Choose a baseline that matches the decision: compare a topic with its own past, a relevant peer, or a recurring period, and explain why that comparison is useful.
Three useful starting points
A previous equivalent period helps identify recent change. The same period in an earlier year can reveal seasonality. A carefully selected peer can provide context, but it must measure a comparable concept in the same place and window.
An illustrative check
A small cycling shop notices interest in puncture repairs rising in April. It compares the same UK location with March and with the previous April. If the spring rise repeats, the owner may plan stock earlier; if it is unusual, they can investigate a local event or other explanation. The example is illustrative, not a measured finding.
Keep the comparison honest
Record the source, query, dates, geography and reason for choosing the baseline before interpreting the result. Google Trends provides relative, sampled data, so changes in its index should not be converted into customer counts. A baseline describes a difference; it does not prove a cause.
Match the baseline to the decision
| Decision | Comparison to consider | Limitation |
|---|---|---|
| Is this week unusual? | Previous equivalent week | A holiday may distort the comparison |
| Is this year's seasonal peak different? | Same period in earlier years | Events and conditions change |
| Is one topic gaining relative attention? | A relevant peer in the same view | The topics may serve different audiences |
Google Trends quick comparisons offer preceding-period and, for some ranges, same-period historical benchmarks. Check the available options for the chosen range in Google's explanation. If the baseline is unusually high or near zero, a percentage change can mislead; show the underlying series and explain the starting point.
Before sharing a result, write: “Compared with [baseline], the relative search-interest index [changed in this way], in [place and period].” Add what the data cannot establish and what evidence could challenge your interpretation.
When two baselines should not be combined
A comparison with last week and one with the same week last year answer different questions. Do not average their percentage changes into one "trend strength" number without a defined method. A high prior-week baseline can make normal interest look weak, while a low one can exaggerate a modest rise.
Keep both results visible, label each question and explain disagreements. If they point in different directions, investigate the calendar, search wording and local context rather than choosing whichever comparison supports your preferred conclusion.
Where to go next
Read Google's guidance for the underlying methodology.
For a broader research workflow, see how to analyse a trend and trend momentum explained. TrendRadar is free to download and use for exploring trends by country.
Free on iPhone and Android
Explore trends by country with the free TrendRadar app
Explore trending searches, popular YouTube videos and Apple Music charts by country. Compare trend scores and momentum, and save topics worth investigating. Free to use on iPhone and Android. No subscriptions or in-app purchases.
