Crossing the Segment Line: What AI Changed, and What It Didn't
For most of SaaS history, you picked a customer segment and built around it. Business or consumer. Enterprise, mid-market or SMB. Your product, pricing, and sales motion were all tailored to your chosen segment. Crossing into a new segment via M&A was where a lot of deals went wrong.
I ran corporate development and integration work at Insider One, Sitecore, and spent years at Thomson Reuters, and the segment question came up in every deal. The safe instinct was to stay in your lane, because the failure points were easy to see. And I did have deals go wrong because we bought businesses that were SMB-focused. Eventually these were viewed as non-core and divested.
The reason these deals went wrong wasn’t about ambition. Companies want to grow into a new segment to expand their addressable market and to drive growth. It was about economics. Four risks killed most attempts to acquire into a new segment.
Feature fit. Your enterprise product was too heavy to configure for a small business. Or your consumer / SMB product was too light on security, compliance, and admin controls to clear an enterprise procurement review.
Cost-to-serve. Supporting thousands of small accounts for a business configured to focus on larger customers was unprofitable. The long tail of customers was margin negative.
Pricing model. Smaller customers typically bought self-serve on a credit card. Larger customers negotiated annual (or longer) contracts. Running both motions created too much complexity for most organizations to navigate.
Go-to-market. Enterprise sells top-down and high-touch. SMB and consumer sell bottom-up and self-serve. Two very different motions, for two very different buyers, with two completely different comp plans.
AI has started to change the first three. Not completely, but enough to make some deals possible that would have been hard to justify a few years ago. The reason is simple. AI absorbs some of the labor that used to sit between a product and a new segment: support, onboarding, configuration, and first-line expertise. That makes some old segment lines cheaper to cross.
You can already see this in the deals getting done.
Feature fit: Atlassian and The Browser Company
In September 2025, Atlassian agreed to buy The Browser Company for a reported $610 million. The Browser Company built Arc and Dia, consumer browsers, and Atlassian said it would rebuild Dia for knowledge workers, not, in its own words, for someone's mom.
AI does reduce part of the feature-fit risk. A consumer product moving up-market used to fail because it lacked the depth enterprises expect. AI assistants that understand your work context close some of that gap. Atlassian said it still must build security, compliance, and admin controls. AI closes the gap on complexity and experience. It does not close the enterprise-readiness gap. If you are underwriting a deal on the theory that AI makes a consumer product enterprise-grade, stop there. It makes the product easier to use. It does not make it safe to sell to a buyer with compliance and regulatory requirements.
Cost-to-serve: Salesforce and Fin
In June 2026, Salesforce agreed to buy Fin, the customer service company formerly known as Intercom, for a reported $3.6 billion. Fin brought more than 30,000 customers, a large share of them small and mid-sized. Fin gives Salesforce an agent that resolves a reported 76% of support volume without a human.
That resolution rate is a massive draw for Salesforce on many fronts. One of those reasons is potentially the impact it has on cost-to-serve for smaller customers. The reason an enterprise company like Salesforce could never make money at the bottom of the market was small accounts creating support load that human teams could not cover profitably. With an agent that handles three out of four tickets, that changes. AI did not eliminate the cost of serving those accounts. It reduced the cost-to-serve barrier enough to make the long tail reachable. Fin is probably a hybrid deal though (more on this later). It gives Salesforce a customer base it can serve more cheaply, but it also gives Salesforce an AI support capability it can push back into its enterprise base through Agentforce.
Pricing model: Salesforce and m3ter
The pricing conflict changed too, not because AI made pricing simple, but because the unit of value changed.
On July 1, 2026, Salesforce completed its acquisition of m3ter, a metering and rating platform built for consumption-based billing. Terms were not disclosed. On its own, it sounds like plumbing. But for Salesforce, it plugs a major gap. The old mismatch was a segment problem: small companies bought self-serve seats, enterprises signed negotiated contracts, and serving both from one product strained packaging and billing. Metering moves the unit from seats to outcomes, and a price per outcome can work for a ten-person shop and a Fortune 500 in a way seats often cannot. (For a fuller examination of why outcome pricing is at the top of the value ladder for AI businesses, see my prior post “How to Monetize Your AI Roadmap”).
So m3ter is really the second half of the Fin story. Fin is the agent that makes the long tail cheap to serve. m3ter is the meter that lets Salesforce charge for what the agent does, regardless of customer size. m3ter is not itself a company crossing a segment. It is the billing layer that makes crossing one easier.
