Back to all articlesRenewals and retention

Keep Missing Account Health Data Separate From Genuine Customer Risk

Alactic Editorial10 min read

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

An account gets marked unhealthy because its renewal plan is blank. Another stays green because nobody has recorded a complaint. Both ratings confuse the condition of the customer relationship with the condition of the evidence.

Keep three signal states separate: known positive, known negative, and unknown. Then record confidence and freshness alongside each signal. Missing information should create a verification task, not automatically a risk penalty. Equally, an absence of recorded problems should not earn a positive rating unless someone has actually checked the relevant outcome.

The practical goal is not to eliminate uncertainty before making a retention decision. It is to identify which uncertainty could change the decision, resolve that uncertainty economically, and avoid presenting guesses as customer facts.

1. Separate the customer condition from your knowledge of it

A useful health review answers two different questions: what evidence suggests the customer will continue receiving value, and how well do we understand the account? Combining those questions into one colour hides important distinctions.

  • Known positive: Relevant evidence supports a favourable condition. For example, a customer has confirmed that a named operational goal was achieved.
  • Known negative: Relevant evidence supports an adverse condition. For example, the customer has said that an unresolved blocker prevents an essential workflow.
  • Unknown: The available evidence cannot establish the condition. A blank field, an unverified inference, or conflicting accounts may all belong here.

These definitions describe a recommended operating method, not a prediction of renewal. A positive signal does not prove the account will renew. A negative signal does not prove it will leave.

Also distinguish a fact from its interpretation. A confirmed departure by an executive sponsor is a fact. Whether that departure threatens renewal depends on replacement ownership, buying authority, and the strength of other relationships. Record the departure without pretending that its consequences are already known.

At the account level, use a short evidence summary rather than forcing a verdict. “Positive operational evidence; commercial intent unverified” is more useful than green when the renewal decision has not been discussed.

2. Define the evidence needed for the decision

Start with the retention decision you need to make: whether to open a recovery plan, seek a renewal conversation, involve an executive, or continue routine account coverage. Do not collect information merely to fill every field.

For each relevant signal, state what would count as evidence. “Engagement” is too vague. “The person accountable for the customer’s target outcome has confirmed progress and the next milestone” is something a reviewer can check.

Use a compact worksheet with these fields:

  • Signal and decision relevance: What condition are you checking, and what action could it change?
  • Evidence: What was observed, who supplied it, and where can the owner verify it?
  • Status: Known positive, known negative, or unknown.
  • Confidence: How directly and reliably does the evidence support that status?
  • Freshness: When was the condition observed, and has anything material changed since?
  • Missing context: What prevents a stronger conclusion?
  • Next verification action: Who will check what, by when, and through which route?

Keep confidence separate from status. You can be highly confident that commercial intent is unknown because nobody has asked the authorised decision-maker. That does not make the account highly risky. It makes the information gap clear.

Freshness should follow the signal, not a universal expiry rule. A past outcome can remain valid historical evidence while no longer describing current priorities. A recent service complaint may need immediate reassessment after a fix. Record both the original evidence and any change that could undermine its present relevance.

3. Completed fictional health evidence worksheet

Fictional worked example: Every account detail, date, quantity, role, and deadline below is hypothetical. These are not Alactic customer results.

Harbour Supply has a renewal decision approaching. Its account owner must decide whether to begin a recovery plan or first clarify a few unresolved conditions. The review takes place on 6 May, with renewal scheduled for 30 June. The following worksheet is completed before any outreach is authorised.

Signal: achievement of the primary operating outcome

  • Decision relevance: Establish whether there is current evidence of customer value.
  • Evidence: On 2 May, the operations lead confirmed that the agreed order-handover process was working for the main team. The account owner retained the meeting note.
  • Status: Known positive, limited to that team and workflow.
  • Confidence: High for the lead’s stated experience; not evidence of company-wide value.
  • Freshness: Recent at the review date; no subsequent change is recorded.
  • Missing context: Whether the finance team recognises the same outcome as sufficient value for renewal.
  • Next verification action: The account owner will check the agreed success criteria internally on 7 May, then include a value-confirmation question in the planned renewal discussion.

Signal: unresolved workflow blocker

  • Decision relevance: Determine whether recovery work is needed before the renewal discussion.
  • Evidence: A support record dated 5 May says the branch team still cannot complete an essential handover step.
  • Status: Known negative for that workflow.
  • Confidence: High that the issue was reported; medium that the recorded scope is complete.
  • Freshness: Recent, but the support owner may have newer information.
  • Missing context: Whether a workaround is usable and how many teams are affected.
  • Next verification action: The support owner will confirm the latest state internally on 6 May. If impact remains unclear, the account owner will ask the affected customer contact to validate the scope.

Signal: executive sponsorship

  • Decision relevance: Decide whether relationship rebuilding requires senior involvement.
  • Evidence: The former sponsor’s departure is recorded in an account note dated 18 April. No replacement is identified.
  • Status: Known negative for continuity of the previous sponsorship; unknown for replacement coverage.
  • Confidence: High for the recorded departure; low for any claim that the account now lacks sponsorship.
  • Freshness: The departure remains relevant, but replacement information may have changed.
  • Missing context: Who now owns the business outcome and who can approve continuation.
  • Next verification action: The account owner will check recent internal correspondence on 7 May, then ask the existing customer contact for an introduction if ownership remains unclear.

