The AI & Automation Process Assessment.

The short version

The AI & Automation Process Assessment is three weeks spent inside how your firm actually works. At the end of it you get a written plan built in four parts:

1. A full picture of where your staff’s time goes.
2. An honest analysis of which problems technology can actually fix today.
3. A roadmap ranking the fixes by what they’ll return for what they’ll cost.
4. An AI Readiness Score — a single number out of 100 that tells you how prepared your firm is to compete in a market where your competitors are already automating.

That’s the whole product. The rest of this page explains why it exists, what happens in each of the three weeks, how I get to the numbers, and what you can and can’t expect from it — with specifics for law firms and accounting firms, because that’s who this assessment was built for.

01 · RationaleWhy this exists.

The economics of a firm are simple and slightly cruel: you’re paid for judgment, and judgment is the smallest part of the day. The rest goes to intake, document assembly, chasing clients for records, reconstructing time entries, and writing the same status update for the eighth time. Every hour of that is either written down at billing, absorbed by salaried staff, or done by a partner at 9pm.

Most firms know something is eating that time. What they don’t know is which thing is eating the most, what it actually costs in realization and write-offs, and which problems are worth fixing with software versus which ones just need a better process — or should be left alone entirely.

That gap matters because the most common way firms waste money on automation is by building the wrong thing first. Someone gets excited about a tool, a project gets funded, and six months later it’s automating a task that was never the expensive one. Meanwhile the client document requests still stop after the second email, the engagement letter still gets assembled by hand from last year’s, and time entries still get reconstructed from memory on Sunday night — because nobody ever measured where the hours were really going.

The assessment exists to answer that with evidence instead of guesses — and to give you one more thing measurement alone can’t: a clear reading of whether your firm is actually ready to use what it learns.

02 · GroundworkWhat happens before we start.

Before week one, I ask for a short list of materials. Nothing exotic: an org chart, a list of the software you use and what it costs — practice management, document management, time and billing, tax workflow, whatever holds the work — and headcount by role, from partners through admin. I also ask leadership to name a point of contact with the authority to unblock things — because at some point mid-assessment, I’ll need a report pulled or a door opened, and the whole schedule shouldn’t stall waiting on it.

For accounting firms, we also schedule around your calendar honestly: I don’t run discovery in the middle of a filing deadline, and an assessment timed to land before busy season — so the first fix is live when the volume hits — is worth more than one that lands during it.

If any of that is hard to produce, that’s useful information by itself, and we talk about it before you pay me anything.

03 · Week onePhase one: Discovery.

Discovery starts wide and then goes deep.

It starts with a firm-wide survey. Every member of your staff gets a short, anonymous questionnaire: where their time goes in a typical week, what they do repeatedly, what they work around, and what they’d fix first if anyone asked them. It takes each person about ten minutes. The survey does two things no interview schedule can — it hears from everyone instead of a chosen few, and it surfaces the problems people will write down but won’t say in a meeting. In firms, that especially means the things associates and staff accountants won’t say in front of a partner.

Then come the interviews. I sit down with the partners to understand where the firm is trying to go — growth, succession, merger, staying exactly the size you are — because a plan that ignores your actual goals is just a list. I talk with the people who lead each function or practice area about where their team’s time goes. And then — this is the part most assessments skip — I talk with the people who do the work: the paralegals, associates, staff accountants, bookkeepers, and admins. Partners describe how a process is supposed to run. The people running it describe how it actually runs, including the workarounds, the retyping, and the steps that exist only because a system broke years ago and nobody ever fixed it. The distance between those two descriptions is usually where the money is.

Finally, I pull the numbers your systems already hold. The software firms run — practice management, document management, time and billing, tax workflow, the shared inbox — can export its history: how many matters or engagements opened last year, how long each phase actually takes, where work sits waiting, how much recorded time gets written down at billing, how many client document requests take three or more follow-ups. Where a system can’t export, I use what your staff reported, and the final plan is honest about which numbers came from where.

04 · Week twoPhase two: Analysis.

Week two turns everything from discovery into findings.

First, I build the picture of where the hours go: every significant repetitive workflow across the firm, with hours per month and cost per year attached — at real loaded rates for salaried staff, and at billing rates for time that could have been billable, because those are two different kinds of expensive. In my experience the ranking surprises people. The loudest complaint is rarely the most expensive problem.

