Proof · Audience rebuild · 12 weeks
The meal-kit brand that was mailing two lists: one real, one imaginary
By Maren Holt, Founder & Principal Strategist · Published May 11, 2026 · Updated July 26, 2026 · Client anonymized; figures representative of the engagement
A meal-kit subscription brand came to us with 32,000 profiles, two campaigns a week, and revenue that had stopped responding to volume. Sending more did nothing. The audit explained why in one number: 64% of their sends were going to profiles with no open or click in 120 or more days. On paper they had a 32k list. In practice, about a third of it was real.
The problem: volume addiction with a hidden tax
Batch-and-blast feels free because ESPs price by profile count, not by damage. It is not free. Gmail reads sustained sends to silent recipients as a signal that mail is unwanted, and this brand's complaint rate had crept to 0.31%, three times the 0.1% ceiling Google recommends for bulk senders and closing on the 0.3% enforcement line. Unsubscribes ran at 0.48% per campaign. Every send was spending future deliverability to buy present revenue that mostly did not arrive.
The approach: score, suppress, rebuild
Weeks one and two produced the audience map: engagement scores for every profile built on click and site-session recency, since Apple's Mail Privacy Protection has made opens unreliable on their Apple-heavy audience. RFM tiers split the buyers from the browsers. Then came the uncomfortable step, and the one clients resist: we suppressed the silent 64% from regular campaigns entirely.
Send volume dropped 34% in three weeks. Campaigns were rebuilt for the remaining tiers: VIPs and steady repeaters got earlier access and no discounts, one-time buyers got a reorder-timed sequence keyed to box cadence, and recent-but-quiet profiles got a three-email re-permission series. Anyone who stayed silent after that left the active list under a written sunset policy.
The result: the same revenue from two-thirds the mail
| Metric | Before | After 12 weeks |
|---|---|---|
| Send volume | baseline | −34% |
| Campaign click rate | 1.3% | 2.2% (Klaviyo avg: 1.69%) |
| Spam-complaint rate | 0.31% | 0.07% |
| Unsubscribe rate | 0.48% | 0.19% |
| Flow share of email revenue | 19% | 33% (platform avg ~41%) |
Revenue held flat through the entire suppression phase, which is the finding worth sitting with: a third of their sending had been producing nothing except complaints. Click rate now sits above Klaviyo's 1.69% cross-account average, and the flow share of email revenue is climbing toward the platform's ~41% norm as the rebuilt automations mature.
The engagement used the same sequence we run for every client: the audit from our Audience Intelligence service first, then the campaign and flow rebuild. If your list has grown faster than your revenue, the segmentation strategy guide shows the scoring model in more detail.