Competitor evidence is too old when it no longer supports the decision you are making now. The date alone cannot tell you that. A record becomes dangerous when the market, product, search result, ownership structure or decision context has moved beyond what the original observation captured.
Why old competitor evidence becomes dangerous
Old evidence usually does not fail like a broken file. It fails quietly. The screenshot still opens, the source still exists and the statement may remain historically correct. The problem is that the decision has moved on. A pricing page can change after a package launch. A result page can gain an AI Overview or lose the competitors that shaped the original content plan. A company can become part of a larger platform while the old competitor profile still describes it as independent.
The page is live, but the package changed
A screenshot from last month can still show what was offered then. It cannot safely support today’s pricing comparison after limits, bundling or discounts change.
The query is the same, but the surface changed
A prior SERP sample can become exposed after an AI Overview, new feature, locale shift or different competitor set changes what users encounter.
The company is the same name, but not the same strategic unit
An acquisition can alter resources, ecosystem position and the relevant competitor boundary without making the older company record false.
Old evidence can stay true and still become unsafe for today’s decision.
How one competitor record changed state
A February 2025 record could accurately describe Semrush as an independent SEO and online-visibility software company. Later product and ownership events narrowed what that record could support.
The important distinction is not whether the old record was “wrong.” It is whether it could still control a July 2026 competitor decision. Semrush launched AI-search products, Adobe announced and completed an acquisition, and Adobe later presented Semrush capabilities inside a broader brand-visibility offer. Those events changed product scope, ownership context and the unit being compared.
The original record moved out of current control
The old record is preserved; newer evidence controls the current classification.
Original February 2025 record
- 1CurrentIndependent SEO-company record
- 2Time-ExposedMar–Oct 2025AI scope expands
- 3RevalidateAcquisition plan
- 4SupersededAcquisition completed
- 5Historical OnlyArchive layerChronology and audit
The original record remains available, but it no longer controls the current competitor classification.
Three moments carried most of the decision weight. AI Toolkit and Semrush One widened the product comparison. Adobe’s acquisition announcement made ownership assumptions uncertain. Completion of the acquisition and the later Adobe Brand Visibility announcement supplied the replacement evidence for a current competitor map.
Complete chronology record
| Date | Documented change | Why the old record became exposed | Lifecycle consequence |
|---|---|---|---|
| 4 Mar 2025 | AI Toolkit launched | AI-search visibility became a material product dimension. | Time-Exposed |
| 29 Oct 2025 | Semrush One launched | The offer expanded beyond a conventional SEO-tool comparison. | Time-Exposed |
| 19 Nov 2025 | Adobe announced an agreement to acquire Semrush | Ownership and strategic-resource assumptions became uncertain. | Revalidate |
| 28 Apr 2026 | Adobe completed the acquisition | Semrush was no longer an independent company. | Superseded |
| 17 Jun 2026 | Adobe Brand Visibility incorporated Semrush capabilities | The relevant competitive unit extended into a broader Adobe offer. | Replacement evidence becomes Current |
Decision change: a current comparison now needs Adobe ownership, ecosystem context and AI-visibility scope. The February 2025 record is retained for history, but it cannot control the present classification by itself.
The method in one view
After the practical problem is clear, the method is simple: test how the evidence relates to the decision—not merely how many days have passed.
- Question
- Can this record still support this exact claim, context and decision?
- Rule
- Combine observation age, condition volatility and decision distance; then search for a superseding event.
- Action
- Use, monitor, revalidate, supersede or preserve as history.
- Boundary
- Freshness governs permitted use. Source reliability and outcomes require separate evidence.
The Three Clocks
No single clock decides whether evidence is usable. The pressure comes from their collision.
Three Clocks Collision Model
Three dimensions converge on Evidence Pressure, which governs the active lifecycle. Historical Only sits outside that flow as an archive category.
Combine the three clocks
Governs active use
Governed output: Current
Read the model: pressure rises as evidence ages, the underlying condition becomes more volatile and the current decision moves farther from the original context. A confirmed material event overrides the score and produces Superseded.
- Observation Age
- How long since the condition was actually observed—not merely when the page was updated.
- Condition Volatility
- How quickly the price, product, result page, ownership or market condition can change.
- Decision Distance
- How far today’s decision has moved from the original query, market, device, product or strategic context.
Past review point + high volatility + partly changed context → Revalidate. A confirmed superseding event overrides the total.
Five lifecycle states
The active decision path has four states. Historical Only is a separate preservation category, not the next operational step after Superseded.
- 1
Current
Context-matched and uncontradicted for the stated decision.
Use in scope. - 2
Time-Exposed
Near or past a review point, or attached to a volatile condition.
Search for contradiction and monitor. - 3
Revalidate
The current decision depends on a claim that may have changed.
Obtain a new direct observation. - 4
Superseded
Newer evidence changes what the old record can prove.
Link the replacement record.
Practical questions about competitor-data age
How old is too old for SEO competitor data?
There is no universal number. Live SERP composition may need a new sample within days, while a dated corporate fact can remain useful for years. Revalidate when the condition is volatile, the decision is material, the context changed or newer evidence creates a credible contradiction.
Should competitor analysis be updated monthly or quarterly?
