When to refresh an industrial prospect list

A list is a decision against a brief, supported by evidence at a point in time. A change can affect one field, one facility or the whole account. Start by naming the question that has reopened rather than treating every record as equally uncertain.

• Your brief changes. A new territory, application, company type or exclusion can change which accounts belong. Record the new profile version before reassessing fit.
• Company identity becomes uncertain. A renamed website, group announcement or different trading name is a reason to check the entity—not evidence that two businesses are the same.
• An operating site changes. An opening, closure or change of activity may affect a facility but not every company in the group. Check the company-to-site relationship and other relevant operations.
• The commercial role changes. A manufacturer, integrator, operator and reseller can share product words while being different customers for your range. A business may have more than one role.
• The evidence no longer supports the decision. A removed page, conflicting statement or unresolved source-use condition should reopen the affected check. A broken link alone does not establish that a company has closed.
• An existing-account or channel exclusion changes. Reconcile the client's current exclusion instructions in its controlled workflow before placing the account back into an active shortlist.

For initial discovery and the company-versus-site distinction, use the Irish manufacturing account-research guide. Maintenance begins after that first decision; it should not silently erase how the original list was built.

1. Keep the starting snapshot. Preserve the record ID, brief/version, prior decision and evidence references. Work on a review copy rather than overwriting the only account history.
2. Scope the change. Identify the field, site or inclusion rule affected. Separate an observed change from an inference or a question still needing an answer.
3. Check the evidence. Verify that it concerns the same company and relevant activity. Record what the source says, its date where available, when it was checked and whether the intended use is permitted. Missing dates stay unknown.
4. Make a bounded decision. Retain the supported account, hold an unresolved one, correct a verified field or remove a confirmed non-match from the current shortlist. Recheck the relevant brief and exclusions before release.
5. Leave an owned next step. Record the reason, changed fields, reviewer and next check or review trigger. Keep held accounts separate; do not let them drift back into accepted totals.

The change log supports a review; it does not automatically validate the source, clear reuse rights or approve delivery. A role match is not equipment suitability, buying intent or permission to contact anyone.

Do not settle an evidence conflict by choosing the page you opened most recently. First compare the same company, the same site, the same field and the period each statement describes. Then record what would resolve that particular disagreement. This exercise adds a field-level decision trail to the refresh workflow; it does not replace the account review.

Name the problem before changing the record

• Stale for this decision: evidence supports an earlier period, but its continuing relevance has not been established. Reopening an old page today does not make its underlying statement current. Preserve the historical observation; reopen the present-tense field.
• Contradictory: incompatible assertions concern the same entity, place, field and overlapping period. Preserve both references and hold the affected conclusion. A newer publication is not, by itself, proof that the older assertion was wrong.
• Missing: a required assertion has not been established. An absent page or absent detail does not supply the opposite fact. Ask for the missing relationship or activity, rather than marking it false.
• Different scope: a registered office and a production workshop, or a group and a subsidiary, may legitimately have different descriptions. Split their fields before deciding that they conflict. Resolving this apparent mismatch does not close a separate freshness gap.

These are issue labels, not mutually exclusive account statuses. One record can have stale activity evidence, a missing operator link and a contradictory location claim. Keep each issue visible; do not compress them into a confidence score.

Work one account through four evidence arrivals

Wholly fictional teaching exercise. All evidence below is invented, not a quotation, real prospect or anonymised client record. Company FCR-C01 and Workshop FCR-S01 are fictional. The abstract brief requires the company to operate relevant machinery production in Ireland. “Earlier period” and “review period” are exercise labels, not real dates. Source URLs, publication dates and retrieval dates are blank in the downloads because no actual sources exist.

