How I work

How I work, from the first call to handoff

I work on the frontier of AI implementation, and I'm making sure local businesses don't miss the train. That work is backed by a decade of doing this by hand: 8 years as a director in digital marketing for a national agency and 2 as a freelance web designer. That experience is how I know what a right answer looks like, and how I check every result before it reaches you.

The way I work: find the one step that is holding the business back, measure it, build only what moves that step, test it until it is dependable, hand it over so it runs without me, and then look again, because the slow step moves.

I work this way because of what I saw at the agency. I talked many customers out of big advertising plans because they didn't have the systems to support them. Otherwise it meant scope creep on a service the agency didn't hang its hat on, or losing the client 3 months later because more traffic didn't produce more business.

The idea

Every business has one slow step

A business is a chain: people find you, some of them get in touch, some of those become customers, some come back. Behind that sits the work of delivering what you sold. The whole chain moves only as fast as its slowest link. This is the old idea from manufacturing called the theory of constraints, and it holds just as well for a roofing company or a dental office.

It matters because effort spent anywhere else is wasted. If leads sit in an inbox for two days, a better ad campaign brings more leads to the same inbox. If the schedule is already full, more leads are a cost, not a win. The first job is to find the link that is actually slow.

Step 1 · The call

The first conversation is about numbers

The call is free and it lasts thirty minutes. I walk through four stages one at a time and ask for the numbers at each:

  • Reach: how many people see the business in a typical week, and where?
  • Acquisition: how many of them get in touch, and how fast does someone reply?
  • Conversion: how many of those become customers?
  • Expansion: how many come back, buy more, or send someone?

The stage that looks weakest next to the others is usually the bottleneck, and relieving it is what lets you close more business.

I'm not limited to those four. I also ask how much the business can handle in a typical month, how much comes in, and what is backed up, late, or turned away, because more leads won't help a schedule that's already full. But these four stages are where the biggest wins usually are.

The call works with rough numbers and points to the stage that looks weakest. Going through the actual accounts and systems behind those four stages, and writing a plan in priority order, is the marketing and sales systems audit.

"I don't know" is a useful answer. A step nobody counts is usually the step nobody is watching. If some numbers do not exist yet, counting them comes first, in writing, before anything gets built.

Step 2 · The scope

Then a baseline and a written scope

Before anything is built, I write down where things stand and what we would change:

  • the step that is holding the business back, measured
  • what the work around it costs today in time and money, with where each number came from
  • the concrete actions that would widen the slow step, broken down until each is something a person or a system can actually do
  • what a system may do alone, what it has to ask about, and what it never touches
  • the order of work, with the first working piece early

It also says what not to build. Sometimes the honest answer is that a process needs fixing before any system is worth paying for. And before anyone pays to widen the slow step, the first question is how much more it can give with the people and tools you already have.

If you want this done across your whole team, with AI in mind, that is the AI implementation roadmap. Most work does not need to start there.

Step 3 · The build

Builds start small and get tested before they run alone

The first build is usually a process you already have and already measure, made faster or more reliable, not a brand new invention. That way there is a number to beat.

AI systems do not give the same output every time. A system that did well once might do badly the next three times. So before anything runs on its own, it gets tested on about ten real examples from your business, including the messy and unusual ones, each scored by a person on your side. The average has to clear a bar, and so does the worst result, because a system that swings between great and bad is not ready. Whenever the model, the instructions, or a connected tool changes, the same test runs again, so you know whether the change helped.

While it runs, anything that spends money, sends something to a customer, or changes a record asks first. Everything it does is written to a log you can read.

Step 4 · Handoff

You own it, and your team learns it

Everything is built on your accounts and your data, and documented so your team can run it. Access I needed for the build is reduced at handoff, and you keep the list of who has what.

Systems only stick if people use them. During handoff we hold a short weekly session with the people who use the system: whoever got the best result that week shows how, and that becomes the written way of doing it. After a few weeks your own lead runs that session. The goal is that you stop needing me for it.

If your own team would rather do the building, the path is the same, except they build and I direct and review.

Step 5 · Look again

Then the slow step moves

Fix the slowest link and another one becomes the slowest. That is a sign the first fix worked. So the last step is to measure again: the same questions as the first call, against the baseline. Sometimes that points to the next thing worth building. Sometimes it says the business is in good shape and nothing more is needed yet.

Ground rules

What I will not do

No guaranteed results. No invented numbers. Nothing built before it is measured, nothing that spends or sends without your say, and nothing that locks you in to me.

Next

Where to start

Book the thirty-minute call. If you would rather pick a place to start first, the four services are laid out side by side, with their starting prices.

The services, one by one →

Want to talk about your business?

Book a free 30-minute call. Bring whatever numbers you have.

Book a call
How I work, from the first call to handoff · Jeremiah Fehrle, Business Systems Specialist in Tulsa