One rising line does not prove local demand. It proves that one tool, under one set of filters, produced an upward line. A defensible rising-topic decision combines differently limited evidence: relative search interest, the website’s current search exposure, questions heard from real prospects or customers and a dated sample of the result environment. Those signals must point towards a decision the business can genuinely serve in a place and period that matter.
The output is not another keyword list. It is one of five actions: create, refresh, merge, watch or reject.
A trend earns an investigation, not a content brief. Until the team can name the buyer decision, current page owner, original contribution and review trigger, it has found something interesting, not something with permission to publish.
TL;DR
- Write the demand hypothesis first: audience, geography, period, topic form and buyer decision.
- Google Trends is relative interest scaled 0-100, not a search count; "Breakout" is a clue on an unknown base.
- Search Console shows this site's exposure; real customer questions add commercial weight; a dated result sample shows interpretation.
- Check who already owns the page, then force one action: create, refresh, merge, watch or reject.
Bound the demand claim before opening a tool
“More people are searching for heat pumps” cannot govern a local marketing decision. More people where, over which period, using which language and with what need? Write a hypothesis first: Among [audience] in [place], interest in [defined problem, service or question] appears to be changing during [period], creating a possible need for [buyer decision or action]. Then define:
- audience: homeowner, tenant, business buyer, landlord, visitor or another real group;
- geography: the area the evidence describes, not the territory marketing would like to claim;
- time: current period, year-on-year comparison, season or event window;
- topic form: exact term, broader topic, product, service or question family;
- buyer decision: what the person must understand, compare, diagnose or do;
- commercial fit: whether the organisation supplies a relevant answer or next step;
- capacity: whether demand can be served during the period;
- current owner: the page that may already owe the answer.
This prevents a national news spike becoming a Brighton service page and a short-lived curiosity becoming a permanent maintenance obligation.
Google Trends reports relative interest
Google Trends is useful for exploring change across time and geography. It does not provide an absolute monthly search count. Google’s Trends data FAQ explains that data is sampled, aggregated and normalised relative to total searches for the selected place and time, then scaled from 0 to 100 within the comparison. That means:
Google describes Trends as one data point rather than a scientific poll. That is the correct business boundary. Use the graph to decide what deserves investigation, not what deserves a URL.
Record whether you measured a term or a topic
A Trends search term and topic do different jobs. Google’s term-versus-topic guidance distinguishes exact entered wording from broader topic groupings which can include conceptually related searches and language variants. Use a term when exact emerging language matters. Use a topic when the underlying concept matters more than one phrase. Record the selection, geography, period, category, search type, comparison and capture date.
If the intended local geography produces no graph, do not convert missing Trends data into missing demand. Google’s troubleshooting guidance notes that insufficient query volume can prevent display. Widening time or geography may help exploration, but the resulting evidence must keep the wider scope attached. The honest action may remain watch, supported by other local evidence.
“Breakout” is not a purchase order
Related searches can expose changing language, but their growth labels require context. Google says rising related searches are terms with substantial growth in the selected period relative to the prior one; Breakout indicates very large percentage growth. A tiny baseline can still generate a spectacular percentage.
For each candidate, record the parent term or topic; location, category and period; rising percentage or label; previous comparison window; the actual buyer problem suggested; corroboration from another source; and whether growth persists when the time range changes. Do not paste “Breakout” into a content calendar as if it were a signed commission. It is a strong clue attached to an unknown absolute base.
Search Console measures this site’s exposure
Growing Search Console impressions can show that the property appeared more often for a defined query set. They do not prove the market grew or explain why exposure changed. Google’s Search Console query-data documentation notes privacy filtering and reporting limits. The table is not a complete census of all searches. Use it to ask:
- Are impressions changing for a manually defined query family?
- What happened to clicks and average position?
- Which existing pages received the exposure?
- Is movement concentrated by country, device or search appearance?
- Did ranking mix, page mix or query wording change?
- Is this new growth, seasonality or recovery?
- Are multiple pages appearing for the same decision?
Compare equivalent windows. Weekly or monthly views can reduce day-of-week noise; seasonal demand also needs year-on-year context. Zero visible query rows can mean no demand, no exposure, anonymisation, low volume, reporting limits, the wrong property or no relevant page. Search Console cannot distinguish those possibilities alone.
Real customer questions add commercial weight
The strongest corroboration often sits outside SEO tools: calls, forms, chats, sales conversations, quotes, support tickets and reviews. Collect patterns rather than anecdotes. Preserve the exact redacted question, the prospect or customer stage, the relevant service and location, date and channel, occurrence count and denominator where available, whether the uncertainty affected qualification, timing or purchase, and whether the current website answers it.
“Sales keeps hearing this” is a useful lead. A robust record sounds like: [count] of [denominator] qualified enquiries between [dates] asked [redacted question]. That still does not prove search volume. It does prove a bounded audience and commercial decision without inventing demand. Protect personal information. The evidence record needs language, counts and context, not copied identities or private correspondence.
My preferred opportunity is a repeated buyer question the website cannot answer while Trends or Search Console begins showing the same direction. The sources are imperfect in different ways; their agreement is valuable precisely because they do not share the same weakness.
The live result reveals interpretation, not market size
A dated search-result sample can show how a phrase is currently interpreted: news, local services, definitions, products, video or mixed intent. It cannot measure the market. Capture exact query, date, stated location, device, logged-in state, result types, leading page subjects and visible related questions. Then identify which current page, if any, could serve the intent.