1. FCR-R01 — an old statement is retrieved again. Card FCR-E01 establishes production at the workshop during an earlier period only. It says nothing about the review period. The activity field is stale for the present decision, not contradicted. Keep the earlier fact, hold current activity and ask for review-period evidence. Do not change the company's activity to “not manufacturing”.
2. FCR-R02 — an office description appears. Card FCR-E02 describes the same company's registered office as administrative-only; it does not describe Workshop FCR-S01. There is no like-for-like contradiction. Keep the office and workshop separate. The workshop's current activity remains unestablished, so the account remains held.
3. FCR-R03 — two statements actually collide. Card FCR-E03 says this company manufactures at this workshop throughout the review period. Card FCR-E04 says this same company performs no manufacturing at this same workshop throughout that same period. Both claims are stipulated as attributable, but neither explains the other. This is a material contradiction. Preserve both, withhold a current activity conclusion and ask which statement is corrected or whether a narrower time boundary resolves them. Do not count sources as votes.
4. FCR-R04 — a scoped correction arrives. In card FCR-E05, the reviewer verifies that the original issuer of FCR-E04 explicitly withdraws its own activity claim as erroneous and confirms FCR-E03 for the same company, workshop and review period. Within this fiction, that resolves the activity contradiction. Record E04 as superseded for this claim, retain its history and correct only the affected field. The current exclusion review is separately uncompleted: the overall account therefore stays on hold. Closing an evidence issue is not automatic shortlist acceptance.

Try the counterfactual: if E05 merely said “we continue to support Irish customers”, would R04 resolve? No. Customer geography does not answer whether this company manufactures at this workshop during the review period. Keep the contradiction open unless evidence addresses the disputed claim.

Leave a resolution record, not an overwritten cell

Use the blank field-level worksheet (CSV), the five invented evidence cards (CSV) and the four-event answer key (CSV). Read the cards before revealing the answer key. Make one worksheet copy per disputed field and review event. Reuse the account ID; these events are not additional companies.

1. Identify the brief version, account, site and precise field. Record the previous value and its evidence reference without altering the original.
2. List the competing evidence IDs separately. For each, distinguish publication time, the period the assertion concerns and your actual retrieval time. Leave unknown dates empty and explain the gap. Check scope and intended source use before retaining any source content.
3. State the unresolved question and the evidence needed to close it. Verify the correction’s origin and authority for the exact statement. It must identify the affected assertion and scope; recency, silence or a more polished website is insufficient.
4. Record the affected-field outcome separately from the whole-account decision. Name the internal reviewer role, next check or chosen review trigger, and every remaining gate. Preserve the earlier reference and explain any supersession; never silently replace the evidence.

The blank worksheet is a manual educational aid, not an approval engine. Keep restricted material outside it; use only a minimal permitted reference while source use is unresolved. Do not copy contact details or confidential exclusion lists into the exercise. None of these outcomes establishes buying intent, equipment suitability, an open package, delivery approval or permission to contact anyone. Real account acceptance still requires the agreed identity, fit, exclusion, source-use and independent review checks.

A source can contain several valid dates and still leave the date you need unresolved. Keep the date's label beside its value. Do not turn a catalogue update into a project update or a recent retrieval into a recent event.

The EPA's River Abstraction Pressures catalogue provides a real, inspectable example: it displays release 2017-11-20, update 2020-07-02, and period covered 2014-01-01 to 2015-01-01.[1] Those are different fields, not an inconsistency to repair. Our check on 9 October 2026 does not make that reporting period contemporary.

The openly licensed Remedial Action List catalogue displays release 2020-04-28, update 2020-07-24, and both period endpoints as 2010-11-09.[2] Treat these as the catalogue's displayed labels. Do not silently expand the identical endpoints into a reporting year, or use them to certify a current supply's status. These catalogues are useful for source selection and date discipline, not a freshly verified account list.

Field: release_label and release_value

Store: Exact visible label and value

Publication rule: Attribute the label; do not imply independent verification of publication timing.

Field: body_dateline

Store: Any separately stated date in the text

Publication rule: Keep it even when it differs from the release field.

Field: modified_value

Store: Explicit source modification date, if supplied

Publication rule: Do not assume that every substantive claim changed on that date.

