Assess before you buy

Buying is a relief. It feels like motion. Assessment feels like delay. For a small shop, the assessment is the only part that keeps the purchase from becoming shelfware.

In short. List what you already pay for, who uses it, and which workflow is actually broken. Then decide whether a new seat fixes that workflow.

What you already pay for

Write the tools that touch the work: email, texting, the job board or CRM, accounting, files, the group chat, the camera in the truck, the spreadsheet that is secretly the system of record. Next to each, write what it is the source of truth for, and who opened it this week.

You are looking for overlap and for ghosts. Overlap is three places that claim to know the customer’s phone number. A ghost is a subscription nobody opened. Ghosts are not a moral failure. They are evidence that “we have a tool for that” did not create a habit.

Do not start the list from a vendor’s category diagram. Start from yesterday.

Who owns the work

For the three workflows that eat the week, write the name of the person who finishes them. If the name is the owner for all three, your constraint is hours, not software. A new assistant that still needs the owner to review every output will add a tool to the same bottleneck.

Also write who is allowed to change the system of record. That person is not automatically the AI owner, but if the pilot wants to write into that system, they have to be in the conversation before you buy write-access.

Risk you will not outsource to a setting

Note the data you will not paste into a general chat: payment details, credentials, employee records you would not forward, anything involving a child, anything a contract says stays in a named system. A purchase that assumes “we’ll figure out the policy in the admin panel” is a purchase ahead of the rule.

One paragraph is enough. The 30-day plan puts that paragraph in week one, before week two’s pick.

Decide what the purchase is for

A new seat is justified when a named workflow cannot be piloted on tools you already have, and when someone has the hours to run the shadow. It is not justified by a demo that felt like the future.

If the workflow can be piloted in a tool you already pay for, do that first. The first-workflow guide is the scoring pass. The 90-day strategy is where the purchase sits in a longer sequence.

When the assessment says the gap is a team, not a login, book discovery and bring the tool list. The useful call starts from what you already run.