High Valuations and limited exit opportunities have some investors struggling to realize profits.

TLDR: High valuations for low-revenue startups with an unproven exit market.

Will you see your local AI startup in the future? Most likely no. And as an investor, you might not realize profits either.

Speculation on new AI startups coming for the blood of tech giants has been forcibly shut down. Big tech has bankrolled the industry seeking to turn enemies into friends, acquiring AI startups and poaching top talent. 

Many of the successful incumbent startups won’t remain independent for long. Most smart AI startups are building niche industry solutions to be acquired. Those that aren’t are viewed as an existential threat against big tech’s economic moat and sought after to be consumed.

However, these acquisitions delivered mixed results for investors—some yielded favorable exits, while others left much to be desired.

Character.ai: Pre-revenue with $1B Valuation. Acquihire for $2.7B by Google; licensed its technology and CEO plus 20% of staff were hired. In this example, investors did well.

Inflection.ai: Made almost no money (>5 million) and raised $1.5B at a $4B valuation. Acquihire by Microsoft for $620 million. Investors in the early $225 million round will be getting 1.5 times their investment; those in the later $1.3B round will get 1.1 times their investment. That math doesn’t add up with a $620 million acquisition, leaving the assumption that investors were paid in equity based on an exaggerated valuation from the skeleton of a startup that remained. Not happy for sure.

Adept.ai: $75 million in revenue. Raised $415M at a $1B valuation. Aquihire by Amazon with Adept receiving around $20 million and investors being made roughly whole. While the entire investment wasn’t lost, investors were not happy.

Honorable mentions: Octo.ai: $900M valuation, acquired by NVDA for $165M (LOSS)  Deci.ai: $200M valuation, acquired b NVDA for $300M (WIN) run.ai: $300M valuation, acquired by NVDA for $700M (WIN) amelia.ai: $189M raised, acquired by SoundHound for $80M (LOSS)

This trend suggests that a portion of AI startups will likely end up as part of larger tech conglomerates - falling into acquisition “kill zones” where theories state that VC investment is known to be discouraged when large players step in. Favorable exits for current investors might not be included in the deal as well.

Despite the existence of these kill zones, their effect on overall VC funding has been positive, not negative, contrary to some beliefs. A University of Michigan study (2010-2020) revealed that total VC funding increased by an average of 30.7% in the four quarters following a rise in Big Tech's startup acquisitions. 

Why don’t the numbers show decreased VC investment? Why do VCs continue to invest without sure exit opportunities?

AI companies receiving high valuations greatly limit their exit opportunities and become risky investments. Most companies valued over $1B will seek exit through IPO. Only 159 companies IPO’d in 2024 - a small number compared to the 1,549 unicorns waiting to offer liquidity to their investors. In a bottlenecked IPO market, what is the goal of funding AI companies to such high valuations where the main hope of exit is IPO?

Take pika.art - one of the best video generation AI models out there.

• Raised $135 million with a $500-700 million valuation (Business Insider). The exaggerated assumption that every single one of their 500,000 users has the 8$ paid version puts Pika’s revenue at $48 million.

• Pika’s valuation is too high to be acquired by large competitors and it needs a clear path to profitability and its valuation justified to IPO.

• The IPO market is its best shot at liquidity, however, inflation and economic uncertainty could continue to impact it.

You could look toward the M&A market, with Character.ai as the winning example for an AI unicorn. The acquisitions of Inflection.ai and adept.ai however represent losses for all investors involved, especially those in earlier rounds with less favorable rights.

What about AI startups being built to be acquired?

A lot of AI startups are seeking to solve specialized problems that exist in specific industries. These, most likely, will exit through a strategic acquisition within their target industry.

These startups’ successes are predicated on the companies in which they cater to. Whatever they are willing to pay for them is the going rate, not the valuation. Investors who get in at inflated valuations might not be happy with the exit offerings that companies can provide — existing in a more realistic business space.

Thanks for reading this far. If you’re interested in joining GTVC or being part of one of our events, please reach out to Xander Coles at [email protected]

Your feedback helps us improve.