HubSpot, Pipedrive and Salesforce model leads moving through stages towards a deal. Accountancy does not work like that. The work is recurring, deadline-driven, regulatory: onboarding, AML refresh, MTD VAT, year-end, personal tax, payroll, Companies House. A CRM built around that cycle is a different artefact, and most practices have spent years bending generic tools to fit. We build the artefact, fixed-quote.
A typical pattern across UK accountancy practices: most have tried at least one generic CRM. Three months in, the CRM is abandoned, partially configured, with the deadline tracker workbook back on the practice manager's laptop running everything. The reason is not training, not change resistance. Generic CRMs model the wrong thing.
Leads enter a pipeline, move through stages, each with a probability, towards a deal. The dashboard shows expected revenue. When a deal is won it becomes a customer, and most of the CRM goes quiet because the deal is done.
A client signs an engagement letter. Then, every year forever, the same cycle runs: AML refresh, year-end accounts, corporation tax, personal tax for directors, VAT quarters, payroll runs, Companies House confirmation statement, annual fee review. Each piece has a statutory or contractual deadline, a job stage (records received, draft prepared, partner review, signed, filed), and time recorded against the fee. New work onboarding is occasional. The recurring lifecycle is everything.
A sales CRM has no native concept of "this client has a 31 January self-assessment deadline, records arrived on 14 December, the senior has the draft, partner review is queued, fee budget £450, time recorded £380, AML refresh due in March." That is a custom-fields kludge on HubSpot or Salesforce, unmaintainable within six months. On Pipedrive it is impossible because the pipeline metaphor is hard-coded into the UI.
The mission-critical artefact in almost every UK practice: a single Excel workbook on the practice manager's laptop. It is the client deadline tracker. It runs the practice. The columns, near-universal across firms:
Colour-coded: red overdue, amber within 30 days, green comfortable. The practice manager filters it every Monday. The partner uses it for the Friday WIP review. Juniors are told to update their lines but half do not, so the workbook drifts and the manager re-syncs it weekly.
It breaks predictably. 200+ clients and Excel is sluggish. Edit conflicts as soon as two people open it. The January peak is when it is most stale. Any "overdue across the firm, by fee-earner, by deadline type" view is a 90-minute pivot-table exercise. When the practice manager is on holiday, nobody else understands the formulas.
The CRM the practice needs is the digital version of this workbook, with the manual parts done by software, and the parts AI can take on (record-request drafting, deadline triage, AML refresh chasing, fee-leakage surfacing) done by AI.
These are illustrative shapes drawn from UK accountancy practice patterns. The fixed-quote bands are the real ranges we quote for builds of this shape; the workflows described are the patterns that fit each context.
Profile: sole principal, two seniors, three juniors, two admin. Around 350 active clients. On Excel for deadlines and IRIS for tax. Typical pain: January is a near-breakdown every year, the partner spends every evening on WIP review, 15 to 20 clients are billed at fees set five years ago that have never been reviewed. The shape that fits: a client lifecycle CRM with job-stage tracking, capacity dashboard against the January peak, fee-leakage surfacing. Typical band: £18,000 to £26,000 fixed quote, ten to fourteen weeks. Expected outcome: WIP review off the partner's evenings; fee leakage surfaced before renewal.
Profile: two partners, four seniors, six juniors, three support. 600+ clients. On Karbon, but only two of the team genuinely use it. Typical pain: Karbon licence costs around £18,000/year and the team will not adopt it. The shape that fits: replace Karbon with a custom client lifecycle CRM matching the firm's actual workflow, AI-drafted record requests, onboarding portal. Typical band: £35,000 to £48,000 fixed quote, fourteen to eighteen weeks. Expected outcome: Karbon cancellation saves around £18,000/year ongoing.
Profile: solo principal, one part-time admin. Around 90 clients, almost all owner-managed Ltds. Excel-only. Typical pain: principal is the bottleneck, admin spends two days/week on record-chasing, AML refresh is in arrears. The shape that fits: a focused client lifecycle tracker with AI-drafted record-request emails and automated AML refresh reminders. Not a full CRM, just the deadline-tracker workbook replaced with software. Typical band: £8,000 to £12,000 fixed quote, six to nine weeks. Expected outcome: admin time on record-chasing drops sharply and AML refresh stops drifting.
