A practical research workflow

Use Google Trends to Research a SaaS Idea

Use Google Trends to examine the timing and geography of search attention around a software task. Save the settings with every observation; a chart without a scope is not reusable evidence.

By TopTrending · Reviewed October 29, 2026

Start the workflow Research your keyword

Reproduce a software-topic investigation

Our worked investigation starts with the search term “product feedback”. Use United States, past five years, Web Search. Leave category at All categories. Record whether you chose a search term or a Google topic before comparing anything.

Open Explore and enter product feedback. Choose “Search term” for exact query framing; a topic can include related concepts and languages, so it is a different series.
Set United States, past five years, All categories and Web Search. Save the URL, date and timezone. Note whether the most recent interval is incomplete.
Inspect the entire period before zooming into the past 12 months. Look for repeated peaks across multiple years; one spike does not establish seasonality.
For sparse output, broaden the period or region one setting at a time and save a separate observation. Do not substitute a broader topic’s values for the original query.

For an exact recorded run and any access or data limits, see the dated operation log and the observations below.

Read the recorded run with its limits

We operated Google Trends for product feedback and the narrower product feedback digest for small SaaS teams, United States, past five years, all categories, Web Search. Both browser-generated weekly data requests returned 262 points. The broad term had continuous relative-interest data; the narrow phrase was almost entirely sparse with one isolated index spike. These are relative search-interest observations, not monthly volume or evidence of paid SaaS demand.

Recorded weekly relative index, US, Web Search. 261 complete data points; 1 incomplete point excluded from the line. Peak: 100, week starting 2026-02-22 UTC. Latest week (2026-10-25) is incomplete and its current index of 28 is not a completed-week comparison. Weekly buckets can straddle the requested date boundary. Download all recorded points and settings.

Settings: search term product feedback; country US; requested dates 2021-10-29 to 2026-10-29; all categories; Web Search; weekly series in UTC.
The classic interface currently links to the new Explore interface. The new interface was operated first, but chart and query cells remained blank. The classic interface supplied widget configuration; its generated multiline request returned HTTP 200. No chart was invented from the blank interface.
262 returned weekly points begin with the UTC week of 2021-10-24 (8) and end with 2026-10-25 (28). Weekly buckets may overlap requested boundaries. The final point is explicitly partial and should be excluded from complete-week comparisons.
The observed maximum is 100 in the UTC week of 2026-02-22. This is an index normalized within this request, not 100 searches. No returned week has a zero value.
Complete-year peak weeks were 2022-10-09 (17), 2023-10-29 (20), 2024-11-17 (26), and 2025-11-02 (82). Earlier autumn peaks are a reason to investigate seasonality, not a confirmed recurring seasonal effect or a forecast; the observed 2026 peak is in February.
Autocomplete offered product feedback as a search term and five unrelated book topics. We retained the term rather than substituting an unrelated topic. A topic can group a concept across languages; it is not interchangeable with this term.
A second actual run used product feedback digest for small SaaS teams with the same US / five-year / All / Web settings. It returned HTTP 200 and 262 weekly points: 261 have value 0 / hasData=false; one week (2025-03-16 UTC) is 100 / hasData=true. This isolated spike does not establish sustained interest or search counts. Google warns sparse queries can show statistical noise. We leave volume and demand unknown and return to the broader relevant seed before seeking customer evidence.

Download the sparse-query comparison. Keep the provider’s hasData flag alongside every reported zero; it is not zero monthly search volume.

Observed 2026-10-29. Reopen the recorded settings. Conditions can change; repeat the run before relying on it. The recorded link uses a rolling five-year window. To revisit the original requested period, set October 29, 2021 through October 29, 2026; preserve the downloaded record because sampling and historical revisions can still differ.

Turn relative interest into a research question

Google normalizes sampled search interest within the selected time and geography. A score of 100 represents the peak in that selection, not 100 searches. A displayed zero can reflect insufficient data; it does not establish zero demand. Downloaded values from different settings are not interchangeable or additive.

A repeated pattern could suggest when to interview users or publish a resource. It cannot tell you how many teams pay for feedback software. A flat or sparse niche query can still describe a valuable task; record the gap and test it with other evidence.

Read Google’s Trends data FAQ and term and topic comparison guidance. These are official explanations; our operation log separately states what we actually observed.

Connect attention to an actual software decision

Next, research the candidate in TopTrending, retaining its source, country and period. Use the validation ledger to connect queries to existing supply and a user experiment. Use the Google Trends alternatives comparison if you need a different discovery workflow.

Download the settings and observation worksheet. No absolute volumes, seasonal conclusion or market validation should be filled in without actual supporting evidence.

Check your own seed keyword

Move from a proposed task to keyword evidence. Custom research follows the existing account and credit limits.