You receive the AI mandate in a short briefing, maybe over a slide deck, maybe over coffee. The targets look like any other operational goals. Then someone mentions the earn-out. And you realise the list you were handed is the list the founder gets paid on.
That discovery changes the shape of the work. The stakes are different. The conversations you need to have are different. The questions you should have asked on day one are different.
What is an AI milestone in a founder earn-out?
When a founder sells part or all of their business, the deal often includes an earn-out, a deferred payment contingent on hitting agreed targets after the transaction closes. AI milestones are a growing variant of this structure, where some portion of the earn-out depends on specific AI adoption goals, such as systems deployed, processes automated, or cost reduction attributable to AI.
Earn-outs are common in founder-led deals where the seller stays involved post-completion. This is the typical structure in investor-backed owner-managed businesses where the founder retains a stake and an operational role for some period after investment or partial sale. AI targets began appearing in these agreements as AI capability became a valuation lever, with advisors and acquirers treating demonstrated adoption as a signal of operational maturity and reduced founder dependency.
The milestone language varies widely. Some earn-outs use activity targets, a number of tools in production, a percentage of team members actively using an AI system, or a pilot completed by a given date. Others use outcome targets, documented cost savings attributed to AI, a reduction in decisions requiring the founder’s personal involvement, or revenue generated with AI support. The distinction between the two matters considerably for the person responsible for delivering them.
Why does this change what you prioritise?
When your operational AI targets and the founder’s earn-out milestones overlap, the work stops being purely about what the business needs. The founder has a financial stake in specific outcomes being recorded by a specific date. That creates pressure to hit the milestone language, which may or may not align with the AI work that will produce the most lasting operational value.
The priority conflict is real and concrete. If you have limited budget and limited team bandwidth, and one AI project scores well against the earn-out language but a different one would deliver greater operational impact, the earn-out creates a financial incentive toward the first. The founder rarely names this directly. They frame it as urgency, a deadline, a board expectation, or a conversation about momentum.
M&A research consistently finds that owner dependency is one of the most significant discounts to exit value in founder-led businesses. Buyers apply material reductions when the operations, relationships, and decisions of a business are founder-centric rather than systematised. AI adoption that genuinely reduces that dependency adds exit value. Milestone compliance that leaves the founder-dependency unchanged is a different proposition entirely.
That distinction is yours to manage. Your job is to keep the founder’s exit interest and the business’s actual interest aligned. In a well-structured situation they mostly do. Where they diverge, you need to know which one you are optimising for.
Where will you find the actual milestone language?
The milestone wording that governs the founder’s earn-out rarely appears in the brief you receive. It lives in deal documentation, typically the share purchase agreement, the investment term sheet, or the shareholders’ agreement. In many cases the founder has not shared this detail with their senior operators, either because it was not considered relevant or because the personal financial exposure felt awkward to surface.
Asking to see the relevant clause is a reasonable professional request. Frame it as wanting to make sure the work you are leading maps clearly to what matters for the deal. If the founder resists, that itself is useful information about how much authority has genuinely been delegated.
Once you have the language, read it with three questions in mind. First, are the milestones activity-based or outcome-based? Second, what is the measurement mechanism and who controls it? Third, what timeline is tied to the payout? Activity milestones measured by a third party leave little room for interpretation. Internally-reported outcome metrics are more flexible but also more exposed to challenge during due diligence.
When should you push back on what the milestones require?
Pushing back on earn-out milestone targets is legitimate when the language as written would incentivise the wrong work. AI milestones that reward activity, such as tools deployed, pilots launched, or user activation counts, rather than measured impact can produce exactly the shallow adoption that M&A advisors flag as adding little to exit value and that due diligence teams learn to discount quickly.
The scope of the pushback matters. You are questioning whether specific milestone language will produce the result the founder actually needs, not the exit plan itself or the deal structure. That distinction makes the conversation easier to initiate.
Ask yourself whether hitting the milestone as written would produce something an acquirer or investor would read as genuine operational capability. BCG research on AI adoption found that companies capturing meaningful value embed AI across multiple workflows connected to business decisions, rather than in standalone projects. MIT research suggests only around 5% of generative AI pilots achieve rapid revenue impact; the gap is almost always a workflow integration problem rather than a technology one. A due diligence team familiar with that research will look for exactly that distinction.
Push back by proposing outcome equivalents rather than refusing the milestone structure. Replace an activity target with an outcome-based alternative that is still measurable and timebound, but tracks something that will survive scrutiny. Founders who understand the distinction, particularly those close enough to exit that due diligence is a near-term reality, will generally accept the reframe.
What does this connect to across the broader mandate?
The earn-out milestone question is one layer of a broader challenge every delegated AI lead eventually faces. How to produce AI adoption that serves both the exit date and the business that comes after it is a question with two answers that usually, but not always, point to the same work. Holding both in view under the pressure of a payout deadline is the real skill the situation asks for.
Three related threads are worth carrying into any earn-out situation.
The first is the founder-dependency discount. M&A advisors flag this as one of the most persistent drags on exit value in owner-managed businesses. Buyers apply material discounts when the founder’s presence is required for operations, decisions, or client relationships to function. AI that genuinely reduces this, by documenting the founder’s judgment, automating routine decisions, and systematising what was informal, helps close that discount. AI that creates a new dependency on the founder’s personal involvement in tool selection or process sign-off does not.
The second is delegation quality. Research on technology adoption consistently finds that visible, sustained executive sponsorship is among the strongest predictors of whether AI actually changes how a business operates rather than sitting alongside it. Spencer Stuart research on CEO-level AI engagement frames this as the delegation-versus-abdication distinction. Full founder withdrawal removes the strongest signal to the organisation that adoption is a genuine operational shift, not a project assigned to someone else.
The third is the measurement gap. BCG research identifies the difference between deploying AI and capturing value from it as primarily a workflow integration and measurement problem. Milestones that miss this reward the wrong activity. Getting measurement right from the start, rather than retrofitting it during due diligence, is the more reliable path to an exit value that holds up under scrutiny.
The earn-out is the founder’s mechanism. Your mandate is wider. Keeping that distinction clear is what allows you to serve both.
