The accounting roll-up rests on cross-sell and margin expansion. AI is about to test both, and most platforms are not measuring either one well enough to know how they will hold up.
Private equity has reshaped the top of the accounting profession in five years. The investment case depends on numbers few platforms can see today.
About half of the 30 largest U.S. firms have accepted outside capital, and ten of the top 20 are, or will soon be, owned by private equity. Deal volume set a record last year and is on pace to do so again.
The investment case usually rests on two assumptions. The first is cross-sell: acquired firms bring tax and audit clients who will buy advisory, technology, and wealth services from the combined platform. The second is margin expansion: shared services, pricing discipline, and technology will make each hour of work more profitable. AI is about to test both.
Most PE-backed accounting firms operate under an alternative practice structure. A licensed CPA firm owned by partners performs attest work, while a separate services company, owned in part by the sponsor, provides tax, advisory, and everything else. Independence rules keep the two apart.
That separation is necessary, but it splits the client. One business may appear in an attest system, a tax system, and an advisory CRM, often inherited from different acquired firms with different client IDs and different definitions of what counts as a client. A platform cannot measure cross-sell if it cannot see the same client across those systems, and it has to build that view without breaching the attest wall.
The simplest test of the cross-sell thesis is to measure client retention twice. Under a narrow definition, a client is retained if it keeps buying the service it originally bought. Under a broad definition, a client is retained if it keeps buying anything from the platform. The gap between the two is the cross-sell the investment case depends on.
The cross-sell thesis is working, and the platform has evidence to show at exit.
The platform is a collection of firms that share a logo, and the growth story will be hard to defend at exit.
AI is compressing the hours required for tax preparation, review, documentation, and parts of audit. For a firm that bills by the hour, getting more efficient means billing less. The margin gain the sponsor underwrote shows up only if pricing moves away from hours toward fixed or value-based fees.
That shift will not happen all at once, but the pressure will build within a typical hold period, and firms will need to plan for it now. The platforms that redesign pricing ahead of the change will keep the efficiency gain. Those that wait will hand it to clients.
There is a second, less discussed effect. Clients are adopting AI too. As they automate their own bookkeeping, close, and first-pass tax work, they will come to their accountants with narrower requests. All else equal, narrower scopes mean smaller engagements and lower realization, even if the firm’s own productivity is unchanged.
Together, these two forces mean that hours, realization, and scope per engagement are now leading indicators of value, not just operating statistics.
The same questions apply to consulting, engineering, legal support, and other professional services roll-ups. Each depends on clients buying more services after an acquisition, each is built on billable time, and each faces clients who can do more of the work themselves. Accounting is simply further along, because so much capital has moved into it so quickly.
Each acquired firm arrives with its own practice-management system, client IDs, and definitions, and the attest wall keeps the same client in separate systems by design. The foundation is a governed data layer: a semantic layer and a shared model of clients, engagements, and service lines that holds across every acquisition and respects independence rules.
On top of it sits the economics. That means estimating how much retention cross-sell actually creates once client size and tenure are accounted for, separating the firm’s own AI efficiency from narrower client scopes when realization moves, and modeling how a shift from hourly to fixed or value-based fees would change margin by service line. A dashboard is only the last step, where that work becomes visible to the people who act on it.
One model of clients, engagements, and service lines across every acquisition, with the attest wall designed in.
Retention created by cross-sell, AI efficiency versus narrower scopes, and margin under new fee models.
Governance, a semantic layer, AI, and training let practice and finance teams build their own analytics.
The same foundation also changes who can do the analysis. With clear data governance in place, defining who owns each data domain, who stewards its definitions and quality, and who maintains the systems as custodian, the business no longer has to queue every question behind IT. A well-built semantic layer, AI tools that work against it, and the right training let partners, practice leaders, and finance teams build the analytics and tools they need themselves. IT, in turn, can concentrate on keeping systems reliable and available.
For a platform still absorbing acquisitions, that is often the difference between measuring cross-sell every quarter and measuring it once, for the exit.
The accounting roll-up story can be a strong one at exit. The platforms that can show measured cross-sell, and a pricing model that keeps AI’s gains rather than giving them away, will have the evidence to support it.
A short conversation is enough to see which of these measures your data can support today, and where the foundation needs work.
A read on whether a platform’s cross-sell and pricing story will hold up at exit.
Compare NotesOne governed view of clients across every acquired firm, and the economics to price for AI.
Map the Platform