This check catches semantic traps. A promising phrase in a tool may refer to a television programme, national policy debate or unrelated product. One search sample remains variable evidence, but it can disprove a lazy commercial interpretation. Do not mistake current competitors for the content brief either. Their headings show what exists. They do not prove what the local buyer still needs or what the business is entitled to claim.
Local events can create temporary demand
Weather, regulation, school calendars, road works, grants, tourism, product releases and local incidents can all move search behaviour. Attach the suspected driver to a source and date. Ask whether it affects the actual audience and service area, whether it is recurring or one-off, whether the business can serve demand and what happens when it expires.
A festival may justify a temporary logistics update without granting permission for an evergreen location article. A recurring winter fault may justify a maintained diagnostic. A proposed regulation may deserve a watch record until its text and effective date are confirmed. The event explains timing. It does not create permanent page ownership.
Got a rising line and a content calendar waiting for it? We run the evidence card with you and tell you whether the honest action is create, refresh, merge, watch or reject.
Put the signals on one evidence card
Use a Rising Local Topic Card:
| Field | Record |
|---|---|
| Hypothesis | Defined problem, service or decision, not a synonym bag |
| Audience and geography | Who and where the evidence concerns |
| Decision and capacity | What the reader needs and whether the business can serve it |
| Trends | Term/topic, settings, periods, direction and capture date |
| Search Console | Property, query set, pages, metrics, windows and limitations |
| First-party evidence | Redacted questions, count, denominator, channel and dates |
| Result sample | Query, place, time, device, result types and interpreted intent |
| External driver | Event, season, policy or local condition and applicable date |
| Current owner | Existing page and the decision it owns |
| Conflicts | Missing data and alternative explanations |
| Action | Create, refresh, merge, watch or reject |
| Reopening trigger | Date or event |
| Owner | Accountable person |
Do not calculate a fake precision score. A total cannot repair unbounded geography, duplicated evidence or a missing buyer decision.
Inspect page ownership before creating
A rising phrase may already belong to a service page, location page, comparison, product or existing article. Growth does not revoke that ownership. Check whether the current page answers the question directly, matches audience and geography, carries current evidence and already earns exposure. Ask whether a new URL would perform a distinct task or repeat the same claim with fresher wording.
Default to the smallest sufficient change. Refresh the existing owner when it already owns the decision. Merge fragments when several pages compete. Create only when the topic carries a separate task which cannot be completed properly inside the current page. This is the same discipline applied to content-gap analysis, but the evidence object differs. A gap tool begins with competitor or keyword absence. A rising-topic card begins with temporal change and must prove that the change belongs to the local business.
Force one of five actions
The card ends in one decision, and each has its own condition:
Action 1
Create
Differently limited sources support a bounded distinct need; no adequate owner exists; original value is available.
Action 2
Refresh
Demand or language changes, but a current page owns the decision.
Action 3
Merge
Several pages or fragments split one emerging question.
Action 4
Watch
The signal is early, local data thin, sources conflicted or the driver unresolved.
Action 5
Reject
Topic is irrelevant, unserviceable, misinterpreted, expired or differentiated only by wording or place.
Watch is not indecision. It is evidence discipline before the market becomes clear. Reject matters too: without a rejection record, the same shiny query returns in the next keyword export wearing a different row number. A create decision also needs information gain: operational experience, a real local constraint, measurements, decision tool, accountable specialist explanation, project evidence or useful synthesis. A chart screenshot and rewritten definition are not enough.
Let timing govern the content decision
Recurring seasonal topics need lead time before demand peaks. One-off urgent topics may need quick publication and a clear expiry route. Record production and approval time, operational lead time, review before the expected peak and expiry or evergreen-conversion conditions. Do not keep editing because a daily graph wiggles.
Each action needs its own observation plan. Create tracks discovery and qualified actions without claiming causation. Refresh annotates the release and compares equivalent periods. Merge checks whether ownership consolidates. Watch returns to the same bounded signals. Reject reopens only when new evidence changes the reason.
The graph does not get the final vote. The decision belongs to the combined evidence: a real audience, bounded geography, recurring or timely problem, distinct buyer decision, service capacity, current-owner check and an original answer worth publishing. At first glance, a missed wave and a manufactured wave can share the same upward line. The evidence card exists to tell them apart before the content calendar pretends it already knows.
Good to know
FAQs
Not on its own. Trends reports relative interest scaled 0-100 within the selected view, sampled and normalised, and a missing local graph can mean insufficient displayable data rather than no demand. Treat it as a reason to investigate, not a search count.
A related search with very large percentage growth compared with the previous period. A tiny baseline can still produce a spectacular percentage, so record the comparison window, check whether growth persists across time ranges and corroborate it from another source.
Only when differently limited sources support a bounded distinct need, no existing page owns the decision and you can add original value. Otherwise refresh the current owner, merge competing fragments, watch the signal or reject it with a recorded reason.
A repeated buyer question the website cannot answer, recorded with count and denominator, while Trends or Search Console begins showing the same direction. The sources fail in different ways, so their agreement carries weight.
Next step
Want a trend decision you can defend?
We build the Rising Local Topic Card from Trends, Search Console, your enquiry records and a dated result sample, and return one action with its reopening trigger.