Reading demand: the free data that shows what customers are doing

Trend analysis is usually described as understanding where the market is going, which is an ambition rather than a method. The version that produces decisions is narrower: find out what people are searching for, how that has changed, and whether the change is real or seasonal.

Four sources will do it, all free, and each one answers a question the others cannot.

1. Google Search Console: what you already win and nearly win

Start here rather than with market research, because this is data about your own customers rather than about people in general. The performance report gives you every query that showed your site, its impressions, its clicks and its average position.

Two things to pull out. First, queries with high impressions and low clicks: you are visible and not chosen, which is a title and description problem or a wrong-page problem. Second, anything sitting between positions 8 and 15 on a commercial term, because moving to the top of page one from there is the cheapest traffic available to you.

Compare a period against the same period last year, not against last month. Almost every retail category has a seasonal shape, and month-on-month comparisons mistake it for a trend.

2. Google Trends: direction and season, not volume

Trends reports relative interest on a 0 to 100 scale, not a number of searches. That constraint is the thing most people using it get wrong, and it decides what the tool is good for.

• It answers: is this rising or falling, when in the year does it peak, and which of these two terms is bigger.
• It does not answer: how many people search this. A term at 100 might be a thousand searches a month or a million.
• Set the region to United Kingdom. The worldwide default mixes in markets with different seasons, and it will show you a summer peak for a category that peaks in a British winter.
• Use a five-year window for seasonality and a twelve-month one for direction. A short window turns normal variation into a story.

3. The search terms report: what people typed before they bought

This is the highest-value source on the list and it is only available if you run paid search. It shows the actual query behind each click, alongside whether that click converted.

That combination is what nothing else gives you. Every other source tells you what people search. This one tells you which of those searches turns into money, which is how you find out that a high-volume head term costs you a fortune and a specific, boring, four-word phrase is where the sales come from.

It is also the fastest way to find products people want that you do not sell, and language your customers use that your site does not.

4. Official statistics: the backdrop your own data cannot see

The Office for National Statistics publishes UK retail sales monthly, including the internet share of total retail. It will not tell you anything about your store, and that is the point: it tells you whether a soft quarter was you or the category.

Statcounter is worth the same kind of use. Its UK figures for July 2026 put desktop at 50.6% of traffic, mobile at 45.05% and tablet at 4.35%, which is a useful corrective to the received wisdom that mobile is a clear majority. Your own analytics still overrule it, because the split moves by category and by channel.

The method

1. Write the question down first. "Should we stock more of X before autumn" is answerable. "What are the trends in our market" is not.
2. Check your own data before anyone else’s. Search Console and your own sales history know your customers. The market does not.
3. Test whether it is seasonal. Five years in Google Trends, region set to the UK.
4. Look for the same movement in a second source. One source is an observation. Two agreeing is a finding.
5. Say what would change your mind, then set a date to look again.

What this looks like when it works

On the Skribl Box account the research came before the content rather than after it, which is why the keyword work and the traffic moved together rather than one arriving without the other.