Two kinds of AI deals
So far, the examples are mostly about segment access: the buyer is trying to reach customers it could not serve, sell to, or price for before.
But AI deals do not fit neatly into one box. Some are capability deals. In those, the buyer is not chasing a new segment. It is buying an AI capability to push across customers it already has.
Some deals do both. Fin is a good example. It gives Salesforce both segment access and capability leverage: a customer base it can serve more cheaply, and an AI support capability it can push into Agentforce.
The distinction still matters because the diligence focus changes depending on the main reason for the deal.
Segment access
The one risk that AI hasn’t solved is whether you can earn trust and run the sales motion in a segment you have never sold to. AI makes onboarding, enablement, and first-line support cheaper, but it does not hand you the buyer relationship, the channel, or the domain credibility that wins a segment.
There is a well-worn argument that cross-segment expansion kills incumbents because skills do not port. Expertise in enterprise supply chain does not become expertise in consumer commerce because you wrote a check.
The buyers who get this right do not assume their own motion will carry. They buy the motion with the asset. When Canva, a prosumer design tool, wanted to reach professional designers, it bought Affinity, a professional suite with its own base and its own go-to-market. That deal has almost no AI in its logic. Canva did not bet that its prosumer playbook would stretch up-market. It bought a business with a go-to-market motion that already worked up-market.
Capability deals
Capability deals do not focus on whether you can sell into a new segment, but whether the model and its data can add value to the customers you already have.
In 2023, Thomson Reuters bought Casetext for a reported $650 million. Casetext had built CoCounsel, the first legal AI assistant on GPT-4, and Thomson Reuters ran that model across Westlaw, Practical Law, tax, and audit. One acquisition became a capability the current install base could benefit from. That move was rarer in 2023 than it is now.
The diligence does not stop at whether the model works. The harder question is whether the target has the rights to use the data the model depends on. Whether you can train on it, under the contracts and consents the target signed, is a critical diligence question. Usable, proprietary, domain-specific data becomes a massive differentiator for such deals.
The impact on M&A playbooks
To recap, AI reduced cost-to-serve as a primary reason not to cross a segment, and softened feature fit and pricing to where they are solvable. It did not solve go-to-market or domain credibility. This allows businesses to broaden the aperture of deals they might consider adding growth and creating barriers for players moving from other segments into their core market.
It also has implications for diligence given the two different types of deals, each of which fails for different reasons. If you are buying segment access, the question is whether you can earn trust and run the motion in a segment you have never sold to. If you are buying a capability, the question is whether the model and its data can legally and practically travel across your existing install base. AI is the trigger in both cases. What differs is where business cases break.
If you are building, buying, or operating in this space, I would love to compare notes. You can reach me at faraaz@inorganicedge.com or on LinkedIn.
Author's Note: I had no involvement in the transactions discussed here and no inside knowledge of them. The analysis reflects publicly reported information and my own experience running corporate development and integration.
Sources
Salesforce to acquire Fin (formerly Intercom): Salesforce press release, June 15, 2026. https://www.salesforce.com/news/press-releases/2026/06/15/salesforce-signs-definitive-agreement-to-acquire-fin/
Salesforce / Fin deal terms and customer count: TechCrunch, June 15, 2026. https://techcrunch.com/2026/06/15/salesforce-acquires-ai-customer-service-platform-fin-for-3-6b/
Atlassian to acquire The Browser Company: Atlassian / BusinessWire, September 4, 2025. https://www.businesswire.com/news/home/20250904645125/en/Atlassian-Enters-Into-Definitive-Agreement-to-Acquire-The-Browser-Company-of-New-York
Salesforce completes acquisition of m3ter: Salesforce, July 1, 2026. https://www.salesforce.com/news/stories/salesforce-signs-definitive-agreement-to-acquire-m3ter/
Thomson Reuters completes acquisition of Casetext: Thomson Reuters, August 17, 2023. https://www.thomsonreuters.com/en/press-releases/2023/august/thomson-reuters-completes-acquisition-of-casetext-inc
Canva acquires Affinity and Leonardo.ai: TechCrunch, July 29, 2024. https://techcrunch.com/2024/07/29/canva-acquires-leonardo-ai-to-boost-its-generative-ai-efforts/
Tool-to-outcome and TAM expansion framing: PwC, How AI Is Reshaping Software Valuations in M&A. https://www.pwc.com/us/en/services/consulting/deals/library/ai-software-valuations-ma-private-equity.html
