Back to all articlesMessaging permission

Review a Historical Contact List Before Treating It as Messageable

Alactic Editorial11 min read

Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.

A historical list arrives with phone numbers, source labels and a column called “permission.” Some rows say yes; others are blank. The team wants one decision so work can move forward. The useful answer is not to approve or reject the whole file: review each record against evidence for the intended messaging channel and purpose, separate withdrawals from missing evidence, and hold records whose status cannot be supported.

A sample can reveal recurring problems, such as a missing form version or an import that replaced blanks with yes. It cannot establish permission for records nobody reviewed. The workbook below shows how to make that distinction operational, with evidence categories, named ownership and release-or-hold decisions.

This is an operating method for reviewing an existing list, not a legal determination. A permission-review release also does not establish campaign readiness, template approval or likely delivery.

1. Define what the review is actually deciding

Before opening individual records, write down the proposed use. “Can we message these people?” is too broad. A useful review question identifies the business, channel and purpose: does the available evidence support this business contacting this person through this channel for this stated purpose?

A phone number supplied for a service appointment does not, by itself, establish permission for unrelated promotional messages. An email subscription does not, by itself, establish WhatsApp permission. A previous purchase, old conversation or salesperson’s familiarity with the contact may help locate evidence, but it does not explain the scope of permission on its own.

Use four evidence statuses rather than a single yes/no field:

  • Supported permission: accessible evidence connects the recipient, business, channel and proposed purpose, with no unresolved contradictory evidence found in the review.
  • Purpose-limited permission: evidence supports a narrower use than the one proposed. Record that boundary explicitly.
  • Withdrawal: evidence shows the recipient asked to stop the relevant messaging. Preserve the scope and timing of that request.
  • Unknown: evidence is missing, inaccessible, ambiguous or contradictory enough that the reviewer cannot support a decision.

Keep the evidence status separate from the decision. A purpose-limited record can be supported for its documented purpose while remaining on hold for the proposed broader use. Unknown is not a synonym for withdrawal, but both can require a hold.

2. Establish the evidence standard before reviewing rows

Agree on what a reviewer must be able to establish. Otherwise, one reviewer will accept a legacy checkbox while another will demand the original submission, making the final totals difficult to interpret.

Use these evidence categories in the workbook:

  • E1 — Direct permission evidence: a preserved submission or message showing the relevant choice, its wording, the recipient identifier and timing.
  • E2 — Supporting context: source labels, transaction records, internal notes or conversation history that help explain a record but do not independently establish the required permission.
  • E3 — Withdrawal evidence: a recipient’s stop request or another preserved record of withdrawal, with the affected channel or purpose where known.
  • E4 — Missing or conflicting evidence: an unexplained imported flag, an unavailable source, contradictory records or an uncertain identity link.

These categories describe evidence, not its strength by filename. A document called “consent export” may contain only unexplained yes values. Conversely, an archived message may contain clear, narrow permission. Review the content and its connection to the contact, not just the label.

For each decision, retain a retrievable reference, the relevant wording or an accurate scope summary, the event date where available, and the reason the evidence belongs to that record. Do not copy entire conversation histories into a broadly shared workbook when a restricted reference and concise summary will do.

A shared form or script can support multiple records only where each record is reliably linked to that version and its recorded response. Finding one good form does not establish which contacts saw it or what they selected.

3. Use sampling to find defects, not confer permission

Start with a discovery sample across source groups and known uncertainties. Include records from different collection periods, forms, events and manual imports, plus records with blank fields or conflicting flags. This is a diagnostic exercise, not an estimate of how much of the whole list is safe to use.

For each sampled record, ask two questions: what does the evidence support for this person, and what recurring problem might affect other records?

If an event record lacks the original collection wording, investigate that source group. If a manual import shows yes values but no source references, look for the import mapping and original evidence. If a withdrawal disappeared from the current file, check whether other withdrawals were omitted too.

Document what the sample cannot tell you. A sample containing no withdrawals does not establish that the rest of the list contains none. A high proportion of well-supported sampled records is not a reason to fill every remaining row with supported permission.

