The hidden cost of running a 30-person firm without a real CRM

Most 30-person accounting firms believe they have a CRM. They have a Xero contact list, a referral spreadsheet, an intake inbox, and notes spread across Outlook and the practice manager. The cost of this is rarely calculated. We do the maths the way we do on a diagnostic: hours per week per person, at fully loaded cost, before counting the deals that fall through the gaps. At one Melbourne firm a bookkeeper was losing 40% of her week to fixing other people’s data entry. Nobody was surprised.

What “having a CRM” usually means in practice

When a partner at a 30-person firm tells us they have a CRM, we ask which system they mean. The answer is almost never one system.

It is usually the contact list inside Xero, plus the client list inside XPM or Karbon, plus a referral spreadsheet a partner started in 2019, plus a shared inbox at intake@firm.com.au, plus the marketing list someone exported to Mailchimp two years ago and has been topping up by hand. The firm calls this collection “the CRM” because they need a word for it, and because each piece does something a CRM is supposed to do.

The reason it gets called a CRM is the part that costs money. Once a thing has a name, the firm assumes it has an owner, an audit trail, and a definition of what is in it. None of those exist. The owner is “whoever updates it.” The audit trail is the email chain. The definition of what is in it is whatever the last person typed.

A real CRM is one place where the firm’s view of a client lives. Not the ledger’s version. Not the practice manager’s. The firm’s. The relationship, the history, the next conversation, the partner who owns the relationship, the children who are about to inherit the family group. That object does not exist in Xero, in XPM, in Karbon, or in Outlook. At most firms we sit down with, it lives in someone’s head and gets reconstructed by hand on the morning of the partner meeting.

The five places client data lives in a typical firm

A diagnostic always starts with the same exercise. We ask the partner to list every place client data lives. We get five answers, in roughly the same order each time.

The ledger comes first, because it is the system everyone agrees is canonical for the books. Xero or MYOB. Names, ABNs, billing addresses, invoices, paid status. The ledger is excellent at being the ledger and is treated, incorrectly, as the source of truth for “who is the client.”

The practice manager comes second. XPM, Karbon, or a Karbon-XPM hybrid the firm has not finished migrating. This is where the work lives: jobs, deadlines, time entries, lodgement status. Client records duplicate the ledger’s records but are not identical. The phone numbers drift. The trading name and the legal entity get filed differently. The contact who signs the BAS is recorded here but rarely matches the contact recorded in the ledger.

The document store is third. FYI, SuiteFiles, SharePoint, or a Dropbox folder structure that predates everything. The folder name carries a client identifier. The identifier rarely matches the practice manager’s client code, because the document store was set up first.

The fourth place is the partner’s email. Outlook, mostly. Conversations with the client, attachments the client sent, decisions the partner made on the phone and never wrote down. The Smith family’s BAS strategy for the 2024 financial year exists in a thread between two partners and the client’s son. Nothing else in the firm knows about it.

The fifth place is the spreadsheet. There is always a spreadsheet. It is the referral tracker, the fee-setting model, the quarterly capacity planner, the partner-bonus calculation, the new-business pipeline, or some combination. It was started by one partner, owned by no one in particular, and is now the only document that answers a question the partners ask every month.

Five places. None of them know about the others. The work of keeping them aligned is the work that is rarely costed.

The hours-per-week calculation, by role

A diagnostic produces an hours-per-week figure for each role in the firm. The numbers are calibrated against actual conversations with the team, not pulled from a benchmark report.

A bookkeeper at a 30-person firm typically loses 12 to 16 hours a week to data hygiene. Fixing client names that arrived through three different systems with three different spellings. Reconciling a phone number a partner updated in Outlook against the one the practice manager has. Re-keying details from the new-client form into XPM, then into Xero, then into FYI’s folder structure. The bookkeeper at Fresh Networking we mentioned at the top was at the upper end. 40% of a 38-hour week is 15.2 hours, every week, on a problem nobody has acknowledged is a problem.

A senior accountant or partner loses 3 to 6 hours a week to the same shape of problem, at a much higher cost per hour. The work is different. They are searching for context that should be one click away. “What did we do for the Smith family last year?” sends them through XPM, FYI, and an old email thread. “Has Wong Brothers been onboarded properly for the new financial year?” requires three places be checked. The hours rarely register as wasted. They show up as Friday afternoons that vanish.

The practice manager or operations partner loses 4 to 8 hours a week to reporting. Producing the partners’ weekly capacity view. Collating the new-business pipeline. Writing the BAS pipeline view from a SUMIF across the lodgement system, the practice manager, and a spreadsheet a partner maintains. The hours are largely spent moving data between formats so a question can be answered. The question is asked again next week.

