Accounting & CPA Firms · Marketing & Reporting
The client everybody likes is losing you money and nobody has run it
They call constantly. They are pleasant. Their fee has moved twice in six years, both times by inflation-adjacent amounts nobody argued about. The work has quietly expanded to include three things that were never in the engagement letter, and the staffing mix has drifted upward because the client asks for the partner. Firm-level realization looks fine, so the question never gets asked at the client level.
The data to answer it is sitting in your time and billing system right now. What is missing is the two days of joining and the willingness to look — and in a partnership, the willingness is often the harder half.
Four numbers your firm has and cannot see
Client-level profitability has never actually been run
You know realization by partner, by department, and for the firm. You do not know that eleven percent of clients consume a disproportionate share of senior time at fees that have not moved since before the merger, and that a handful of them are outright losses once you account for who is actually doing the work.
The partner scorecard is a spreadsheet somebody rebuilds monthly
Utilization, realization, WIP, and AR by partner, assembled by the firm administrator from three exports, formatted, distributed, and discussed at a partner meeting two weeks after the month it describes. Half the meeting is spent agreeing on what the numbers mean because the definitions live in the spreadsheet rather than anywhere durable.
Every proposal gets written from scratch at nine at night
A senior manager assembles twenty pages for a manufacturing prospect, half-remembering paragraphs from a proposal that won last year and cannot be found. The firm has written a hundred of these. None of them is retrievable in a way that makes the hundred-and-first faster.
The website describes a full range of accounting services
Meanwhile the firm has a genuine contractor niche with fourteen years of depth, three partners who could speak to it authoritatively, and nothing online that would tell a prospect any of that. The expertise is real. The published evidence of it is a services page that could belong to any firm in the country.
Running the analysis nobody has had two spare days for
Client profitability is the first build because it is entirely derivable and consistently uncomfortable. Join time entries, billing, write-offs, and staffing mix per client, rank the bottom decile, and — this is the part that makes it usable — attach a reason to each one. Scope creep, stale fee, staffing drift, or a client who genuinely costs what they cost. A ranked list without reasons produces an argument; a ranked list with evidence produces a repricing plan.
The partner scorecard is pure assembly and gets treated as such: pulled on your cadence, definitions encoded once so they stop being relitigated, and the variance commentary drafted so the meeting starts from what changed rather than from what the numbers mean.
Proposals become retrieval instead of composition. Your own prior proposals, indexed by industry and service, become the source for the next one — assembled against this prospect's scope and reviewed by the manager whose name goes on it. And the niche expertise gets published: drafted from actual partner knowledge, in their voice, with a partner reviewing every substantive claim before anything about tax or audit treatment goes on your website.
What we build toward
- Bottom 10%Client profitability, ranked with reasons
- Not just which relationships lose money but why each one does — scope creep, stale fee, or staffing drift — which is the difference between a report and a repricing conversation you can actually have.
- Meeting-readyPartner scorecard narrated, not assembled
- Utilization, realization, WIP, and AR by partner on your cadence with variance commentary drafted and definitions encoded once, so the partner meeting argues about decisions rather than about denominators.
- Hours → 40 minProposal turnaround
- Assembled from the firm's own winning proposals against this prospect's scope, so a senior manager edits at nine at night instead of composing.
The profitability analysis is only as good as your time entry discipline, and every firm's is worse than it thinks. Expect the first run to surface entry problems alongside client problems — both are worth knowing and only one of them is fixable this quarter.
What actually gets built
Client profitability with attributed causes
Joins time, billing, write-offs, and staffing mix per client, ranks the bottom decile, and attaches a probable cause to each — scope creep against the engagement letter, a fee that has not moved, or work drifting to more expensive people than it should require.
Partner and department scorecards
Assembled on the firm's cadence with metric definitions encoded once, variance commentary drafted against the actual numbers, and the underlying query logic visible so a partner who disputes a figure can see how it was produced.
Proposal assembly from your own archive
Indexes the firm's prior proposals by industry and service line and drafts the next one against the current prospect's scope — your language and your differentiators, not a generic template with the firm name inserted.
