What's Changing in the Stack You Own

Every Head of Technology at a mid-sized European accounting and CFO advisory firm has had the same Tuesday. The fortnightly practice operations meeting opens at half past eight. Four things on the screen. The fourth compliance RFP lost to Aiwyn this quarter, and the pitch partner's note: "Their unit economics are below our junior rate." A senior tax partner's retirement timing email arrived over the weekend, earlier than anyone wanted to hear. Your second-best advisory lead took the Deloitte AI-and-advisory package on Friday at a compensation number structurally out of your firm's range. And the Managing Partner's tuck-in proposal lands on the agenda: €5M of firm capital for a sub-scale peer, three-year practice-equity consequences.

You have run this function for seven years. The pattern used to be simple. Keep Karbon and CCH working. Ship the Microsoft upgrade on time. Pilot the new AI tool when the practice heads ask. Hold uptime. Meet the Managing Partner on licence cost. The seat was understandable.

Something changed. The compliance RFPs you are losing are losing on unit economics AI-natives can hit and your firm cannot. Aiwyn, Botkeeper, Blue J Legal, TrullionAI, Intuit's advisor layer. The advisory work that is supposed to replace the compliance book requires the client-specific intelligence your senior partners carry in their heads, and the retirement cohort starts in eighteen months. Ninety-five percent of corporate AI deployments produce no measurable P&L impact. Your firm sells owner-managed clients a practice-intelligence transformation every week. You have not yet built one for yourselves. The cobbler's children have no shoes.

You are not running one tech-spend envelope. You are running two, and only one is on your scorecard. One funds what your practice already uses: Karbon, CCH, Wolters Kluwer, Microsoft 365, the AI-compliance layer to defend the shrinking book. The other funds what has to exist by 2028: the partner-IP extraction programme, the firm-owned client-intelligence graph, the agent architecture that encodes the methodology, the integration depth that outlasts three vendor swaps. On one hurdle rate the first wins every quarter. On one scorecard the second does not exist.

This briefing covers what is happening in your sector, what it means for the practice-technology function you run, and the three questions your partnership should answer before the next council meeting.

The squeeze on mid-market accounting and CFO advisory is three-directional and AI is accelerating each side. From above, the Big 4 are shipping AI-augmented audit and tax tooling that compresses advisory margins. Deloitte Consulting, PwC, EY, and KPMG are each walking into the same owner-managed clients you serve, with AI-and-advisory offers priced for a different cost structure. From below, the PE-backed consolidators (Aspia at SEK 3.5B, Accountor, Forvis Mazars) are absorbing compliance at scale. Their AI-enabled compliance factories run at unit economics your partnership model cannot match. From the side, AI-native point vendors are absorbing individual billable-hour streams: Aiwyn on audit, Botkeeper on bookkeeping, Blue J Legal and TrullionAI on tax research, Intuit's advisor layer and Digits on SME compliance.

The compliance revenue base that funded the partnership model for forty years is being automated and competed away on three fronts simultaneously. The advisory revenue that is supposed to replace it requires the client-specific knowledge partners accumulated over twenty years. That knowledge is approaching retirement. The 2028-2030 cohort is not theoretical. Most mid-market firms have never mapped which partners carry how much of which book, and those who have quietly stopped sharing the map with the partnership council because the numbers are unsettling.

Your practice stack shipped on Karbon and CCH is on a five-year commoditisation curve. Every competitor buys from the same vendors. The moat lives in what your firm has encoded on top of the stack, in infrastructure no vendor can see.

The industry's speed mismatch is the third force. AI-natives build from a clean sheet with no partnership politics, no twenty-year client book to protect, no retirement cohort to manage. They ship on unit economics that compliance-heavy mid-market firms cannot match on their current cost structure. Your firm cannot pivot at their speed on their terms. What you can do is identify what is genuinely defensible (the twenty years of partner knowledge, the client relationships, the methodology) and stop defending what is not. The firms already moving are quiet about it. The ones renewing Karbon and calling it an AI strategy are the ones the PE consolidator writes a polite email to in 2027.

Your build engine. Every vendor in the stack is shipping AI this year. Karbon adds workflow copilots. CCH adds tax-research AI. Wolters Kluwer adds document automation. Microsoft adds Copilot. Your team is consumed by integration and configuration. Velocity on renewals is up. Velocity on anything the firm owns is flat. Twelve months in, your practice is more AI-enabled on the dashboard and no more defensible on the substance.

Your platform and data. You hold twenty years of client files, partner notes, WIP records, email threads, and meeting transcripts inside Karbon, CCH, SharePoint, and three partner-owned filing systems nobody migrated. That corpus is the firm's most valuable asset. It is structurally inaccessible to any agent the firm has deployed. The AI-compliance platform quote is on your desk. Necessary for meeting the next RFP on price. It encodes nothing the firm owns.

