It is one job.

Not three services. I have done this inside very different companies and it has been the same work every time. Find where revenue actually stops, clean up what is in the way, then run it. The industry changes. The sequence does not.

Same four moves, every time

The order matters more than any single step. Most of what goes wrong is doing the fourth one first.

Find where it stops

Read what actually happened, not what the plan said. There is almost always one place things die, and it is almost never the place people assume.

Clean up what is in the way

The records, the duplicate accounts, the pricing nobody can explain, the deal from April that is still marked open. This part is boring and it is where the value is.

Run it

Follow up that goes out on a schedule. Things that stalled get chased. You start seeing movement you did not personally have to cause.

Show it honestly

A weekly read in plain numbers, including the ones that went the wrong way. A report that only ever shows wins gets ignored inside two months.

What changes, and what never does

Changes every time

  • ~The industry, and the vocabulary that comes with it
  • ~Which tools are already in the building
  • ~Whether the problem is demand or the machinery behind it
  • ~How much has to be rebuilt before anything can run
  • ~What the words say, because that is the client's voice and not mine

Never changes

  • +Look first. Nothing gets built in week one.
  • +The cleanup happens before the clever part
  • +One place where the truth about revenue lives
  • +Something runs at the end, not a document about running
  • +The number that went down is on the page

If a description of the work only makes sense for one industry, it is a case study wearing a service page. This is the part that travels.

What I take on, and what stays yours

Worth agreeing on in the first conversation rather than finding out in month two.

Mine

  • +Wherever the truth about revenue is supposed to live
  • +Follow up, written and actually sent
  • +The cleanup, all of it
  • +Content and email, built and then run
  • +The weekly read on what moved

Yours

  • xPricing, and what a piece of work is worth
  • xWho you will and will not take on
  • xDelivering the work itself
  • xAnything a lawyer should decide
  • xSaying no to a bad fit

One problem, six vantage points

This is the part a timeline cannot show, and it is the part you are actually hiring. It is also why the work is not limited to whatever engagement happens to be running this quarter.

2024 janitor 2020 founder 2018 vendor 2012 seller 2006 builder 1997 user
2024Now
The janitor
That most of it is cleaning up.

Pricing nobody can explain, a pipeline that is a guess, follow up that depends on one person remembering. Somebody has to sort that out before anything clever is worth doing. It is not a glamorous title and it is the honest one.

2020Founded and exited
The founder
What it costs.

Making payroll for five people out of revenue you personally brought in changes how you price, how you scope, and how fast you tell a client something is not worth doing.

2018Six plus years at HubSpot
The vendor
What finished looks like.

Six plus years inside a product at that scale is the only way I know to calibrate what good actually is, and it is the one thing you cannot get from working on small systems for a long time.

2012Account exec to sales director
The seller
What a buyer will and will not sit through.

The largest deal in the company's history, won on wireframes drawn in house with no deal team behind it, which is what you can do when you know the product well enough to invent the half that is missing.

2006Employee four
The builder
Why a thing does what it does.

Helping architect a CRM out of nothing teaches you that a field exists because somebody decided it should, usually for a reason nobody wrote down. That is still true of every system I inherit.

1997At thirteen, doing data entry
The user
What it is to be the person who has to type it in.

Every system I have built since answers to that, and most software is designed by people who have never once been that person.

Pick a seat. The janitor is driving, which is the seat this practice is in now. There is a trailer.

Keep these apart, because most people mash them together

One is something measured for the client. The other is how the work gets done. They are both real and they are not the same claim.

Whether the assistants name you

  • +A buyer asks ChatGPT, Claude, Perplexity or Gemini who to use, and a name comes back.
  • +Whether it is yours is measurable, weekly, against the questions your buyers actually ask.
  • +Mostly it is fixable, and the fixes are ordinary: what you publish, where, and how clearly you say who you are for.
  • +There is a one time audit for finding out, and a weekly run for watching it move.

How the work itself gets built

  • +Built with AI tooling, to a production standard. Said plainly rather than discovered later.
  • +It is the reason a small practice can carry work that used to need a room of people.
  • +It is leverage on the boring half, the cleanup and the plumbing, which is where the hours actually went.
  • +It does not replace the judgment, which is the part being bought.

Measuring whether AI recommends you is a service. Using AI to deliver is an operating decision. Selling the second as though it were the first is how this gets a bad name.

Start with whatever is stuck

Thirty minutes, no deck. If the honest answer is that you need someone in house, I will say that.