Search for the best CRM for accountants and you will find two kinds of article: vendor roundups that place their own product at the top, and comparison tables that list twelve platforms without telling you which problem each one solves. Both skip the question that actually determines the answer.

A CRM does one of two jobs in an accounting firm. It either manages the pipeline of people who are not yet clients, or it manages the workflow of people who already are. Most practice management platforms are excellent at the second job and thin at the first. Most sales CRMs are the reverse. Firms that buy the wrong one usually conclude that CRMs do not work for accountants, when the real issue is a mismatch between tool and task.

What follows is how we assess this with the firms we work with, including the cases where we tell them to keep the spreadsheet.

Practice management platforms such as Karbon, Senta, Pixie, BrightManager and TaxDome are built around the client lifecycle after signature. Recurring jobs, deadline tracking, document requests, AML checks, e-signature, portal access, billing. Their CRM module is usually a light layer for storing prospect records and sending the odd email.

Sales CRMs such as HubSpot, Pipedrive or Salesforce are built around the opposite half of the relationship. Lead source tracking, pipeline stages, follow-up sequences, call logging, conversion reporting. They know nothing about a VAT deadline and they do not care.

The practical consequence is this. If your problem is that jobs slip, deadlines get chased manually and clients email documents into a black hole, a practice management platform solves it. If your problem is that enquiries arrive and you cannot say how many converted, where they came from, or which ones went cold because nobody followed up on day four, a practice management platform will not solve it, no matter how good the software is at everything else.

Some firms genuinely need both. That is a normal outcome, and it is cheaper than forcing one system to do work it was never designed for.

We say this to more firms than you would expect. If you take fewer than around eight enquiries a month, all from referral, and one person handles every one of them, a spreadsheet with six columns will hold your pipeline perfectly well.

The columns that matter are: name, source, date of first contact, stage, date of last contact, and next action date. Sort by next action date every Monday morning. That is a pipeline system.

A CRM earns its place when one of three things becomes true. Enquiry volume exceeds what one person can hold in their head. More than one person touches an enquiry between arrival and proposal. Or you start spending money on acquisition and need to know which channel produced which client.

Until then, the CRM will not increase your conversion rate. It will give you a place to record the fact that your conversion rate is unknown. Firms often buy software hoping it will impose a process they have not yet defined. It does not work that way round. Write the process down first, run it manually for a quarter, then automate the version that survives contact with reality.

Across the firms we have built acquisition systems for, the single largest gain has come from speed of first response, not from the sophistication of the tool holding the record.

An enquiry that gets a reply within five minutes converts at a materially different rate to one answered the next working day. That is not a CRM feature, it is an operational decision. The CRM only helps because it can send the acknowledgement automatically, alert whoever is on duty, and book the call without an email exchange.

The second gain comes from follow-up persistence. Most firms stop after one attempt. A structured sequence over two to three weeks recovers enquiries that would otherwise be recorded as lost, and that sequence has to be automatic or it will not happen during January.

The third is qualification. One firm we work with, Annabel, runs an AI screening layer on the enquiry form so her calendar only carries opportunities worth the hour. Fewer calls, higher value per call. None of this depends on which brand of CRM sits underneath. It depends on whether the sequence has been designed, and whether anyone is accountable for it.

Before signing anything, run these:

• Data export. Ask specifically what you can extract if you leave, in what format, and whether email history and file attachments come with it. Vendors who answer vaguely are telling you something.
• Integration with what you already run. If it does not connect cleanly to your bookkeeping software, your email, and your calendar, someone will end up rekeying data, and that person will eventually stop.
• Who administers it. Every CRM needs an owner inside the firm. If nobody has time to maintain pipeline hygiene, the system degrades within two quarters and you will blame the software.
• Migration effort, honestly costed. Moving client records, templates and recurring jobs takes real hours. Ask for a realistic estimate in days, then assume it is optimistic.
• UK compliance fit. AML workflows, MTD-linked deadlines and GDPR handling vary considerably between platforms, and US-built tools sometimes treat these as afterthoughts.

Trial with real data, not the demo dataset. Load thirty actual client records and one live enquiry and see whether the system tells you anything you did not already know.

If your priority is running the practice and your pipeline is small, choose the practice management platform your team will actually open every day and accept that its CRM module is basic. Karbon, Senta, Pixie and TaxDome all clear that bar for different firm shapes, and the deciding factor is usually workflow style rather than feature count.

If your priority is growth, and particularly if you are spending on search or paid acquisition, you need a system that tracks source, stage, response time and conversion by channel. That means a proper sales CRM sitting alongside practice management, with the handover point clearly defined at proposal acceptance.

The best CRM for accountants, in other words, is whichever one matches the constraint you are currently hitting. Firms that grow past twenty enquiries a month almost always end up running both, connected rather than merged. We build that acquisition layer for firms and connect it into whatever practice software they already run, because ripping out working compliance infrastructure to fix a marketing problem is a poor trade.