Then comes the part that separates this from a sales pitch: pairing each inefficiency with what technology can actually do about it — and what it can’t. For every workflow I answer three questions honestly. Can today’s tools realistically take this over, at a quality your clients — and your professional obligations — would accept? Is the underlying process sound enough to automate, or would automating it just make a broken process run faster? And is your data in good enough shape to support it — because inconsistent matter naming, duplicate client records, and documents scattered across a DMS, email attachments, and someone’s desktop are the single most common reason firm automation projects fail, and almost nobody checks before they build.

One rule shapes every recommendation in a firm: nothing client-facing goes out without a professional’s review. The value of automation in a practice is the first draft, not the final word — the engagement letter assembled and waiting when someone sits down, the document summary ready before the meeting. Any workflow that can’t support that review checkpoint is one I’ll tell you not to automate, in writing.

Some findings won’t involve AI at all. Every assessment turns up problems that need a process change, a setting adjusted in software you already own, or a task that should simply stop being done. Those go in the plan too. You’re paying me to find where the week goes, not to sell you automation.

05 · In practiceWhat this finds in a law firm.

Every firm is different, but the same patterns show up because the economics are the same. Intake that takes a week of back-and-forth — forms, conflict checks, engagement letters — before anyone does billable work. Partners doing work below their rate because delegating it feels slower than doing it. Client status updates written when someone finally gets to them, which is usually after the client calls. Document collection from clients that dies after the second request, stalling the matter and the billing behind it. And time entries reconstructed days later from memory and calendar scraps — which is where realization quietly leaks, because reconstructed time is undercounted time.

The industry data here is blunt: the average attorney captures fewer than three billable hours in an eight-hour day, and published estimates put time leakage anywhere from a few percent of revenue to considerably more. Whatever your firm’s real number is, the assessment measures it — that’s the point of measuring instead of estimating.

06 · In practiceWhat this finds in an accounting firm.

The accounting version of the same pattern: client document collection — the PBC list, the missing 1099, the bank statements chased four times — consuming staff hours and compressing everything downstream into deadline weeks. The same client questions answered individually all season when the answers live in prior-year files. The exceptions and cleanup your tax and audit software’s rollforward doesn’t touch, handled by hand engagement after engagement. Status inquiries — “where’s my return?” — interrupting the exact people trying to finish it. And write-downs at billing because the time to do the work outgrew the fee, year after year, without anyone re-examining why.

Most of this is document-and-deadline work with clear rules — which is precisely the shape of work today’s tools handle well, and why the fastest-payback items in a firm’s roadmap tend to come from it.

07 · Week threePhase three: The roadmap.

The roadmap is where analysis becomes decisions. It ranks every worthwhile improvement by one standard: what it returns for what it costs, in your numbers, on your timeline.

For each recommended item, the roadmap states what it would save or recover per year, what it would cost to build, how long until it pays for itself, and what has to be true for it to work. For firms, the return is stated in the terms partners actually decide with: realization, utilization, and capacity — not vague efficiency. The recommendations are sequenced deliberately — starting with fixes that pay back fast and prove the approach, then moving to the deeper changes that return more but take longer. For accounting firms, the sequence also respects the calendar: what must be live before busy season, and what should wait until after. If something popular shouldn’t be automated, the roadmap says so and explains why.

The top recommendations come with a fixed price and a timeline — not a vague range, but a number you could put in front of a partner meeting or an executive committee. And the roadmap always includes the short list of things to stop, fix, or leave alone: the non-software findings that are often the cheapest wins in the whole document.

08 · The instrumentThe AI Readiness Score.

Alongside the roadmap, you get a single number out of 100: your AI Readiness Score. The score measures your firm across six areas:

• The quality and accessibility of your data.
• Whether your processes are consistent enough to automate — including whether every partner runs their matters or engagements their own way, which is the most common firm-specific ceiling.
• Your technology’s ability to connect to modern tools.
• Your staff’s capacity to absorb new ways of working.
• Whether anyone actually has the authority to approve and govern this kind of change, or whether every decision needs unanimous partners.
• Your track record with change itself — because the firms that struggle with automation are almost never struggling with the technology.