There is a speed-versus-certainty trade-off. A source-based sample can expose a systemic defect quickly, allowing a whole affected group to remain on hold while the defect is investigated. Releasing individual records still requires evidence connected to those records. Where that connection cannot be recovered, preserve the hold rather than substituting the sample’s result.

4. Fictional worked example: the filled review workbook

All contacts, dates, references, roles and numbers in this workbook are hypothetical. The example uses a small historical list so every record can be shown. The proposed use is WhatsApp product announcements from the same business named in the original collection material.

Worksheet A — Review control

  • Review ID: HIST-01.
  • List: Legacy contacts extract, frozen for review on 6 May.
  • Population: 12 contact records across web forms, service requests, an event and manual imports.
  • Discovery sample: R01, R04, R07 and R09, selected to inspect each source group; not a representative statistical sample.
  • Review owner: Maya, operations lead.
  • Evidence retrieval owner: Theo, records coordinator.
  • Decision checker: Priya, messaging operations reviewer.
  • Control: No list-wide permission release. Decisions must reference reviewed record IDs and a stated purpose.

Worksheet B — Record-level evidence register

Each entry below contains the record ID, evidence reference, finding, status and current decision. Full recipient identifiers remain in the restricted source register rather than this working summary.

  • R01 — Web form; E1, FORM-A/101: archived submission links the recipient to wording offering WhatsApp product announcements from the named business. Status: supported permission. Decision: release from permission hold for that purpose. Reviewer: Maya; checked by Priya.
  • R02 — Web form; E1, FORM-A/102: same verified form version, separately linked affirmative response and recipient identifier. Status: supported permission. Decision: release for product announcements. Reviewer: Maya; checked by Priya.
  • R03 — Web form; E1, FORM-A/103: preserved response supports WhatsApp product announcements; no conflicting evidence found in the reviewed sources. Status: supported permission. Decision: release for that purpose. Reviewer: Maya; checked by Priya.
  • R04 — Service request; E1, SERVICE/204: recipient requested WhatsApp updates about one repair. Status: purpose-limited permission. Decision: hold for product announcements; preserve repair-only scope. Reviewer: Maya.
  • R05 — Service request; E1, SERVICE/205: recipient requested appointment coordination through WhatsApp. Status: purpose-limited permission. Decision: hold for product announcements. Reviewer: Maya.
  • R06 — Service request; E1, SERVICE/206: archived message asks for updates about one order. Status: purpose-limited permission. Decision: hold for product announcements. Reviewer: Maya.
  • R07 — Event list; E2, EVENT/307: attendance record contains a phone number but no preserved messaging choice or collection wording. Status: unknown. Decision: hold; Theo to seek original collection material and response record.
  • R08 — Event list; E2 and E4, EVENT/308: internal note says “interested,” with no channel or purpose stated. Status: unknown. Decision: hold; Theo to locate underlying evidence.
  • R09 — Manual import; E3, THREAD/409: later recipient message requests that WhatsApp announcements stop. Status: withdrawal. Decision: hold for those announcements and retain withdrawal reference. Reviewer: Maya; checked by Priya.
  • R10 — Manual import; E3, THREAD/410: recipient requested no further WhatsApp messages from the business. Status: withdrawal. Decision: hold for WhatsApp messaging within that request’s scope. Reviewer: Maya; checked by Priya.
  • R11 — Manual import; E4, IMPORT/411: yes flag has no source, wording or event date. Status: unknown. Decision: hold; Theo to trace the original record.
  • R12 — Manual import; E4, IMPORT/412: permission evidence is attached to an identifier that cannot be reliably matched to this recipient. Status: unknown. Decision: hold; Theo to resolve the identity link without assuming a match.