The intake person, where one exists, loses 2 to 4 hours a week to onboarding rework. New clients arrive through three doors: the website form, the partner’s email, and a referral conversation that ends with “I will get my assistant to send through their details.” Each door has a different format and creates a record in a different system first. The intake person reconciles them by hand.

At a 30-person firm with one bookkeeper, two operations or admin staff, and around eight senior fee earners, the diagnostic adds up to roughly 50 to 70 hours of CRM-shaped work every week. At fully loaded cost, that is between $4,000 and $6,500 a week, before counting any error correction or rework. Annualised, $200,000 to $300,000 of effort goes into keeping the five places aligned. The firm has been spending it for years and has not seen it on a P&L line, because no line was ever created for it.

The compounding cost: the deals that fall through the gaps

The hours are the visible part. The deals are the part that does not show up in any spreadsheet, because the firm never finds out about most of them.

A referral from a financial adviser comes in by email to a partner who is on holiday. The partner’s auto-reply names someone else in the firm. The financial adviser does not follow up because they have other accountants to refer to. The firm does not know it lost the referral. There is no record. The financial adviser has stopped sending work this firm’s way and nobody can tell you when it started.

A long-term client mentions, on the BAS call in March, that her brother is selling his business in October and will need an accountant. The partner makes a note in his email. The note never makes it into a system. October arrives. The brother gets a different accountant because nobody at the firm remembered to follow up.

A new staff member starts in May. Three of the firm’s most active referrers have drifted away over the last 18 months because the firm got slow on intake and never noticed. The staff member would not know to ask, because there is no list of “referrers we used to get work from.” The list exists in two partners’ heads and they have not compared notes in three years.

The compounding cost is the part that breaks the maths. The hours figure is uncomfortable. The deal figure is harder to face, because it points to revenue the firm assumed it had earned by being good at the work. A 30-person accounting firm in a Melbourne suburb typically loses two to five new-client opportunities a quarter to gaps in this layer. At an average lifetime value of $15,000 to $40,000 per client, that is $120,000 to $800,000 a year that the firm does not see, because it cannot count what it never received.

We are deliberate about how we present this in a diagnostic. The hours number is provable. The deal number is an estimate, framed as a range, with the assumptions written down. The point is not to scare anyone. The point is that the second number is bigger than the first, and the firm needs to look at both before it decides what to do.

What a real CRM has to do to be worth more than the sum

A CRM is worth doing when it removes more cost than it creates. The brief, on a 30-person firm, has a specific shape.

The CRM has to know which person at a client is the right contact for which conversation. The signer for the tax return, the contact for BAS, the family member who handles ASIC correspondence, the spouse who asks for the year-end pack. This information exists across Outlook threads and partners’ memories. A real CRM models it as a structured field, set at intake, updated when the relationship changes, and visible to whoever is doing the work.

It has to carry the audit trail. Every change to a client record, who made it, when, and why. The audit trail is the difference between “we have a CRM” and “we have a list.” It is also the part that off-the-shelf tools handle inconsistently, because each one has a slightly different idea of what counts as a change.

It has to integrate with the systems the firm already runs, not replace them. Xero stays. The practice manager stays. The CRM sits underneath them and treats them as inputs. A change to the client name made in the CRM propagates to Xero and to XPM, with a timestamp. A change made in Xero is captured in the CRM with a note about where it came from. The firm stops having to choose which system is right when they disagree, because the CRM is the place that decides.

It has to give the partner one view. Not a dashboard with twelve tiles. A page, per client, that answers the questions a partner asks: where are we in the relationship, what is due, what was the last conversation, what is at risk. The view has to load in under three seconds and survive a partner who refuses to learn a new tool. If it does not, it gets the same fate as the last CRM the firm bought.

It has to be cheaper to maintain than the current arrangement. This is the test most firms do not run. The current arrangement has a cost. We have just spent five sections describing it. A real CRM that costs more to maintain than the five-system mess it replaces is just a different mess with a better label. The maths has to land in the firm’s favour at month 24 and stay there at month 60.

We run this calculation against firms in the diagnostic before recommending any build. The diagnostic puts the hours figure, the deal figure, and the maintenance figure on one page. Some firms walk away with the diagnostic and act on it themselves. Some come back when they are ready to look at a custom CRM that fits the firm. For partners and practice managers at Melbourne accounting firms specifically, the broader engagement model lives at our services for accounting firms page.

The firms that have done this work are not running a different CRM. They are running their existing tools with one place that decides what the firm thinks is true. The bookkeeper still exists. The hours she spends fixing data entry have moved from 15 a week to under 4. The other 11 hours are not a savings line item, because she was never billed out. They are time that has gone back to work the firm wanted her to be doing in the first place.