Use a calendar review to prevent neglect, then add event triggers. Monthly or quarterly reviews can be reasonable for a stable market, but pricing changes, launches, acquisitions, market entry and major search-surface changes can require immediate review.
Is a recently updated webpage automatically fresh evidence?
No. A recent timestamp may sit above older data, a live URL may describe a past condition, and a derivative summary may omit the collection method. Record when the condition was observed, what was observed and how it was collected.
Can old evidence still be useful?
Yes. A dated acquisition announcement can remain strong evidence for the historical event it records. The same document may be unsuitable for describing the company’s current product scope. The claim and decision determine the permitted use.
Should superseded evidence be deleted?
Usually not when auditability matters. Preserve the original record, link the evidence that replaced it, record the state change and prevent the old record from controlling current recommendations.
Check a record in 30 seconds
Use the same three inputs as the model, then add any superseding event. The result is a decision aid—not a probability of truth.
How often should competitor evidence be refreshed?
Use scheduled reviews as reminders, not expiry laws. The correct cadence depends on the evidence type and the events that can invalidate the decision sooner.
| Observed public guidance | Typical cadence | Important qualification |
|---|---|---|
| Panoramata | Weekly tracking plus an annual comprehensive analysis | Frequency varies with niche, goals and market pace. |
| CopyPress | Quarterly smaller analyses plus an annual comprehensive analysis | Fast-changing markets may need more frequent review. |
| Crayon | Continuous monitoring of material competitor changes | Event signals complement, rather than replace, structured review. |
| SERPIFAI operating rule | Calendar review + immediate event trigger | A material launch, pricing change, acquisition or search-surface shift overrides the calendar. |
Evidence-type revalidation matrix
| Evidence type | Main decay mechanism | Decision-safe verification trigger |
|---|---|---|
| SERP composition | Rankings, features, location, device and visible competitors change. | Replay the current query, locale, device and decision window. |
| Pricing and packaging | Plans, limits, discounts and bundles change. | Verify the pricing page and checkout near the decision. |
| Product capability | Features launch, move, integrate or retire. | Review a direct product source after a release or rebrand. |
| Ownership | Acquisition, disposal or restructuring changes strategic context. | Check the announcement, completion status and current corporate source. |
| Market geography | Availability, regulation and local competitors vary. | Collect market-specific evidence for the target geography. |
| Customer perception | Reviews and sentiment shift with releases and incidents. | Use a recent, bounded sample with collection dates and source mix. |
| Traffic or keyword estimate | Provider models, coverage and estimates change. | Record provider, date, market and model context; refresh before material use. |
| Historical corporate fact | Usually stable, but corrections or legal changes can occur. | Verify against a direct filing or authoritative dated announcement. |
Claim-specific expiry examples
These are review defaults for the worked examples, not universal market laws. Material events can override them immediately.
| Claim type | Review default | Immediate trigger | Permitted use |
|---|---|---|---|
| Live SERP composition | 7 days | Major search feature, locale, device or competitor change | Only inside the recorded query and observation context. |
| Public pricing or package | 14 days | Pricing-page, checkout or contract change | Public packaging only—not negotiated terms or realised revenue. |
| Product scope | 30 days | Launch, integration, rebrand or retirement | Current only after direct product verification. |
| Ownership transition | 30 days during transition | Announcement, completion, regulatory change or disposal | Current ownership context plus dated historical event. |
| Historical chronology | Annual review | Correction or source withdrawal | Chronology and audit; not an unstated current-product claim. |
Preserve what changed with an Evidence Passport
Do not overwrite the old record. Preserve enough information to reconstruct why its permitted use changed.
Evidence Passport field specification
| Field | Required entry |
|---|---|
| Evidence ID and original claim | A stable identifier and the proposition the record directly supported. |
| Observation date and context | When, where and how the condition was observed. |
| Previous and new state | The lifecycle transition and transition date. |
| Superseding evidence | The linked source or observation that changed permitted use. |
| Decision impact | The classification, recommendation or scope that changed. |
| Reviewer and next trigger | Who changed the state, why, and what must be checked next. |
Evidence boundaries
What this release establishes—and what still requires separate evidence.
Published decision method
Three Clocks logic, active lifecycle states, populated reference tables, worked chronology, checker rules and workbook artifacts are present and reproducible.
Independent reviewer agreement
No blind reviewer study or chance-corrected agreement result is included. Reviewers may still disagree about when a context change is material.
Performance outcomes
No permissioned dataset establishes ranking, traffic, lead, revenue, ROI or causal performance outcomes.
When Evidence Becomes Too Old
How to read it: select a checkpoint to isolate that part of the method. The canvas shows the relationship; the inspector explains what the checkpoint changes in the decision.
Sources and method context
- UK Government — Data Quality Action Plan implementation guide, including timeliness as fitness for the intended period of use.
- W3C — PROV Data Model, distinguishing generation, use and invalidation in provenance records.
- Barrett and Turner (2025), on time-dependent value and relevance in a defined context of use.
- Semrush AI Toolkit announcement — 4 March 2025.
- Semrush One announcement — 29 October 2025.
- Adobe acquisition announcement — 19 November 2025.
- Adobe acquisition completion — 28 April 2026.
- Adobe Brand Visibility announcement — 17 June 2026.