Every services firm now says it is AI-native. Ask what that means and you get a slide of tool logos and a line about delivering faster. In a valuation conversation, that slide is worth nothing. A claim that costs nothing to make proves nothing, and investors and acquirers do not pay for it. They pay for what shows up in the numbers and in the business they can inspect.
Four years ago we wrote a ten-part series, Demystifying ETS Valuations, on how emerging tech services firms are valued: growth, revenue, gross margin, EBITDA, IP, addressable market, revenue predictability, team loyalty, badges and readiness to scale. Most of it still holds. Three parts need revising, because AI changed what drives value in each and what a buyer or investor can check. These are the three we want to revise, and here is what changed in each.
Revision 1: Margins
In Part 3 we argued that gross margin is the best single measure of the core strength of a business. It is what is left after the work is delivered, and it funds everything else. That has not changed. What has changed is how high the number can go.
Eventually your use of AI has to turn into money, and gross margin is where it shows. Traditional services firms run near 50% gross margin. Software runs near 85%. AI-Native Services (AINS) firms are pushing toward 70%. AINS businesses look more like software businesses in their economics.
Moving that number means fixing the whole value chain. An obvious part of that, which we have talked about before, is how your contracts are written. Do they let you capture the value of the efficiencies you create? But it goes deeper than that. Improving gross margin requires a constant search for the next bottleneck as you solve problems, and a team that keeps transforming along the way.
We saw the difference between point-in-time and concerted, consistent AI adoption at Red Buffer. Its engineers adopted AI coding tools and reported working 50 to 60% faster, but team performance improved only marginally. The problem was that each handoff required rebuilding context. So they stopped giving each engineer an individual AI and gave the project one instead. This change caused productivity to rise three to four times, and gross margin climbed from about 60% to over 80%.
Your gross margins show the core health of the business. But for an investor that is not enough. An acquirer may not care as much, because they will change your sales and operational processes anyway. An investor needs to see how well the current setup of the business will scale. So cost of sales and operating expenses matter to them too. We showed what that looks like here and here. Managing each of these costs shows how well the business is using AI to make a better business.
Revision 2: From team loyalty to a team aligned on AI-first
In Part 8 we wrote about team loyalty and the value of culture. In a services business, the people are ultimately what a buyer is trying to acquire. The problem is that their value is not clear. A buyer is looking for two things:
- The loyalty of people with skills and expertise.
- The ability to modernize and enable the larger organization.
Some data can help with the first, like employee satisfaction and retention, but it is not reliable, because acquisitions cause chaos. The second is very hard to define beyond the skills of the delivery team.
Team alignment on AI-first gives the second a definite meaning for the whole team. In a bolted-on firm, AI lives in pockets and nothing connects. In an aligned firm, everyone collaborates to build together. This is necessary to create an interconnected solution that delivers the margins an AI-native firm delivers. But it also joins the team together. The team gains cross-domain understanding and forms relationships.
You can see it in how the work connects. Bitsol Technologies links lead research to SOW drafting to delivery to the agents that watch production and propose fixes. Stormatics runs every customer call through one pipeline that surfaces the problems customers keep raising. At Brainforge, designers and engineers draw on a shared library of past SOWs. That is what modernizing a larger organization looks like, and every person on the team has already done it.
In addition, the scale of AI adoption indicates a dynamic and innovative team. That leads to a strong culture, and it points to inherent features of the business that make the team likely to stay through the leadership and other changes that come with an acquisition. If the only builder is a Head of AI, the rest of the firm was never expected to change.
Revision 3: AI-first processes and IP
In Part 5 we defined IP broadly: any asset that standardizes, streamlines or automates how the business runs, whether code, templates or checklists, and we showed it can lift a valuation multiple by 20 to 30%. AI has changed both the definition of services IP and its importance.
For starters, IP makes the difference between an AI-native services business and a tech services business. Everyone has access to the same models. The firms that are actually AI-native are building IP to make their businesses run more effectively. Both of the earlier sections are really about IP. Your margins are high because your team has come together to build IP. From our work with investment bankers and acquirers, and the negotiations we have seen in this field, the difference in valuation between an AI-native services firm and a tech services firm is easily 2 to 3x.
AI also makes it so easy to automate processes that process IP now needs to be encapsulated in code and agents. A checklist someone remembers to follow is no longer IP. In the firms furthest along, software factories already generate most of the code. Bitsol's Prototype-to-Production engine is one example: agents convert, test and secure the code, while people lead discovery and design and review every gate.
Show how much of the work runs through the system and how little still runs outside it, what the system costs to run, and how both are trending. A buyer or investor will check whether each engagement starts from what the last one built. When it does, the firm is running a machine that gets better with every project.
What the three revisions share
All three revisions point the same way. The firms that will earn a higher valuation changed how they are built, not what they bought. That is what makes the evidence hard to fake and easy to check. It is also why the old services multiple does not fit these firms, which is where we pick up next week: why services-as-software needs its own investor category.
To work through your own valuation with founders and investors in the room, join us at the AI Native Services Summit on November 16 at Pershing in Austin. Request your spot.