Resolve Conflicting Company Facts with Evidence

A company website lists three offices, an older brochure lists five, and a directory lists four. Picking the number that fits a sales hypothesis may feel efficient, but it leaves the next person with an unreliable record. Start by asking whether the sources are answering the same question. Offices, operating sites, registered addresses, and service locations may all be counted differently.

Define the fact before ranking the sources

Write the disputed field in precise language. “Number of active customer-facing branches in the country” is more useful than “locations.” Include the relevant geography, entity, time period, and unit. Once the question is clear, a source that initially looked contradictory may turn out to describe another part of the business.

Compare publication dates, observation dates, and effective dates separately. An undated page is not automatically current because you opened it today. A dated filing may be authoritative for a legal identity but unsuitable for a recent product launch. Choose authority by the specific claim, rather than treating one source as best for everything.

Create a compact evidence comparison

Use one short entry per claim. Record the value, source, date, scope, and any qualification. Preserve enough context to explain the difference without copying whole pages into the account. W3C's provenance overview links data to its origins; your comparison should make those origins easy to inspect.

• Does the source describe the correct legal entity or group?

• Is the statement first-hand, attributed, or an estimate?

• Does its date match the decision you are making?

• Could different definitions explain the disagreement?

• Would choosing either value change the next business action?

If you are considering an additional research service, ask about the US$15/month offer. Leads Automation is an independent service and is not affiliated with LinkedIn; this is not LinkedIn’s official price. Confirm the plan, eligibility, and terms before ordering.

Choose an outcome other than forced agreement

There are several useful resolutions. You can select a supported value, store a range with its meaning, split one field into two better-defined fields, or leave the fact unresolved. A range is appropriate only when the values measure a comparable thing. Combining group revenue and subsidiary revenue into a range would conceal the actual issue.

If uncertainty does not affect the immediate decision, document it and move on. If it changes service eligibility, ownership, or a material claim in a proposal, escalate. The research escalation path should identify who can decide whether more evidence is worth obtaining.

Fictional example: conflicting staff counts

In a fictional review of Birch Analytics, an old presentation mentions 120 employees, a current company page describes a team of about 90, and a parent-group report lists 260 people. The researcher does not average the figures. The group report covers several businesses, while the presentation predates a business-unit sale described in a later company announcement.

The account record keeps the current company description as an approximate company-level observation, with its source date and wording. The exact count remains unknown. A qualification rule requiring more than 100 employees cannot be marked satisfied solely from these sources. The fictional account goes to review rather than being accepted because the oldest figure was convenient.

Record why a value won

A short decision note should state why the selected evidence better matches the question. “Used the latest source” is weaker than “used the current company-level page because the annual report covers the parent group.” This explanation helps a later reviewer notice when the underlying reason no longer applies.

Add the resolution to an account change log, especially when a prior decision depended on the old value. Identify downstream records or plans that need review. Do not silently correct the main account while leaving an outdated claim in a handoff document.

Use the same evidence standard when evaluating the US$15/month offer: ask for clear written details about the service rather than assuming particular data sources, accuracy levels, or product features are included.

Stop researching when the remaining uncertainty is clearly stated and proportionate to the decision. A credible unresolved answer is more useful than a precise number that the evidence cannot support.