How to compare three candidate sites side by side

To compare candidate sites fairly, measure each one the same way at the same moment: the same reach, the same distances, the same kind of competitor, then lay the figures out in one table with a column per site. In xMap Flows, Compare sites does exactly that for two or three spots: it measures each, marks the best value in every row, and writes a short verdict naming the strongest site, using only the figures in the table.

This article is for the moment a shortlist is down to a few addresses and someone has to choose. It walks through setting up a comparison, what each row means and which way is better, how to read the verdict, what to do when figures are missing, and how to keep watching the sites after the decision.

Setting it up

1. Open Results, then Map, and press Compare sites.
2. Click two or three spots on the map. Drag a pin to move it, or press the x on its chip to remove it. You can also add a result from its card with Compare.
3. Press Compare. The cost is shown first; nothing runs until you press Compare them.
4. Read the table, the map and the verdict. Download the PDF or share a link.

You can also ask in words. Ask xMap understands a request such as compare Rua Augusta, Avenida da Liberdade and Alfama for a bakery, places each address, shows the cost, and opens the same comparison.

What the rows mean

The rows are grouped by question. Each row says whether higher or lower is better, and the best value is marked only when the sites actually differ.

• People. Who lives within 10 minutes by car of each spot, and how many people there are for each place of your kind. More people, and more people per rival, is better. Population figures exist only in some countries; elsewhere the row is empty and says why.
• Competition. Places of your kind within 1 km: how many, how many are chains, their average rating and how many opened in the last year. Fewer rivals is better; a lower average rating means weaker rivals.
• Getting there. The road speed beside each spot right now, the nearest station, bus stops and car parks within 1 km, and charging stations within 2 km.
• Around it and risk. Where the data is switched on, nearby schools and supermarkets, and natural hazards, rivers, the sea and air quality. A row whose data is switched off is left out.

The kind of place counted as a rival is the one xMap remembers for your company, else restaurants. Set it before you compare, or a bakery will be measured against restaurants.

A worked example

A made-up example: a bakery chooses between three spots in Paris. The table might read like this:

Illustrative figures only, to show how a comparison reads.

Site 3 has the most people but also the most bakeries, well rated, with three new arrivals: a crowded, improving field. Site 2 has fewer people but almost twice as many per bakery, and its four rivals are rated lower. Counting marked rows alone would mislead; the verdict weighs them.

Reading the verdict

At the top, Best site names the winner and why in a sentence or two. It is written by xMap AI from the figures on the page and nothing else: a sentence that quotes a number the table does not show is thrown away. When AI is not available, the same kind of sentence is worked out from the figures by rules, giving more weight to people, people per rival, rivals and risk. When the sites are too close to call, it says so rather than inventing a winner.

When a figure is missing or estimated

• Sample figures appear where live data is not switched on for a row. They are free and said so on the page; do not decide on them.
• Estimated travel areas are circles at a typical speed, used when the road area cannot be drawn. The page says so; compare people figures between sites with care if one is estimated.
• Empty population rows mean the country has no population figures yet.

Mistakes to avoid

• Spots too close together. Two pins 200 m apart share most of their rivals and roads. The table will show near-identical columns.
• Comparing on one hour of traffic. The road speed is measured when you compare. Repeat at your busy hour if it matters.
• Forgetting the visit. The table cannot see the frontage, the rent or the lease.

After the decision

PDF downloads the comparison with your logo for the committee. Share a link makes a public page with no name or email on it, which you can stop under Shared links. Hire an employee to watch them starts an AI employee on a 1 km circle round each site: a Rival watcher, an Opening watcher or a Road watcher, so a competitor opening next to the chosen site reaches you the week it appears. For whole lists of candidates rather than three finalists, a flow with Travel area, Demographics, Nearby places and Traffic Flow measures the same rows for every address in a table.