Home

Halden & Rowe · Customer · Loyalty

Lapsed silver-tier members respond 4.2× to a single re-engagement mailer.

92k eligible members; rolling out April's control cell to the full cohort is a low-risk £210k.

Ana · Digital Retail AnalystPrepared Mon 29 Jun · 08:02Decision by Fri 10 JulEstablished

Reading

What Ana sees this morning.

What happened

April's control cell to 8,000 lapsed silver members lifted 60-day repeat purchase from 3.1% to 13.1% with no observed cannibalisation of gold-tier spend.

Why it matters

The cohort has 9–15 months of purchase history and a clear response signal to lifecycle triggers. The mailer template is already approved and CRM capacity is available in week 29.

Evidence

What Ana is reading.

The signals behind the prompt. Every line is drawn from the trading data available to Ana this morning.

  • Control cell size

    8,000 members

  • Repeat-purchase lift

    3.1% → 13.1%

    60-day window

  • Gold-tier cannibalisation

    none observed

  • Eligible population

    92,000

  • CRM capacity

    available week 29

Estimated impact · next 90 days

£210Kincremental gross profit

How Ana sized it · Apply the observed lift to the eligible 92k population at April's average order value, discounted 30% for mailer fatigue at scale.

Recommended action

What Ana would do.

One clear next step, and the alternatives Ana considered before landing on it.

Schedule the mailer for week 29 across the full 92k eligible cohort with a 10% holdout for measurement.

Why · The April cell already carried the risk. Scaling with a holdout preserves measurement without leaving cash on the table.

Alternatives Ana considered

  1. 1.Stage the roll-out in two waves of 46k with a two-week gap to re-measure fatigue.
  2. 2.Move budget into a paid social lookalike instead; slower to measure and higher CAC on this cohort.

Decision

Close the loop.

Accepting records the intent and starts measurement against the estimated impact. Dismissing files the reasoning so the same prompt does not re-surface without new evidence.

Close the loop

Accept to record the intent and begin measurement. Dismiss to file the reasoning. Assign to route the decision to someone else on the team.

Context

Metrics in play.

  • Customer churn
  • Average order value