Field: period_begin and period_end

Store: The stated reporting window

Publication rule: Keep its original precision and any unresolved ambiguity.

Field: event_date

Store: Date of the particular decision or event, if evidenced

Publication rule: Do not substitute the release, update or reporting date.

Field: retrieved_at

Store: Actual retrieval date or tool-recorded timestamp

Publication rule: Describes your check, not the age of the underlying event.

Field: date_resolution

Store: Clear, different scopes, conflict or unknown

Publication rule: A research judgement, not a revised publisher date.

An absent value stays unknown. A day-only source date must not acquire an invented time or timezone. If your system requires a single publication field, leave it unresolved and preserve both candidates in the evidence record; do not choose whichever makes the record appear fresher.

Fictional exercise: a regulator-style evidence card displays a release label of 12 May, exercise year and a body dateline of 13 May, exercise year. Its report concerns the preceding calendar year. These are invented labels, not assertions about any regulator publication.

Keep the two publication candidates, the separate reporting period and the actual date you reviewed the card. The output date label is publication timing unresolved. Reject a statement that the underlying supply conditions were assessed on the day you retrieved the page. Hold only the claim that requires the unresolved timing; do not discard separately supported background merely because a date label needs clarification.

The next check is narrow: seek a correction or authoritative publication record for the same item and version. A newer unrelated news release, a file-download timestamp or a search-engine date does not resolve these labels. If neither date changes the buyer's decision, explain that boundary rather than inventing precision.

An internal analyst can withdraw their own unsupported inference without claiming that the external publisher has issued a correction. Use different fields for analyst withdrawal and issuer correction. This distinction prevents a responsible research correction from becoming a false claim about what the source has admitted.

The following correction note is wholly fictional. It is not a correction to an actual DemandRange delivery or client brief.

Correction field: Affected assertion

Completed example: Claim COR-01: “Every location not recorded as completed is an available component-supply prospect.”

Correction field: Origin

Completed example: Analyst inference from fictional programme summary PRG-A; not a statement the programme publisher made.

Correction field: Action

Completed example: Withdraw COR-01 from the active brief. No revised total or available-package estimate.

Correction field: Reason

Completed example: Aggregate categories were unresolved; no dated project evidence established the remaining locations, stage or equipment-supply route.

Correction field: Retained observation

Completed example: PRG-A is a programme-level reference. Its disputed total is not used for account acceptance.

Correction field: Dependent field

Completed example: Clear the project-signal acceptance flag wherever its only support was COR-01. Mark it held, with the same claim ID.

Correction field: Unaffected evidence

Completed example: Separately verified company identity and industry fit remain attributed to their own sources. Their retention is not a finding of buying intent.

Correction field: Next evidence

Completed example: A dated project-level source resolving identity and stage, followed by separately evidenced supply scope and route.

Correction field: Remaining gates

Completed example: Source-use permission, account exclusions and human review remain separate.

Follow the claim ID into the brief, shortlist, downloadable report and any maintained summary. Record which versions were corrected and which superseded copies remain outside your control. Do not write “all copies corrected” unless you have checked every declared target. Keep the old statement in an authorised private audit trail; remove it from current decision-facing copy.

Use the programme correction register in the resource library as a manual claim-to-output inventory. It does not validate a project or issue a correction automatically. Before delivery, compare each affected output with the revised wording and capture the readback result. A rights hold can remain even after the factual problem is resolved.

For a scoped research enquiry, supply the old claim ID, source URL, date labels, affected versions and the decision that needs to be remade. Specify whether you need a historical explanation, a current project check or a revised account shortlist; those are different deliverables.

Wholly fictional teaching examples. Every business and source statement below is invented, not a researched prospect or client result. The example brief is company-level research on food-and-beverage manufacturers operating production facilities in Ireland. Treat the stated facts as stipulations within the exercise; real records still need identity, evidence, exclusions and source-use review.

