AI Coaching Before the Build
We recently added AI coaching, workshops, and consulting as a formal service. It fills a gap we kept running into.
A business owner reaches out. They've seen what AI can do, they know their shop has manual work that shouldn't be manual, and they want to do something about it. But they're not ready to commission a build. They're not sure which problem to solve first. They haven't mapped their workflows in a way that lets them hand us a clear spec. They just know something should change.
That's not a build engagement. That's a coaching engagement.
What It Looks Like
A coaching session is a focused working session: usually two to three hours, structured around their actual operation.
We start by mapping the manual work. Not the high-level summary ("we handle quotes and scheduling"), but the specific steps: who does it, how long it takes, what triggers it, what happens when it's done. Most businesses have never written this down. Doing it once surfaces things that have been invisible for years.
From that map, we identify where the bottlenecks are and which ones AI can actually address. Not every manual step is worth automating. Some are judgment calls that belong with a human. Others are pattern-matching work that a model handles well. We help draw that line.
By the end of the session, the business has a prioritized list of what to tackle first, rough estimates of the time and money currently being spent on each item, and a clear sense of what a build would actually look like. Some clients take that list and run with it internally. Others hand it back to us as a build spec.
The Problem With Skipping This Step
Custom builds that skip this step often go wrong in a specific way: they solve the visible problem instead of the underlying one.
A client asks for a quote follow-up system. They build it. Six months later they realize the real bottleneck was that their initial quote was taking four hours to produce, so fewer quotes got sent, so there was less to follow up on. The automation helped, but they built the downstream thing instead of the upstream thing.
A workshop that maps the full workflow before any build starts catches this. "You want the follow-up system, but look what's upstream from it" is a conversation that's a lot cheaper to have in a planning session than after the build is done.
The Workshop Format
For teams, we run it as a half-day workshop. We bring the whole process: a structured interview flow, a workflow map template, a scoring rubric for prioritization, and the output format that becomes the brief if a build follows.
The output is a written document: workflows mapped, estimated hours per week, estimated annual cost of the manual work, prioritized list of builds, rough effort and ROI estimate for each. It's the same format we use internally when we're auditing a prospect's site before a sales call. The only difference is that it's produced collaboratively with the team rather than by us from the outside.
Who This Is For
This isn't a fit for every client. If a business has already done this thinking and comes to us with a clear problem and clear data, we go straight to the build. That's the faster path when the groundwork is done.
Coaching is for businesses that haven't done the groundwork: they know they're leaving time on the table, they're not sure where, and they want to be deliberate about where to start. It's also for teams that want to start using AI tools themselves and need someone to help them identify what's actually worth their attention versus what's noise.
The build follows from the coaching. The coaching prevents the build from solving the wrong problem.