Your product thesis. Your Managing Partner asked for the firm's three-sentence technology thesis last quarter. Your answer was the one from three years ago: "We run a best-in-class practice-management stack and we pilot AI where it makes sense." That reads as "we buy what every competitor also buys." The thesis that matters is about what the firm builds on top of the stack, not what it buys inside it.

Your R&D bench. Your internal practice-technology team is three or four senior admins plus a handful of integrations engineers. Their day is vendor tickets and the next rollout. The one person who could run the partner-IP extraction programme left for a Big 4 transformation practice in Q3. The Deloitte package was structurally above what the firm could match on compensation alone.

Your board-level authority. Heidrick February 2026: thirty percent CIO, twenty-three percent CDAO, twenty-one percent CAIO, one percent COO own AI strategy in Europe. Your seat is on the list only if you claim it. The Managing Partner is considering a Chief AI Officer or a Big 4 transformation partner in Q3. Your Q1 boundary agreement lands before that conversation, not after.

Your sector is repricing from below because AI-natives are shipping compliance work at unit economics your partnership model cannot match. Your function is compressing under vendor-AI consolidation where every tool in the stack is trying to become the AI orchestration layer. Same force, two altitudes. But there is a third thing happening that most mid-market advisory-firm CTOs have not named yet.

The moat that matters in 2028 is not in the vendor stack. It is in the layer underneath it that only your firm can build. Every advisory firm your size buys from the same four or five vendors. Karbon or Ignition for practice management. CCH or Wolters Kluwer for tax. TaxDome or Xero for client-facing work. Microsoft 365 for everything else. Your "stack" is a commodity. What is not a commodity: the twenty years of client-specific decisions your senior partners have made, the owner-manager family succession context they carry in their heads, the methodology that distinguishes how your firm advises from how Grant Thornton advises, the client-intelligence graph that captures what every engagement taught the firm. A two-person AI-native vendor cannot build that. Big 4 cannot replicate it without twenty years of your client relationships. Aspia cannot consolidate it. You can build it, and the window closes when the 2028-2030 retirement cohort walks.

This is the argument your advisory partners have delivered to owner-managed clients for a decade. Every senior partner in your firm has said to a family-business CEO: "Your three senior people carry the knowledge. When they retire, it walks out. Let us help you build the infrastructure that holds it inside the firm." The partnership has never turned that flashlight on itself. The cobbler's children have no shoes. Your retiring cohort carries more knowledge than any of your clients' retiring founders, and the firm has not extracted a single partner's twenty years of reasoning into infrastructure you own.

You have twelve to eighteen months before the AI-native cost curve and the retirement cohort arrive at the same table. The CTOs who claim the authorship of this decision in Q1 build the infrastructure that survives both. The ones who wait until the Managing Partner hires a transformation partner inherit a seat that has narrowed to stack-admin, and a scorecard that stopped mattering the quarter the fourth compliance RFP went to Aiwyn.

Every advisory-firm CTO has had the AI conversation. Most are making the same four decisions. A smaller group is making different ones.

Three questions every advisory-firm CTO is tracking. The third is the crux. The first two are how you earn the right to answer it.

Four concrete moves any advisory-firm CTO can start this quarter. Each addresses one sub-function. Pulled together, they become one architecture the firm builds on.

The advisory-firm CTO has four places to go. None can do what we do.

The Big 4 tech-transformation arms ship a standardised AI-practice playbook across a sector book. Their cost structure requires reusability. Your practice is idiosyncratic. A reusable framework on an idiosyncratic practice is a deck your Managing Partner cannot defend.

The practice-management platform vendors have a renewal quote on your desk. Karbon, Ignition, CCH, TaxDome, Wolters Kluwer. Their AI layer is a seat-penetration play. Telling you to build outside their stack is structurally impossible for them.

The AI-native point vendors are what is repricing your compliance book. Aiwyn, Botkeeper, Blue J Legal, TrullionAI, Digits. Asking any of them how to build the moat against themselves is not a conversation that ends well.

A fractional CIO or outsourced IT provider operates the stack against uptime and cost. That is the job they are scoped for. It is not the capital-allocation document your partnership council is waiting on.

The four options you have already looked at. None can run the practice-intelligence build the seat requires. GRAIL is the fifth.

We start with your practice-technology leadership because the change cannot outrun the team's personal understanding of what is possible. Three workshops. Each delivers standalone value. Each earns the right to the next.

The questions where your practice-technology leadership cannot agree are the ones worth an hour on the agenda.