Profile: two partners, three seniors, four juniors, three support. Around 480 clients. On TaxDome for documents and IRIS Practice Engine for compliance. Typical pain: TaxDome works for documents but the practice runs a separate Excel for workflow. Three sources of truth, none complete. The shape that fits: a workflow layer over TaxDome and IRIS, pulling from both, adding the practice's actual lifecycle view. Typical band: £15,000 to £22,000 fixed quote, ten to thirteen weeks. Expected outcome: single live view of workflow across documents, compliance and lifecycle.
Profile: three partners, six seniors, seven juniors, four support. 800+ clients. On Senta (now Iris Elements). Typical pain: per-user pricing at 20 staff is around £24,000/year and rising, AI features are weak, the team's workflow has drifted from how Senta wants them to work. The shape that fits: a custom client lifecycle CRM with AML automation, onboarding portal, AI-drafted client comms, fee benchmarking dashboard. Typical band: £45,000 to £60,000 fixed quote, eighteen to twenty-four weeks. Expected outcome: Senta cancellation saves around £24,000/year; project pays back around year three.
Shared pattern: a practice management tool that was the right answer once becomes wrong at scale, either because per-user pricing has crept past the amortised cost of a custom build, or because the tool's opinionated workflow does not match how the firm actually runs.
Most practices are best served by one of the off-the-shelf tools. A custom CRM only makes sense in specific circumstances. Below is how we frame it in scoping conversations.
| Option | Typical cost | When it fits | When it does not |
|---|---|---|---|
| Senta (Iris Elements Practice Management) | £40 to £80 per user / month | Practices already on the Iris ecosystem, 3 to 12 staff, willing to adopt Senta's opinions about workflow | When the team will not adopt it, or when 15+ staff makes per-user pricing painful |
| Karbon | £50 to £100 per user / month | Practices that want the email-led workflow and the polished UX, 5 to 25 staff | Practices that find the workflow rigid, or where the per-user pricing breaks the model |
| TaxDome | £40 to £70 per user / month | Document collection and client portal are the primary pain; willing to live with TaxDome's job-management opinions | When workflow is the real problem and TaxDome's job model does not match it |
| Pixie | £40 to £80 per user / month | Smaller practices (3 to 10 staff) that want an opinionated, simple tool without the IRIS or Karbon overhead | Larger practices, complex workflows, or AI-led automation needs |
| IRIS Practice Suite (full) | £15,000 to £80,000+ per year | Established practices already deep in IRIS for tax, accounts production and compliance | Practices that do not already need the full IRIS stack |
| HubSpot, Pipedrive, Salesforce | £20 to £150 per user / month | Tracking new business pipeline only, as a layer on top of a real practice management tool | Running the recurring lifecycle work. They were not built for it. |
| Custom CRM (this category) | £15,000 to £55,000 fixed quote | 15+ staff, per-user pricing on the existing tool is painful, the team's workflow does not match any off-the-shelf tool, or AI automation is core to the next phase of growth | Sub-10-staff practices where Pixie, Senta or TaxDome would just work |
The honest recommendation in three quarters of scoping conversations: stay on the practice management tool you have, or adopt one of Pixie, Senta or Karbon if you are still on Excel-only. We will turn down the work where a £15,000 custom build is not the right answer.
Patterns that recur in scoping conversations. Each is a project we would refuse to build.
"Build us a CRM in Notion / Airtable." Both are great for prototyping and small teams. Neither is a CRM. By client 150 the latency, lack of native Xero or Companies House integration, and absence of proper access controls bite hard. Practices end up rebuilding in a real database eighteen months later, paying for two builds.
"Layer HubSpot Operations Hub over the practice management tool." Keep Karbon for delivery, add HubSpot for marketing. In practice the two systems duplicate client data and drift apart. You pay for two SaaS subscriptions and get a worse data model. The fix is one source of truth.
"Wait for Karbon / Senta / TaxDome to release the AI feature we want." Each vendor is shipping AI features, but none on your specific workflow on your specific timeline. If a specific automation is genuinely holding the practice back, building it ourselves is months faster than the vendor roadmap, and the practice owns it forever.