Retain: the source moved, but the supported fit did not

Fictional Beverage Manufacturer A. Its old activity page has moved. In this scenario, the reviewer confirms the same company identity and a new permitted source describing the same Irish production operation; the other required checks remain satisfied. Retain the account and update the evidence reference. Keep the old reference in the history rather than presenting it as a newly checked source.

Hold: the change raises a question, not an answer

Fictional Food Manufacturer B. A newer statement describes a facility as a distribution location, while the earlier inclusion depended on manufacturing there. Other Irish production activity is unresolved. Hold the account for this brief and ask whether it still operates a relevant production facility in Ireland. Do not infer either continued manufacturing or company-wide closure.

Correct: a verified name change is not a new account

Fictional Process Manufacturer C. The scenario confirms that a changed trading name belongs to the same company; its relevant activity, territory and remaining checks still satisfy the brief. Correct the name field, retain the stable record ID and preserve the previous name as an alias in the change history. Do not count the renamed company twice.

Remove from the current shortlist: a criterion is no longer met

Fictional Ingredients Company D. The scenario explicitly confirms that the same company has ceased all relevant manufacturing activity and now only resells finished goods. That contradicts this brief's manufacturing requirement. Remove it from the active shortlist with the reason recorded. This is not a decision to delete every business record, and it says nothing about fit under another brief.

These are four separate cases, not four changes to one company or an expected distribution of outcomes. If a supposed correction makes identity or fit uncertain, use hold instead of silently accepting the amended record.

Set a review rule around the decisions you intend to make, the importance of the fields and the evidence available. Combine change-triggered checks with a review point you can actually resource. Do not choose an interval simply because a generic list-decay statistic sounds precise.

Before reusing an older shortlist for a different campaign or market, check the brief version, exclusions and unresolved holds. Prioritise accounts whose identity, operating activity or route is uncertain. A recent review date is not a substitute for evidence that supports the field.

Record a review due date only when it is genuinely assigned. If timing is unknown, leave the date blank and state the next trigger or question. Do not turn a planned check into a claim that monitoring is already operating.

Download the blank review log (CSV). It contains one blank starting row with a usage notice, ready to copy for additional review events. It includes no real companies, source dates, contacts or pre-approved decisions.

Keep one row per review event. Reuse the company record ID when several events affect the same account; event rows are not distinct-company counts. The fields separate the trigger, previous status, evidence references, identity/fit/source-use checks, decision, reason, changed fields, next action and owner.

Reference evidence held in your controlled source ledger. Keep source publication dates separate from your review date. Use retain, hold, correct or remove_from_active_list for the decision, with an explanation. The worksheet is an educational starting point, not a substitute for your retention, access or source-use controls.

For the underlying record structure, inspect the fictional industrial account-research sample. Maintenance status and research class answer different questions: retaining an account does not create a project signal, and a project signal does not establish an available equipment order.

Define the industry, territory, customer role and exclusions before commissioning a new list or a refresh. Keep the existing record IDs and explain which decisions need to be revisited. Any proposed scope should state the fields, evidence boundaries and acceptance checks rather than promise that every old record will remain usable.

Explore company-level industrial research or request a written scope. Named contacts and outreach are separate from the company-level examples in this guide.

The two actual catalogue records were checked on 9 October 2026. Their displayed dates are reproduced as metadata, not harmonised or certified as current project dates.[1][2]

Contains Irish Public Sector Data licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) licence.[3]

Metadata summaries are shortened and paraphrased; the identified summaries remain available under CC BY 4.0 without additional DemandRange restrictions.[4]

The field dictionary and correction exercise are original. No endorsement or regulatory interpretation is implied.

1. River Abstraction Pressures - Dataset - data.gov.ie
2. Environmental Protection Agency Remedial Action List - Dataset - data.gov.ie
3. Open Data Licence - data.gov.ie
4. Deed - Attribution 4.0 International - Creative Commons