UK firms are recovering nearly a quarter of their capacity through AI and automation, yet many partners are seeing no margin improvement. We sit down with Sage experts to discuss why the traditional billing model is failing the modern firm and how to pivot from compliance processing to high-value advisory before the full weight of MTD arrives.
In Partnership With
Published in partnership with Sage. Based on Accountancy Age’s Leading Voice Broadcast with Chris Downing and Jack Choppin from Sage.
Most partners and senior managers in UK accountancy firms know this feeling well. The workflows are cleaner than they were twelve months ago, certain tasks that used to take hours are taking minutes, and yet the conversation in the partners meeting about margin improvement keeps getting deferred.
Around 25% of a typical accountancy firm’s time goes on chasing and reconciling data. Chris Downing, Director for Accountants and Bookkeepers at Sage, raised this figure during our recent broadcast, and the point was to ask what happens when AI starts returning that time to your practice, and whether your firm is positioned to capture it commercially. For most UK firms right now, the honest answer is that they are not.
Jack Choppin, Senior Accountant Success Manager at Sage, works directly with UK accounting firms every day, and what he sees consistently is firms adopting AI and filling the returned capacity with more of the same work. The accounts pile does not shrink, the client queries do not reduce, the deadline pressure does not ease, and the time comes back, gets absorbed, and the P&L stays exactly where it was. The billing model in most practices was built around hours and compliance outputs, was never designed to capture speed as a form of value, and so when AI makes the work faster, the efficiency passes straight through the firm and lands with the client.
Jack drew a direct parallel with cloud accounting adoption during the broadcast, and it is the most instructive comparison available to the profession right now. When firms adopted receipt capture tools like AutoEntry and Dext, and when bank feeds became standard, delivery times reduced significantly and the technology worked exactly as promised. What many firms did not do was adjust their pricing to reflect the improved service, so the client got faster turnaround, better organised records, and more accurate data, while the firm continued charging the same fee it had always charged, now delivered at lower cost but without a corresponding improvement in margin.
AI adoption is following the same pattern, and the scale of the opportunity being missed is considerably larger. If your practice is using AI to process compliance work faster and your pricing has not changed, you are delivering a better service for the same money, which is good for your clients but does nothing for your firm. Both parties should benefit from efficiency gains, and making that happen requires a deliberate decision to capture the value rather than generate it and pass it on.
Chris made a point in the broadcast that is worth taking at face value before reaching for a broader AI strategy. Firms that try to move into advisory and value-based pricing while their underlying workflows are still inconsistent and deadline-driven find that the commercial model does not hold. The gains stay partial, the review burden grows, and the promised improvement in margin stays out of reach. Getting the efficiency baseline right is the prerequisite, not a stepping stone to something more important.
The practical starting point he suggested is deliberately simple. Look at your client list, find the limited company accounts being filed six, seven, eight, nine months after the year end, and ask what it would take to bring that window down to three or four months. That shift reduces the deadline pressure that prevents senior staff from doing anything other than processing, creates consistent and predictable capacity across the year, and puts your firm in a position to have a different kind of client conversation, one based on current numbers rather than data that is already most of a year old. Start there, build the consistency, and then decide what the capacity is for.
The question the broadcast kept returning to is the one most firms are not yet answering clearly. If AI is reducing the time spent on compliance work, what is the client paying for?
For most of the profession’s history, the answer to that question was implicit. Clients paid for the time it took to prepare accurate accounts and returns, delivered on time, by professionals they trusted, and that model worked when the time involved was genuinely significant and the complexity was genuinely high. AI is changing both of those variables rapidly, and clients, even if they cannot articulate exactly why, are beginning to sense that the relationship between price and effort is shifting.
The firms building sustainable commercial models in this environment are the ones that decided clearly what they are selling now that compliance is no longer the primary deliverable. They are selling confidence, clarity, and the ability to make better business decisions faster, which carries a different price point and requires the firm to have answered the question rather than deferred it.
The partners meeting conversation about margin improvement does not have a technology answer. The technology is already working. The question is whether your model is set up to benefit from it.
