Every week we turn away two or three potential clients. Not because they couldn't pay. Not because we didn't like them. Because their business wasn't ready yet — and we'd rather lose the work than charge $1,000 for a Strategy Day that wouldn't stick. Here are the five questions we use to figure it out. Run them on yourself before you spend money with anyone.
Workflows
Do you have documented workflows?
AI works best when workflows are clear and repeatable. If your team does the same task the same way every time, that's a candidate for automation. If your workflows are ad-hoc, inconsistent, or change every Monday, AI won't help yet. You need to standardize first.
Document how you actually do things. Get the team to follow the same process. Then AI can improve it. Ready? Yes if your core workflows are documented and consistent. No if everything is different every time.
Data
Is your data consistent?
AI learns from data. If your data is messy, inconsistent, or incomplete, the AI produces messy results. Example: half your customer records have phone numbers and half don't. Dates entered as “05/16/26” in one row and “May 16, 2026” in the next. Customer names sometimes uppercase, sometimes not. The AI will struggle.
You don't need perfect data. You need consistent data. Clean it once, then AI can work with it. Ready?Yes if your data is reasonably consistent. No if it's a mess and nobody's standardized it yet.
Measurement
Can you measure the result?
If you can't measure whether AI is working, you can't know if it's worth the investment. “We want to automate customer follow-up” is a good start. But what does “success” look like? More responses? Faster responses? More sales from those responses?
You need specific metrics. Processing time before and after. Accuracy before and after. Cost before and after. Revenue before and after. Vague success means you'll never know if it worked. Ready?Yes if you can define success in measurable terms. No if “success” is hand-wavy.
If you can't define what 'success' looks like in numbers, you can't tell whether AI worked — and you can't tell whether to keep paying for it.
Scale
Does the problem cost real money?
AI is only worth doing if the problem is significant. A task that costs you $200 a year isn't worth automating if it costs $5,000 to build. Ask yourself: how much time or money would we save if this was fully automated?
If the answer is “maybe a few hundred dollars a year,” it's not worth doing. If the answer is “5–10 hours per week” or “tens of thousands a year,” it's worth exploring. Ready?Yes if the problem is big enough to matter. No if it's a nice-to-have.
People
Is your team on board?
The biggest failure in AI implementation isn't technical. It's adoption. If your team thinks AI is going to eliminate their jobs, they won't use it. If they don't understand it, they won't trust it. You need buy-in.
Not everyone has to be excited. But the people who'll use the system need to believe it makes their life better, not worse. Ready?Yes if your team is at least open. No if there's strong resistance and no plan to address it.
The Off-Ramp
What to do if you're not ready.
Most of the businesses we turn away are perfectly fixable. Here's the order of operations:
- Document your workflows. Get the team to agree on how things should be done. Write it down. This is the first step before any technology.
- Clean your data. Spend time standardizing customer records, dates, naming conventions. Good data makes everything easier, AI or not.
- Find the pain. What's actually costing you time or money? Don't automate nice-to-haves. Find the bottleneck.
- Build the culture. Help your team understand AI is a tool, not a replacement. It makes their work easier, not threatening.
The businesses we end up doing the best work with usually look a lot like these:
Ready
3+ trucks, clear scheduling process
Workflows live in a scheduling system, even if it's clunky. Data lives in QuickBooks. Owner knows exactly which 4 hours a week are bleeding.
Ready
15–60 person firm with intake SOPs
Already standardized client intake. Already using a practice management system. The bottleneck is review and draft work, not the unknown.
Ready
Clinic with a written admin playbook
EMR is established. Reception roles are defined. The team can already name the three workflows that eat their week.
FAQ
What people ask before they decide.
How long until we're ready? Close businesses can clean up in 1–2 months. If you need to rebuild workflows and standardize data, 3–6 months is realistic. The time is worth it.
Can we do a Strategy Day if we're not quite ready? Sometimes yes — a Strategy Day can help you sort what to fix first. But we'll be honest in the 30-minute session before you book. Sometimes we recommend getting fundamentals right before paying us anything.
Do we need to hire someone to help us get ready? Usually no. It's mostly discipline: documenting what you do, cleaning data, identifying the real bottlenecks. A consultant can speed it up. They're not necessary.
What if my team is openly hostile to AI? Then start there. Listen to why. Most resistance is about job security or a previous bad experience. Address both directly before you spend any money on tools.
If you're not sure where you land — the 30-minute strategy session at the bottom is the right next step. We'll work through these questions together and tell you honestly which side of the line you're on.
