Anthropic might like to buy your side project

3 min read Original article ↗

You may have seen that Anthropic and OpenAI are building consulting firms. You may assume the thesis is some combination of the following:

  1. MBB, the Big 4, Accenture, etc. are ill-equipped to consult on AI

  2. AI increases the TAM of consulting

  3. OpenAI and Anthropic want more companies to spend more tokens, and building AI products for them is one way to do that

#1 is interesting, and likely true1. This is probably why any professional services company gets founded, if you replace “AI” with whatever the founders think they’re good at.

#2 is maybe the least interesting, but also likely true, at least in the short term (either if AI-fueled consulting is a fad, or if AI-fueled consulting somehow supersedes non-AI-centric consulting). It seems unlikely that, idk, Shell seeking advice on Capex for oil rigs is in anyway affected by AI (except in the execution path of the consultant doing the work), or that the daily deluge of due-diligence engagements Bain gets from PE-firms is affected in any way. So really it’s that Shell now needs to also pay consultants to “modernize” them, and PE-firms need to pay for advice on how to modernize internally and how to modernize their port-cos2.

#3 is obviously true. It’s sort of the reverse-Palantir model. Instead of “this problem you’re paying us to solve can be solved with this SaaS product” becomes “this problem you’re solving with our SaaS product is made much better if we just build a solution that uses our SaaS product [tokens]”

So why are there so many new consulting firms then? Because they all want to be bought by OpenAI or Anthropic? What does Blackstone have to do with this? Why is McKinsey investing?

My theory3 is that OpenAI, Anthropic, and a torrent of new “consulting” firms are actually private equity firms, that sell a little consulting. Their ambition is not simply to increase token spend (though that’s an added benefit) or add ~$10B in consulting revenue. They want a pipeline of M&A targets that fit this criteria:

  1. Could be made more valuable

  2. Could be transformed with AI

  3. Is for sale

And they want to buy these companies.

Pipeline is a huge problem in PE. PE firms pay a lot of money for lists of companies to buy, or invest in4. Let them come to you! Charge them for it!

This is not an entirely new model, obviously. Thrive Holdings is one of the more famous modern examples, though AFAIK, does not sell consulting as a service5. Tiny, Red, and other holdings companies don’t either.

This has interesting implications for the SaaS-Pocalypse, though. Does Salesforce need to worry more if OpenAI buys a CRM? Does Shopify need to worry if Anthropic buys a shopping platform? It certainly changes the economics of consulting, and maybe more interestingly, the economics of small technology companies.

If this is the gold rush, we all seem to be obsessed with pickax supply chains all of a sudden. I guess OpenAI is investing in refineries6.

The new meta: start a company, pretend you’ve never heard of AI, pay for a small consulting contract, sell your company.

Discussion about this post

Ready for more?