Competitor research answers a defined evidence question about an already chosen rival or rival set. Its deliverable is a compact packet of observations, limits, and unresolved facts that an analyst can inspect without reopening a pile of tabs.
Write a brief that can end
Use the competitor discovery method if the set is unsettled. For a research project, record the precise question, accepted competitors, buyer and market scope, decision deadline, exclusions, and person who will receive the packet. Ask what observation would answer the question and what evidence would still leave it open. A search result mentioning another company is a lead for later validation, not permission to expand this project.
Consider a hypothetical workflow vendor, Vale Works, preparing an October packaging review. Its question is narrow: as of late September, does Northbend publicly include approval history in its UK Team plan? Northbend is already a validated rival. The researcher's assignment is to document visible eligibility and any conflicting terms by September 22, not to judge whether Vale should change its package.
Specify the evidence and the stop rule together
For the public-offer question, a dated capture of Northbend's UK plan page can establish what Northbend currently displays. Current terms or plan-specific documentation would strengthen the eligibility claim. An earlier comparable capture could show a change, if its date and scope are known. None of those records proves actual customer access, adoption, or buyer demand. Write those proof limits into the request before collecting.
The competitive intelligence sources guide explains how to select and assess different inputs. This project needs only the sources capable of answering its stated question. It can stop when the public display is documented and the eligibility conflict is resolved, or when the deadline arrives with the conflict and next possible check recorded. More pages cannot substitute for missing terms.
Record observations so someone else can retrace them
Assign each accepted item an ID. Keep its original location, bounded observation, observed date, any stated effective date, entity and offer scope, and a caveat that matters to this claim. Preserve a failed or inaccessible check as a gap rather than recording 'no change.' A URL without the inspected passage and date is a weak handoff because the page may change later.
| Evidence | Dated observation | What it can establish |
|---|---|---|
| E-01 · UK plan page | Captured September 18: Team column lists approval history | Northbend's visible public claim on that date |
| E-02 · UK help page | Captured September 19: eligibility text says Business; page update date absent | A conflicting public claim, not proof that Team lacks access |
| E-03 · earlier UK plan capture | Dated June 11: Business column lists approval history | Earlier displayed placement, if the archived plan is authentic and comparable |
Those records come from different pages, but E-01 and E-02 still originate with the same company. They do not amount to independent verification of working entitlement. An internal note that one Vale prospect asked for approval history could be logged separately as a scoped buyer report; it cannot establish how often UK agencies value the feature.
Resolve the scope before resolving the contradiction
The researcher checks whether E-01 and E-02 refer to the same region, plan name, product version, and current period. If the help page describes an old version, the disagreement may be explainable; if it does not, both claims remain in the packet. Do not silently choose the more convenient page or treat the older capture as a precise change date.
For a material claim, triangulate with evidence that adds something new: dated terms, a plan-specific support answer, or a permitted first-party customer observation of actual access. A syndicated copy of E-01 adds no independent support. Here the available public record supports 'Team is listed on the current UK plan page' but leaves actual eligibility and change timing unresolved.
Close the project with a usable packet
At the September 22 cutoff, the researcher hands off the question and scope, E-01 through E-03 with their original locations, the exact supported public claim, the contradictory help text, the unknown eligibility and effective date, and the check that could settle each gap. The stop reason is deadline with an explicit dispute, not complete verification. The packet contains no recommendation about Vale's packaging.
A short cover note can say which evidence request is answered, which remains open, and whether the missing fact could affect the pending decision. The competitor analysis framework takes over when the team interprets the evidence, weighs alternative explanations, and records an action. Research should leave that judgment visible rather than smuggle it into the observation ledger.
Where Content Radar fits
Content Radar can organize the competitor and bounded public publishing inputs from supported RSS, Atom, and compatible sitemap paths. Structured-source items may become Articles; user-created Google Alerts RSS and manual Candidate URL submissions use a review queue. Source health and ingestion runs can reveal collection gaps.
Content Radar does not verify competitor claims, capture private buyer evidence, resolve conflicting terms, or assemble this research packet automatically. The researcher still owns the question, dated captures, proof limits, and handoff.
Keep the supported public inputs traceable
Use Content Radar for bounded public publishing observations while your team owns the research question and evidence standard.