said that it remains unclear how the startup will make money in the long run, as it will have a hard time capitalizing on the lucrative advertising model used by Google.

How do you see an environment where you can achieve profitability earlier than the previous generation of companies?In a conversation with Stripe's CTO, Srinivas said that with revenue you don't have to keep raising money, you can keep expanding the top of the funnel and keep optimizing conversions in order to build a more sustainable and long-lasting business, rather than a short-lived fad. If you really want to build a company, it's better to become profitable as soon as possible, and it's better to be as efficient as possible, which also allows early stage entrepreneurs to raise more money later. When good milestones are reached, investors will assume that this will indeed work. That's why Srinivas has set Perpleixty's goal for this year at 100 million active users.

HeyGen was recently valued at $440 million with an ARR of over $20 million, a growth that is particularly notable among many generative AI startups that are struggling to commercialize and fund, and its new valuation is about 22x forward revenue compared to many other generative AI startups' valuation multiples.

Together.ai recently raised more than $100 million in new funding, bringing the company's valuation to more than $1 billion after only 20 months of existence.

Together AI Likely to face margin pressure due to the high cost of renting AI chips, its revenue multiple was 38 in February, up from 24 in October, Together is also one of the more controversial AI upstarts within the overseas investor community, with some VCs seeing it as making important strides in helping developers run and train open-source models at a low cost, but others seeing it more as a chip reseller.

Cohere is a Toronto-based LLM developer currently in talks with investors, including Canadian pension fund PSP, and expects to raise more than $500 million; Cohere's business is not as fast-growing and doesn't generate as much revenue as other companies, and its latest round of valuation is at 220x, so why?

Cohere is one of the few major non-U.S.-based LLM developers, and Canada is eager to back a project in its home country, see Mistral. while Cohere's revenue growth has been slow so far, Cohere also accelerated its sales growth with the launch of its latest model in March. If the new funding round goes well, Cohere will become the third most funded LLM startup after OpenAI and Anthropic, giving it more time and money to expand.

Additionally, last week Accel led a round of investment in Scale AI, a shovel company founded by a 27 year old Chinese man and valued at $13 billion! With the latest primary and secondary market insights from Accel partners. Last year, Scale AI generated over $675M in revenue, up ~$1,50% from the previous year, and the new growth has prompted Alexandr Wang, the 27-year-old founder of Scale AI, to accelerate expansion and seek new funding.

Accel said it is more interested in "pick and shovel" businesses like Scale, where Scale AI is the data engine behind many of the models, not just the base players, but also the people trying to tweak them for their vertical applications. Customers.

Reference:

https://www.theinformation.com/articles/ai-valuations-may-be-coming-down-to-earth-a-glimpse-of-openais-search-engine?rc=z9mejq