Worksheet C — Discovery findings and unresolved work

  • Finding F01: sampled service record R04 supports only service updates. Action: inspect each remaining service record’s actual scope. Completed review found R05 and R06 also purpose-limited.
  • Finding F02: sampled event record R07 lacks collection wording. Action: review R08 separately and request the original event materials. Both records remain unknown.
  • Finding F03: sampled import R09 contains a withdrawal absent from the headline permission flag. Action: inspect the remaining imported records individually. Review identified another withdrawal and two unknowns.
  • Unresolved queue: R07, R08, R11 and R12. Owner: Theo. Next checkpoint: 9 May. If evidence is not found, retain unknown status and hold; the checkpoint is not an automatic release date.

Worksheet D — Release-or-hold decision log

  • D01, 6 May: hold the entire extract pending review. Owner: Maya. Reason: legacy yes values have inconsistent provenance.
  • D02, 7 May: release R01–R03 from the permission-review hold for WhatsApp product announcements only. Approved by Priya against the referenced evidence.
  • D03, 7 May: hold R04–R06 for announcements because evidence supports narrower service purposes. Retain those boundaries rather than overwriting them as no permission.
  • D04, 7 May: hold R09–R10 in line with their withdrawal scopes. Preserve the evidence even if the historical permission flags say yes.
  • D05, 7 May: hold R07, R08, R11 and R12 as unknown. Reopen only when new evidence is linked and reviewed.
  • Reconciliation: 12 records reviewed; 3 supported, 3 purpose-limited, 2 withdrawn and 4 unknown. For the proposed announcements: 3 permission-review releases and 9 holds.

5. Make a hold useful, not merely cautious

A defensible hold should tell the next reviewer what is missing and what would change the decision. “Needs checking” is not enough. “Original event collection wording and this recipient’s recorded response are unavailable” identifies a specific gap.

Assign a retrieval owner separately from the decision owner. The person finding an old form should not silently turn its existence into permission for everyone in the file. The decision owner must verify the record-to-evidence link and scope.

Set a checkpoint to prevent unresolved work from disappearing, but do not make elapsed time a release condition. If evidence cannot be recovered, close the retrieval task with the status still unknown and the hold intact. Avoid contacting an uncertain list merely to repair its permission history as part of this review.

6. Recover when a review assumption fails

Suppose the team later discovers that a form archive was labelled with the wrong version. Pause the decisions that relied on that version, identify affected record IDs and reopen them. Do not invalidate unrelated evidence automatically, but do not preserve releases solely because they were already approved.

Keep the earlier decision visible in the log, add the new finding and record the replacement decision. This makes the correction understandable without disguising the original mistake.

Other failure cases need similarly specific responses:

  • Conflicting permission and withdrawal: retain both records and honour the applicable withdrawal rather than allowing an older yes flag to override it.
  • Uncertain recipient identity: hold the affected record until the evidence can be reliably linked; do not borrow permission from a similar name.
  • Pressure for a batch approval: report supported records and held records separately. A deadline does not fill an evidence gap.
  • A cleaner overwrites unknown with no: restore the distinction. Missing evidence and an explicit refusal require different explanations and follow-up handling.

7. Hand off a bounded decision and measure completion

The handoff should contain record IDs, evidence references, approved scope, hold reasons, decision dates and owners—not just a file named “approved contacts.” Any subsequent use must remain within the recorded permission boundary.

For WhatsApp, recipient permission and opt-outs remain relevant independently of template handling. Approved templates are used to initiate Platform conversations and outside the customer-service window, which opens or resets with a user message. Template approval is not a delivery guarantee. Check the current WhatsApp Business Messaging Policy for precise requirements; a permission workbook does not replace that check.

Measure the review by observable work: records with retrievable evidence, unresolved identity links, outstanding retrieval tasks, documented withdrawals and releases with a checked scope. Do not turn a target release percentage into an incentive to reinterpret uncertain evidence.

The next action is concrete: freeze the review copy, name the proposed channel and purpose, assign a review owner, and complete a discovery sample across the source groups. Then reconcile the population so every record has either a supported, bounded decision or an explicit hold. Completion means no unaccounted-for records—not that the whole list becomes messageable.

Messaging permissionContact listsEvidence reviewOperations

Put the next conversation to work.

Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.

Explore Communications