Signal: commercial intention

  • Decision relevance: Decide whether a renewal discussion needs earlier attention.
  • Evidence: The renewal-intent field is blank. No direct statement from a decision-maker appears in the reviewed notes.
  • Status: Unknown.
  • Confidence: High that the reviewed record does not establish intent; no basis for confidence about the customer’s intention.
  • Freshness: Not applicable to a missing observation. The record was checked on 6 May.
  • Missing context: Decision ownership, current priorities, and the customer’s decision process.
  • Next verification action: The account owner will arrange a focused renewal conversation by 10 May, asking what must be true for continuation rather than assuming reluctance.

Signal: absence of complaints

  • Decision relevance: Test whether silence provides any usable evidence of satisfaction.
  • Evidence: The account summary says no complaints, but the open support issue contradicts that broad description.
  • Status: Unknown for overall satisfaction. The summary is not positive evidence.
  • Confidence: Low in the summary’s completeness.
  • Freshness: The summary has no observation date.
  • Missing context: Which channels and contacts were considered when it was written.
  • Next verification action: The account owner will correct the summary on 6 May and link the known blocker. No separate satisfaction message is needed before the planned conversation.

Working decision: Begin scoped recovery work on the known blocker while verifying commercial ownership and intent. Do not downgrade the account merely for the blank renewal field, and do not treat operational success as proof of commercial commitment.

4. Choose internal investigation, customer contact, or deferral

Not every unknown deserves a customer message. First ask whether resolving it could change an action before the next decision point.

Investigate internally when the answer should already exist: the latest support state, a prior meeting note, an agreed success criterion, or a colleague’s recent conversation. Give that investigation an owner and a stopping point. Otherwise, “checking internally” can become an indefinite delay.

Contact the customer when the missing information concerns their present intent, priorities, experience, or decision ownership and cannot be established reliably from existing evidence. Explain why the question matters. Instead of asking a broad “Is everything okay?”, ask which unresolved condition could prevent continuation or who should participate in the decision.

Defer deliberately when the gap is unlikely to change the current action. Record the trigger for revisiting it, such as a change in scope or the next agreed review. This preserves attention for consequential uncertainty without quietly treating deferred questions as answered.

Customer contact also has a cost: repeated questions can make an account team appear disconnected. Combine related verification needs into an existing conversation where practical. But do not delay an active, material problem simply to consolidate outreach.

5. Keep confidence and freshness from becoming hidden scores

Confidence labels are useful only when reviewers explain them. “High confidence” should refer to a specific claim supported by evidence, not the account owner’s general feeling about the relationship.

A direct statement may strongly establish what one person believes while providing weak evidence about the organisation’s final decision. An activity record may show that a workflow occurred without establishing whether it produced value. Preserve those boundaries in the worksheet.

When evidence conflicts, do not average it into a reassuring middle position. Record the competing claims and the question that would distinguish them. A satisfied operations contact and a dissatisfied budget owner may both be accurately describing their experience.

Likewise, do not turn old positive evidence automatically into negative evidence. Mark it as historical and reassess whether it still supports the present decision. If it does not, current status may be unknown. The deterioration is in your knowledge, not necessarily in the relationship.

6. Recover when the existing rating is misleading

If a review reveals that an account’s rating rests on blanks or unsupported assumptions, correct the interpretation without erasing the trail.

  1. Preserve the evidence behind the previous rating, including where none was recorded.
  2. Identify which conclusions were unsupported, stale, or too broad.
  3. Replace those conclusions with bounded signal statements and explicit unknowns.
  4. Reconsider actions triggered by the old rating, such as escalation or reduced coverage.
  5. Tell the people using the rating what changed and what remains unresolved.

Two shortcuts are especially damaging. Penalising every unknown encourages people to complete fields with guesses. Rewarding every silent account encourages teams to overlook relationships they understand least. Both can improve apparent record completeness while weakening retention decisions.

Another failure is allowing verification to displace action. The fictional worksheet contains a known blocker as well as unanswered questions. Work on the blocker can begin while ownership and intention are clarified. Uncertainty in one area is not a reason to ignore evidence in another.

7. Run the next review around decisions, not completeness

For the next review, choose accounts with an upcoming retention decision or a rating that lacks supporting evidence. For each account, write a short evidence summary, list the unknowns that could change the next action, and assign verification owners.

Use the following checks at the subsequent review:

  • How many decision-relevant unknowns were resolved with verifiable evidence?
  • How many verification tasks remain overdue or lack an owner?
  • Which findings changed an account action, and which confirmed the existing approach?
  • Which customer questions duplicated information already available internally?
  • Which positive or negative conclusions were corrected because their scope was overstated?

Do not make the percentage of positive signals the success measure. Discovering a real problem can make the worksheet look worse while making the retention decision better.

The immediate deliverable is simple: every consequential unknown gets an owner, a next action, and a review point—or an explicit reason to defer. That leaves the team with a defensible distinction between a customer who is struggling, a customer with demonstrated positive evidence, and a customer whose condition has not yet been established.

Account healthCustomer retentionRenewalsEvidence quality

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