Writing a local reason to call: what makes a cold call feel relevant

A local reason to call is one true fact about the business or the streets around it that connects to what you sell, said in the first sentence. "You opened in March" is a fact; "you opened in March, and the first winter is when new cafes find out their coffee machine was the wrong size" is a reason. The best reasons come from things that changed recently and nearby: an opening, a new competitor, roadworks, hiring, a new building. Generic praise is not a reason.

This is for anyone who calls or writes to local businesses they have not spoken to before. It covers what counts as a local reason, which facts make strong ones, how to write them so they do not sound like surveillance, a made-up example, and how an AI employee drafts one for each business on your list.

What a local reason is

A reason to call answers the question the other person is silently asking: why me, why now? A local reason answers it with something about their place. It has three parts:

• A fact you can see, such as an opening date, a rating, a job posting, a nearby closure or a road that has been closed for weeks.
• A link to a problem the business is likely to have because of that fact.
• A small ask that fits the problem: a question, a short visit, a sample.

Which local facts make strong reasons

Each of these is a fact the flow can find; the reason in the third column is yours to judge.

Two kinds of fact are weak. Facts that are true of every business ("you have great reviews") say nothing about this one. Facts about people ("I saw your manager left") feel intrusive. Keep to facts about the business and its street that the business would expect a supplier to notice.

How to write it so it lands

• Say the fact plainly and early. "I noticed a second bakery opened two doors down last month" beats a long introduction.
• Make the link a question, not a claim. "Has it changed your mornings?" is easier to answer than "You must be losing customers."
• Keep to one fact. Three facts in a row sound like a file on them.
• Never state a number you cannot back. If the flow says three cafes opened within 500 metres, say three; do not round it into "lots".

A worked example

A made-up case with invented details: a company in Paris sells shelving and display units to independent bakeries. Its AI employee finds new bakeries each week and counts other bakeries within 300 metres of each one.

Hello, I am calling about the bakery's opening on Rue des Martyrs in September. There are four other bakeries within a few minutes' walk, so I imagine the window and the counter are doing a lot of work for you. Would you have ten minutes this week to see two layouts other new bakeries in the 9th have used?

Every fact in that opening came from the flow: the opening month, the street, the four bakeries nearby. The link (the window has to work hard on a busy street) and the ask are the salesperson's. The example is invented; the shape is what to copy.

Mistakes to avoid

• Reading the data at them. "Your rating is 4.3 from 212 reviews" sounds like a report. Use the fact to ask a question.
• Old news. An opening from two years ago is no longer a reason. Prefer facts from the last few weeks or months.
• Guessing a fact. If the opening date is missing, say nothing about it rather than inventing one.
• The same reason for everyone. If every call opens with "I saw you are new", the reason has become a script.

How to do it with an AI employee

xMap LLM can draft one reason per business from the fields the earlier blocks found. It cites only what is in each item, so the facts you want it to use must be found first. The Morning call list with decision makers and call notes template does this; Watch new competitors supplies the competitor facts. In order:

1. Schedule and Area of interest.
2. Nearby places for the businesses you sell to, with Only new after it.
3. A second Nearby places set to Each incoming place, with counts, for the competitors within a few hundred metres of each one.
4. Job postings or Traffic incidents if those facts matter to what you sell.
5. xMap LLM with an instruction such as: "Write one opening sentence for a call, using one fact from this business's fields. Do not mention any fact that is not in the fields."
6. Email or Google Sheets for the reps.

What lands in the inbox is each business with its opening line beside its phone number. Read a dozen on a test run before anyone calls: if the lines repeat each other, give the AI more facts to choose from rather than a longer instruction.