Whether a buyer reads your business as a services company or as tech-enabled services depends on what is documented, owned and repeatable before the process starts. The operating work is the same either way. How it gets capitalised is not something operating work can promise.
The operating work you do before a majority recap shapes whether a buyer reads the business as a services company or as tech-enabled services. Same revenue line, same EBITDA, a different operating-model story. What that story is worth is a market outcome: it depends on the buyer, the sector, the period and the comparable transactions, and we do not put a number on it. Earlier versions of this article did, and the figure was neither sourced nor modelled transparently, so it has been removed rather than re-sourced. This post is the dual-lever play that supports the tech-enabled-services narrative, sequenced for a six-month transaction window.
The multiple math
A traditional services business sells on EBITDA. The buyer underwrites the cash flow, applies a sector comp multiple, and gets a number. In professional services, the comp range is typically 5x to 8x EBITDA depending on margin profile, growth rate, and customer concentration. Most mid-market services rollups land in the 6x to 7x range.
A tech-enabled services business sells on revenue. The buyer underwrites the operating model and the durability of the margin profile, applies a revenue multiple, and gets a different number. The comp range is typically 1.5x to 3x revenue, which on a healthy services-business margin structure works out to roughly 9x to 13x EBITDA equivalent.
What the recap turns on is not a software product. It is whether the operating model reads as a labour-arbitrage cost play or as an operating system that compounds. Buyers and their advisers can tell the difference, and they underwrite accordingly. Which of the two they conclude is inside your control. What they pay for the conclusion is not.
What buyers underwrite
Buyers paying a revenue multiple are underwriting three things. The operating model is durable. Not a one-time efficiency gain. A repeatable, documented system that produces predictable margins year over year. Virtual Employees in production with documented governance. The intelligence accrues to the company. Not to a vendor. The prompt libraries, fine-tuning data, exception patterns, and operational residue live on the company’s books, inside an entity the buyer can take operational ownership of. The team is the right shape. Not a thirty-person back office with AI tools layered on top. A redesigned org chart. Smaller. More senior. More expensive per head.
If any of the three is missing, the story falls apart in diligence. The most common kill is the third. The CEO commissioned an AI initiative. The vendor delivered tools. The team adopted them. Headcount stayed flat. The org chart did not change. The buyer’s diligence team looks at the operating model and sees a 2019 services business with software bolted on. Multiple compresses back to EBITDA. The CEO does not understand why.
The dual-lever sequencing
Both levers, sequenced right, support the tech-enabled-services story. The order matters. Lever 1 ships first: the AI-native org chart. The redesign is what makes the story credible. The before is a 30-person back office; the after, illustratively, is 6 senior humans plus 14 governed AI roles at comparable throughput. The shape change is the proof point, and it is documented rather than asserted. Lever 2 ships underneath: the offshore team inside the entity. This is the cost story that supports margin durability. The COPO entity puts the senior India team inside the company’s own books.
If the order reverses (offshore team first, AI second), the buyer’s banker has a harder time. The narrative looks like cost arbitrage with AI on top, which is the EBITDA-multiple story. The work to reposition that to a revenue-multiple story takes another 12 to 18 months that the company does not have.
The six-month timeline
Most majority recap transactions run on a six-month window from initial conversations to close. The work has to fit inside that window.
Months 1 to 3: ship the narrative. A three-to-five-week Blueprint scopes the AI-native org chart. The redesign is documented. The Virtual Employee roster is finalized with scope, governance, and unit economics. The first one or two Virtual Employees go live in the function with the highest visibility. The org chart change is documented in the company’s operating-model materials before the banker is engaged. In parallel, the COPO entity standup begins. Hiring starts on the senior India roles. Facilities and IT get scoped. The legal entity gets registered.
Months 4 to 6: ship the proof. The Virtual Employees that went live in months 1 to 3 now have three to four months of production data. The senior India team is hired and operational. The combined org chart is functioning as designed: 6 humans plus 14 Virtual Employees per function, with persistent memory architecture, governance documented, audit trails in place. The banker is now engaged. The pitch deck has the new operating-model page. The diligence room has the audit trails, the token-cost models, the governance framework, and the entity structure. The buyer’s diligence team underwrites a revenue multiple. The story holds.
This timeline is compressed. It works because the Blueprint front-loads the design work, the Flexi engagement model lets the first Virtual Employees go live in 30 to 60 days without a full COPO standup, and the COPO entity comes online in parallel. We have run this on a six-month window. It is the work.
What kills the narrative
Three failure modes kill the revenue-multiple story most often. Vendor-rented AI. The CEO bought AI from a third-party platform. The diligence team asks who owns the AI layer and the answer is "the vendor." Multiple compresses. AI tools without org chart change. The team uses them. The org chart did not change. The story is "we made our existing team faster," which is an EBITDA story. Pilots that never reached production. Demos. Slide decks. None of the systems made it to production with governance and persistent memory. The buyer assumes the AI strategy is aspirational.
The fix for all three is structural. The org chart has to be redesigned. The Virtual Employees have to run inside the company’s entity, not a vendor’s. The systems have to be in production with governance and persistent memory engineered in.
Where to start
The Blueprint is the entry point. Three to five weeks. Paid engagement. The deliverables: the AI-native org chart for the function with the highest narrative leverage, the Virtual Employee roster with unit-of-work pricing and governance framework, the COPO entity structure, the joint unit economics, and the six-month rollout plan calibrated to the recap timeline. After the Blueprint, the work executes. Months 1 to 3 ship the narrative. Months 4 to 6 ship the proof. The banker engages. The buyer underwrites. The multiple holds.
The earlier you start, the more multiple is on the table.