Ten minutes before the call, your rep has five tabs open. One holds the CRM record. One holds an email thread from March. One holds a spreadsheet nobody owns. The customer history lives in all three places and in none of them. The rep walks into the call half briefed, again. This gap is the most sellable small service in AI right now. Not a platform. Not a transformation. A checked one-page brief before each call, sold as a two-week pilot.
The short answer: the service turns one rep’s approved notes and emails into a one-page call brief. After the call, it drafts the follow-up. The rep checks both before anything reaches the customer. You sell the pilot, the setup and the teaching. You do not sell results. For this exact build, no measured results exist yet, and the honest offer says so.
The offer is a pilot, not a platform
The pilot runs 2 weeks. It covers 1 rep and up to 10 calls. The brief holds four things every time. Customer history. Open promises. Useful questions. Missing information. The missing part matters most. A brief that admits a gap beats a brief that fills it with a guess.
Version one reads approved files only. It does not connect to the CRM. It sends nothing on its own. Every brief and every follow-up passes a human check first. That check is the product, not a step before it. Review AI-generated work with one signed line before your name goes on it sets the same rule for any AI draft in an office. A named person signs the work, or the work does not leave.
The build starts from a free sales plugin for Claude. The plugin costs nothing. Running it inside Claude takes a paid plan. The client pays for that plan directly, the same way the client pays for any software on the bill.
Win the pilot in one meeting
Ask the sales manager to walk you through the last customer call. Ask where the rep found the notes. Ask what was missing when the customer asked a hard question. Ask how long prep took, and how long the follow-up took. Ask what a useful brief would hold. Five questions tell you if the problem is real.
Then ask to see three things. One good brief. One poor set of notes. One follow-up that worked. Build your sample from fictional or redacted notes. The client approves real customer material before it touches the test. In the sample, every fact traces back to a note. Suggestions carry a label. Unknowns stay unknown.
This prep work repeats in every office until somebody writes it down. You already paid for your AI tools. In hours, rewriting the same prompts makes that point about prompts. A brief standard is the same fix on the sales floor. Write the shape once. Reuse it on every call.
Price the shape, not a promise
Charge one fixed fee for the pilot. The fee covers five things. A kickoff. The setup. Up to ten reviewed test cases. A results meeting. A handoff the team can run without you. Custom software sits outside the fee. A CRM migration sits outside it. Automatic sending and endless revisions sit outside it too.
Track your hours from the first day. The hours tell you what the pilot really cost to deliver. They set the price of the second pilot. If the pilot helps, price the follow-on work apart. Training more reps is a new price. A second call type is a new price. Do not fold them into the first fee to look generous.
A small shop can staff this service now. AI is handing the second founder the first founder’s bench shows how cheap the bench has become. One operator with the right seats can sell a service that used to need a team.
Measure three numbers from the first day. Prep minutes per call. Follow-up minutes per call. Briefs the rep had to correct before use. Those three numbers run the results meeting. Without them, the pilot ends in opinions.
Say the quiet part in the pitch
The method behind this offer is honest about its own limits. Its worked example uses invented companies. Its sample brief was written by hand. Its plugin test had not run yet when the method was published. None of that weakens the offer. It defines the offer.
You are selling a measured test with a decision date. At the end of two weeks, you compare two things against the current process. Time. Quality. The manager decides with numbers from the manager’s own calls. Steal that honesty for your pitch. Cold email replies fell to about 3%. A box on the desk still gets opened shows how hard a reply is to earn now. A checked follow-up, sent on time, is the cheap edge still left. A seller who names the limit early keeps the client who stays.
Do this week. Pick one rep and one real call on Thursday. Build the brief by hand from that rep’s own notes. Time the prep and time the follow-up. Then decide if the pilot earns its two weeks.