The number does three jobs. It tells you where you stand against a market where your competitors are adopting these tools whether you do or not — and in both law and accounting, they are. It tells you why you stand there — the score comes with a breakdown showing which of the six areas is holding you back, so a 58 driven by scattered documents means something different from a 58 driven by five partners with five different intake processes. And it disciplines the roadmap: if your score says you’re not ready for a recommendation, the roadmap sequences the prerequisite work first instead of pretending the problem away. A plan that ignores your readiness is a plan that fails on schedule.

Re-measured a year later, it’s also the simplest honest answer to the question every partnership eventually asks: is any of this working?

09 · Standards of evidenceHow I get to the numbers.

A few rules govern every figure in the plan, because a plan full of inflated numbers is worse than no plan at all.

Every number is labeled by where it came from — measured from your systems, or estimated from staff reports cross-checked against each other. When I estimate, I say so, and I use ranges instead of false precision.

I count conservatively. If a workflow eats somewhere between 20 and 40 hours a month, the plan uses the low end for any decision that depends on it. If the case for fixing something only works with optimistic assumptions, it isn’t a good case, and I’ll tell you that.

And I separate two kinds of value that usually get blurred together. Hours saved are only worth money if the freed-up time gets used for something — billable work, capacity that delays a hire, a busy season that doesn’t require the same overtime. Revenue recovered — time captured that used to leak, a stalled matter unblocked, a fee that stops being written down — is actual dollars. The plan treats them separately so you can see exactly which is which.

10 · Privilege & confidentialityWhat happens with your data.

The terms are simple, and for firms they have to be stricter than for anyone else. I’ll sign your NDA and confidentiality terms before anything starts. The survey is anonymous, and nothing any individual staff member tells me is ever attributed to them — that promise is what makes the discovery phase honest.

The assessment does not require me to read privileged client material or client tax information, and it’s designed not to: I work from time data, workflow patterns, system exports, and document templates — not the contents of client files. Wherever possible, data stays in the accounts you already own; I work from exports and screens rather than copying databases. I don’t put client information into consumer AI tools, and anything I hold for analysis is deleted when the engagement ends.

For the roadmap itself, every recommendation that would touch client data states its data handling plainly — what would be processed, where, under what terms — and flags anything your managing partner, your bar’s confidentiality rules, or the rules governing tax return information should weigh in on before you act. I’m not your lawyer or your compliance opinion, and the plan is written so the person who is can review it quickly.

11 · FitWho this is for — and when I’ll tell you to wait.

The assessment is built for law firms and accounting firms of roughly 10 to 150 people — from a five-partner firm with support staff to a multi-office practice — where the same kinds of work happen dozens or hundreds of times a month and nobody has ever measured what the repetition costs. Where you land in the $4,500–$6,500 range depends on headcount, the number of practice areas or service lines I’m assessing, and how many systems hold your data; you’ll know your exact number before you commit.

There are situations where I’ll tell you to wait. If you’re mid-way through replacing your practice management or tax software, if a merger or partner transition is in motion, or if the honest first project is cleaning up a decade of inconsistent client records, the plan will say that plainly — and your readiness score will show you exactly why. An assessment that tells you not to spend money yet is still a useful assessment.

12 · AfterwardWhat happens after.

You have three options, and all three are fine.

• Build with me. Most builds from an assessment are fixed-fee starting at $7,500, live in two to three weeks, with an optional month-to-month retainer if you want me keeping the system running. Because the assessment already did the investigation, the build starts from a scoped plan instead of a discovery phase.
• Build without me. The roadmap has enough detail that a capable developer, yours or anyone’s, can work from it, and it’s written with that reader in mind.
• Sit on it. Budgets have cycles, partnerships vote when they vote, and the ranked list will still be the ranked list in six months.

13 · The next stepHow to start.

It starts with a twenty-minute call. You tell me where you think the hours go, I ask enough questions to confirm the assessment is worth doing for your firm and quote your exact fee, and I tell you honestly if it isn’t worth doing. If someone can already point at the one workflow that’s obviously the problem — the intake grind, the document chase, the Sunday-night time entries — you may not need the full assessment; a direct scoped proposal within 48 hours might serve you better, and I’ll say so on the call.

Either way, you’ll leave that call knowing what it costs, what you’d get, and when it would be done, before you’ve committed to anything. That’s the standard for everything I do, and the assessment is where it starts.