How to turn a business count into a chart people understand
A count of shops or clinics means little as a bare number. Which chart to draw, what to put on each axis, and how to keep the base of every figure honest.
Part of the guide Location intelligence automation
To turn a business count into a useful chart, decide first what comparison the reader should make, then pick the chart that makes only that comparison easy: bars to compare kinds of business or cities, columns to show change week by week, a map to show where, and a single big number when there is nothing to compare. Keep every bar on the same base (the same area, the same date, the same way of counting), start the axis at zero, and write under the chart where the figures come from.
This is for analysts, marketers and founders who have a list of businesses, from a market study, a prospect list or a competitor watch, and need to show it to a manager, a client or a board. It covers choosing the chart, a worked example with real counts from London, the errors that make charts lie, and how to have charts drawn and refreshed for you.
Start with the question, not the data
The same list can answer very different questions, and each wants a different picture:
| The question | The chart | What goes on the axis |
|---|---|---|
| Which kind of business is most common here? | Horizontal bars, longest first | Kinds of business, and their count |
| Which city has more of them? | Bars, one per city | Cities, and a count on the same base |
| Are more opening than before? | Columns, one per week or month | Time, and new places found |
| Where are they? | A map with points or a heat view | Position |
| How many are there? | One number, with its base written under it | Nothing |
A simple guide to matching the chart to the question.
If you cannot say the question in one sentence, the chart will not say it either. "Most supermarkets near central London belong to a chain, most gyms do not" is a question a chart can answer. "Here is our data" is not.
A worked example: chains near the centre of London
Take four kinds of business near the centre of London, from xMap's places database (November 2026). The totals within 10 km of the centre are large, but the split between chains and independents is worked out from the few hundred nearest the centre, so that is what the chart should show.
| Kind of business | Listed within 10 km | Nearest counted | Of those, in a chain |
|---|---|---|---|
| Supermarkets | 1,332 | 437 | 296 (68%) |
| Bakeries | 1,367 | 380 | 121 (32%) |
| Gyms | 1,112 | 353 | 88 (25%) |
| Pharmacies | 866 | 438 | 91 (21%) |
xMap's places database, November 2026. Chain figures are of the places nearest the centre of London, not of every place listed.
A bad chart would put "1,332 supermarkets" next to "296 in a chain" as two bars, inviting the reader to divide one by the other. That mixes two different bases: the total within 10 km and the nearest few hundred. The honest chart is a single row of horizontal bars, one per kind of business, each bar the share in a chain of the nearest counted: 68%, 32%, 25%, 21%, longest at the top. The title says what it shows ("Share of places in a chain, nearest to central London") and the caption says the base.
That one chart carries a clear message: in central London, supermarkets are mostly chain stores, while bakeries, gyms and pharmacies are mostly independent. A supplier selling to independents knows at once where the doors are.
The errors that make a count chart lie
- Different bases side by side. A 5 km count next to a 10 km count, or a count from last year next to one from this month, compares nothing. Put the base in the title or caption, and keep it the same for every bar.
- Cutting the axis. A bar chart whose axis starts at 200 makes 210 look twice 205. Bars start at zero; if the differences are small, say they are small.
- A pie with ten slices. Beyond three or four parts, nobody can compare slices. Use bars.
- Percentages with no count. "68% are chains" means more when the reader can see it is 296 of 437. Put the count in a label or the caption.
- Raw counts for areas of different size. A big district always has more businesses. To compare places, divide by area or by residents first, and say which.
- Colour that means nothing. Use one colour for the bars and a second only to pick out the one bar you want the reader to notice.
Getting charts drawn for you
In xMap Flows you rarely need to draw by hand. Three ways, from quickest to most automatic:
- A chart from a sentence. In Results, type a question such as "new places by week for the last 8 weeks" or "results by category this month". You get a chart counted from the results your AI employees found, and Pin keeps it on a dashboard where it is counted again each time.
- The Data analyst. For a question the ready charts cannot answer, such as the share of chains by city, the Data analyst block writes a small program over the items, runs it on all of them and returns the answer, a table and a chart, with numbers taken only from what the program worked out.
- A report on a timetable. Make a report turns the items that reach it into a PDF with your logo: the figures, charts by kind and by city, a map and the list, emailed or saved to Google Drive.
A typical AI employee for a monthly market chart: Schedule on the first Monday, Area of interest with the cities you track, Nearby places for the kinds of business, xMap LLM on the whole list to write a short summary, then Make a report. What lands in the inbox is a PDF with the counts, the charts and two paragraphs saying what changed, ready to forward.
Questions
What is the best chart for comparing business counts?
Horizontal bars, sorted longest first, with every bar on the same base. Use columns when the comparison is over time.
Should a bar chart always start at zero?
Yes. Bars show size by their length, so a cut axis exaggerates small differences. If you need to show small changes, a line or a table is more honest.
Can I share the chart with a client?
Yes. A dashboard can be shared by link, and Make a report produces a PDF with your organization's logo and colour.
Why do the chain figures not match the total count?
On xMap's market figures, chains and independents are worked out from the places nearest the centre, not from every place listed. Always chart them as a share of that nearer group.
Templates that do this
Read next
More on business and places data · Every article · Templates