Real bands. Every project is scoped and fixed-quoted before work starts.
| What it is | Timeline | Fixed quote |
|---|---|---|
| Scoping (written delivery plan, fixed scope, fixed quote for the build, no commitment to proceed) | 1 to 2 weeks | £1,500 to £4,000 |
| Deadline tracker + fee pipeline replacement (Excel workbook to software, solo to 5 staff) | 6 to 10 weeks | £8,000 to £15,000 |
| Client lifecycle CRM (5 to 15 staff, full workflow, AI-drafted comms) | 8 to 14 weeks | £15,000 to £40,000 |
| Practice management replacement (Karbon/Senta out, custom in, with onboarding portal, AML, document handling) | 16 to 24 weeks | £25,000 to £60,000 |
| Tech partnership (we run, maintain, improve) | Monthly, no lock-in | £450 to £1,500 / month |
The question to ask: what is your current per-user-per-month spend on practice management software, multiplied by users, multiplied by years until you would switch? Most 15+ staff practices are surprised when they total it. A £40,000 custom build pays back in three to four years at that scale, and the software is owned outright. Adjacent: our app and product builds page covers the full custom-software shape.
I am Oliver Wakefield-Smith. More than twenty years as a developer, tech lead, architect and engineering lead, shipping software across satellite communications, UK broadcast media, sports media, water utilities, retail, industrial engineering and the public sector. That experience is now the engine behind a directed team of AI agents that does the building.
A CRM for an accountancy practice is not an accountancy domain problem, it is a workflow and data-model problem. The accountancy specifics (MTD, AML, ICAEW, ACCA, Companies House) are well-documented and learnable. The hard part is the shape of the software, which is what twenty years of delivery experience brings.
Adjacent: AML software for UK accountants (the workflow this CRM integrates with), where AI earns its place, and the tech partnership retainer.
HubSpot, Pipedrive and Salesforce are sales-pipeline CRMs. They model the world as leads moving through stages towards a deal. An accountancy practice does not run on a sales pipeline. It runs on a client lifecycle: onboarding, AML refresh, year-end, MTD VAT quarters, personal tax season, payroll, Companies House filings. The work is recurring, deadline-driven and regulatory. Forcing it into a sales CRM costs you a custom-fields nightmare, missed deadlines that fall between the cracks, and a system nobody on the team actually uses.
Practice management tools are closer to the right shape. Karbon, Senta (now Iris Elements), TaxDome and Pixie all model work as recurring jobs against client lifecycles. They are a better starting point than a sales CRM. But each one has rigid opinions about how work should flow, the per-user pricing gets expensive at 8+ staff, and the AI features are still mostly bolted-on. A custom build makes sense when the practice has worked around the limits of these tools for years and the workarounds are now the bottleneck.
A focused client lifecycle CRM for a 5 to 15 staff practice runs £15,000 to £40,000 as a fixed-quote build, delivered in 8 to 14 weeks. A smaller scope (replacing the deadline tracker and the fee pipeline workbook only) runs £8,000 to £15,000. A practice management replacement with onboarding portal, AML automation, document handling and AI-drafted client comms runs £25,000 to £60,000. Ongoing maintenance and improvement is £450 to £1,500 per month under the tech partnership.
A focused client lifecycle CRM ships in 8 to 14 weeks from kickoff. A full practice management replacement takes 16 to 24 weeks. The single biggest factor is not the build, it is data migration off the old system (Karbon, Senta, IRIS, Excel) and access to integrations (Companies House API, Xero, your tax software, your e-signature provider). If migration is clean and access is fast, we ship at the lower end of the band.
Yes. Xero and QuickBooks have well-documented APIs. Companies House Public Data API is free and reliable. HMRC integration depends on the service (MTD VAT is the most mature, MTD ITSA is rolling, agent services account is API-accessible). E-signature (DocuSign, Adobe Sign, Yoti) and ID verification (SmartSearch, Credas, Veriphy) integrate cleanly. We list every integration in the fixed quote.
Yes. Code, database, integrations, deployment, all of it. Runs on your own environment. If we part ways the system keeps running and another developer can pick it up. We will not build on proprietary platforms you cannot leave.
Tell us about your practice, your current tool, and what is not working. We will tell you honestly whether a custom CRM is the right answer or whether one of the off-the-shelf tools is. No charge for the conversation.
Email oliver@digitalsignet.com