When a Number Becomes a Story

“Sales doubled” is a magnificent sentence if you sell things. It is somewhat less magnificent when sales rose from one to two and your warehouse contains eight thousand embroidered oven mitts.

The number is correct. The impression is incomplete. This gap is where many stories about business, technology, and public life acquire their dramatic lighting.

Numbers are useful because they compress information. They are dangerous for exactly the same reason. Compression leaves things out, and the omitted details may be the ones that determine what the result means.

Ask for the denominator

A rate describes a relationship. Without its base, it can be difficult to judge scale. Ten complaints out of twenty customers tell a different story from ten complaints out of twenty thousand.

The same principle applies to growth. A small operation can post spectacular percentage gains while adding little in absolute terms. A large operation may produce a modest percentage that represents substantial change.

Ask for both views. What is the percentage? What is the actual count or amount? Neither is automatically superior. Together they make exaggeration work harder.

Check the window

A chart needs a beginning and an end. Those choices can change its emotional temperature. Start after a collapse and the recovery looks heroic. Start before the collapse and the same endpoint may look disappointing.

This does not make every selected time frame dishonest. Analysis requires boundaries. It does mean that the boundaries deserve attention.

Try looking at a shorter and a longer period. Ask whether the comparison is seasonal, whether an unusual event affected the starting point, and whether the measure itself changed. Sometimes the dramatic trend survives these questions. When it does, your confidence has earned something.

Separate the measurement from the explanation

An observed change does not arrive with its cause stapled to it. A company changes its website and sales increase. Perhaps the redesign helped. Perhaps a promotion began, a competitor closed, or the season changed.

In a hypothetical shop, revenue might rise while profit falls because the promotion increased discounts and delivery costs. The headline “record revenue” can be accurate while answering a different question from “is the business healthier?”

Good analysis keeps these layers distinct: what was measured, what might explain it, and what decision follows. A persuasive story often slides across all three without pausing.

Make a small evidence card

For the next impressive statistic you encounter, write five items: the measure, the denominator, the period, the comparison, and the source. Then add one sentence explaining what the statistic does not establish.

You do not need to do this for every number in your life. Nobody should require a research assistant to choose a sandwich. Use it when the statistic is being asked to carry a consequential conclusion.

The aim is not to become the person who ruins every conversation by demanding methodology. It is to recognize when an elegant number is doing several jobs at once: reporting a fact, suggesting a cause, and selling you a feeling. You may accept the fact while declining the rest of the package.