Three firms: a property management company, a recruiting agency, and a tax and advisory practice. One added about $340K in a year. One added about $210K. The third returned $99,840 on $81K of spend.
Their builds had almost nothing in common. Transaction categorization in QuickBooks. Maintenance coordination. Resume screening. Not one of those three businesses could have used either of the others' systems.
What they shared was the shape of the problem underneath: a small number of expensive people spending most of their week on work that required none of their expertise. Your industry decides what that work is. It does not decide whether it is there.
Here is what each of them actually installed — and then how to find your own version of it, which takes about an hour and no help from us.
What actually got built
Property management firm — 400 units, $2.8M/year. Four systems: inbound lead capture, tenant communications, maintenance coordination, and a private internal assistant running on their own data. $99,840 of first-year value on $81K of spend — a 123% return, and I am quoting the cost on purpose.
Recruiting agency — 6 recruiters, ~$1.8M/year. Three systems: resume screening, pipeline automation, inbound response. Admin time 13 → 3 hours a week, time-to-fill 40 → 26 days, roughly 3 extra placements per recruiter per year. About $340K in added revenue in year one.
Tax and advisory firm — 2 partners, 6 staff. Three systems, installed one at a time: transaction categorization across QuickBooks and Xero, client-communication drafting, and lead response. Senior repetitive work went from 50% of the week to 18% — about 11 hours back per senior — and that recovered time turned into roughly $210K of added advisory revenue. Same team, no new hires.
Why it transfers when the builds don't
Read those three lists again. Transaction categorization is meaningless to a property manager. Maintenance coordination is meaningless to a CPA. Resume screening is meaningless to both.
The work is not portable. The constraint is.
In all three, the business was capped by the same thing: people whose judgment is the actual product could only sell, advise, or place for a fraction of their week, because the rest of it went to work that did not need judgment at all. A partner categorizing transactions is a partner not doing advisory. A recruiter screening resumes is a recruiter not placing. That ceiling does not care what you sell.
And notice the counts — four systems, three, three. Not one of these was a single build. That is the part most owners get wrong: they go looking for the one clever thing, when the actual pattern is three or four unglamorous ones, sequenced, each one paying for the next. The tax firm installed theirs one at a time on purpose. The second only made sense once they could see what the first had freed up.
How to find yours
You do not need us for this. You need a list and an honest hour.
Write down every task in the business that repeats weekly. Not projects — the recurring work.
Next to each, put who does it, and roughly what that person costs you per hour.
Then ask what the task actually requires: could a competent new hire do this from a written doc?
Now your list has sorted itself. Every expensive row where the answer to (3) is yes is the opportunity, already ranked by what it is costing you.
Where the answer is no, AI cannot do it either — you have found a process problem wearing an AI costume, and automating it would only make it fast and wrong. And if your senior people are genuinely spending their week on senior work, this is not your constraint. Go look somewhere else, and do not let anybody sell you this.
If you want the ranking done properly
That list is the honest, unaided version of what we do. The faster version: we put the sequencing we have used across 60+ businesses into a free plan. About a minute — what you make a year, how big your team is, your one goal — and it hands back three systems in the order to build them, with an honest range on the hours and dollars each returns.
It will not price your business. Nobody can do that off a form. But it will tell you what to build first, which is the decision that determines whether any of this pays you back.
Already filled it out? Then you have the order, and the next question is what the top system on it is worth in your operation — real numbers off your actual volume, not a range. That is a call. We go through where your hours are going, rank every workflow by dollar impact, and give you the build order in writing. You keep it either way, whether or not we ever work together.
— Nick
P.S. The private internal assistant was the fourth thing the property firm built, not the first. It is the one that demos well, so it is the one most people want to start with — and starting there would have put a clever interface on top of processes nobody had fixed yet. It would have been the thing they pointed at a year later when they said AI did not work for them.