Niche content, sourced from partners
Turns the expertise that already exists in three partners' heads into published pages that a prospect can find. Drafted from interviews and prior work product, and never published on a technical point without the partner who owns it signing off.
Referral and origination reporting
Traces new business back to its actual origin — an existing client, a banker, an attorney, a niche page — so origination credit and marketing spend are argued from evidence rather than from recollection at compensation time.
Modeled, not claimed
What repricing half the bottom decile does
A firm with several hundred business clients, a bottom decile that has never been examined, and fees on a meaningful share of it that have not been revisited in years.
The usual finding is not that a handful of clients are catastrophically unprofitable — it is that a long tail is mildly so, consistently, because fees drifted while scope grew. That tail is fixable through pricing rather than through firing anybody, and most of the affected clients accept a corrected fee because the work genuinely expanded and they know it.
Reprice half of a bottom decile toward break-even and the margin improvement arrives without a single new client, a single additional hour, or a single uncomfortable termination. The clients who decline the new fee were the ones the analysis was telling you about anyway.
- Bottom 10%
- where the analysis starts
- Half
- the realistic share that reprices without a fight
- 0
- new clients required for the margin to move
A pattern, not a promise. Whether your bottom decile looks like this is an empirical question we can answer from your time and billing data in a couple of weeks.
What partner groups actually worry about
This data will start fights. Partner compensation is political here.
It will, and pretending otherwise would be dishonest. Client profitability and origination data land directly on the two most sensitive topics in a partnership. The firms that handle it well do the same two things: they run the analysis firm-wide before anyone sees it by partner, so the conversation starts with the pattern rather than with a person, and they agree the metric definitions in advance rather than after somebody dislikes their number. We will build to whatever sequencing your managing partner wants, including keeping partner-level views restricted at first. What we will not do is quietly soften the numbers.
I am not putting AI-written tax content on our website with our name on it.
You should not, and we would not propose it. Technical content carries the firm's professional reputation and a wrong sentence about a treatment is a real problem, not a typo. The workflow is that a partner's actual knowledge is the source — interviews, prior memos, work product — the drafting speeds up the writing, and the partner who owns the subject reviews every substantive claim before publication. The firms that get burned here are the ones publishing generated content nobody read.
Our data spans three systems since the merger and none of them agree.
That is the normal starting condition and it is also why the analysis has never been run. We reconcile rather than requiring you to consolidate first, and the reconciliation itself usually produces a short list of data problems worth fixing permanently. Expect the first pass to take longer than you would like and to be the one that pays for itself.
Straight answers
Which time and billing systems does the profitability analysis work against?
CCH Axcess Practice, Practice CS, and Karbon are all workable, as are the practice management modules most mid-sized firms run. Where a firm is on something older or partially on spreadsheets, we work from exports — the analysis is identical, the refresh is slower.
Can we see profitability by industry niche rather than only by client?
Yes, and it is often the more strategic view. Niche-level profitability tends to reveal that the practice area everyone is proud of is subsidizing itself with senior time, or conversely that an unglamorous compliance niche is carrying the firm. Both change how the growth plan gets written.
Does the proposal drafting have access to client confidential information?
It reads your prior proposals, which are your work product rather than client return information. Where a past proposal contains identifying client detail, that detail is excluded from the index — proposals should be citing your experience in an industry, not naming who you did it for without permission anyway.
How long before the first profitability run produces something usable?
Two to three weeks for a first pass, most of which is reconciliation rather than analysis. The output is deliberately delivered as a working draft to the managing partner first, because the right sequencing of who sees what is a partnership decision and not ours.
Also for accounting & cpa firms operators
Your best leads of the year arrive in the worst month of the year
Operations & AdminYou deliver in February, invoice in May, and collect in July
Different industry? See every industry and function we have mapped.
Find out what your bottom decile looks like
The Pulse AI Operations Index scores intake, operations, and reporting across a firm and ranks what to fix first. For most multi-office practices, client-level profitability is the analysis with the largest gap between what the data could say and what anyone has looked at.