Delete subscribers. Make more money.
Of all the flows I recommend in audits, one gets resistance every single time: the sunset flow — the automation that identifies subscribers who’ve stopped engaging, gives them a last honest chance to stay, and then suppresses them so you stop emailing them entirely.
The resistance is understandable. You paid to acquire those subscribers. Some popup, some ad spend, some discount bought each of those addresses, and suppressing them feels like setting the receipts on fire. Every founder does the same arithmetic out loud: “even at a terrible conversion rate, 20,000 extra addresses must be worth something.”
They’re worth less than something. Past a point, they’re worth negative — and the mechanism is invisible until you understand how mailbox providers decide where your email lands. So let me make the case for the flow nobody wants, with the math, the metaphor, and the exact segment logic.
Your list is a fruit tree, not a warehouse
The mental model that gets this wrong treats an email list like a warehouse: every address is inventory, inventory is an asset, and more inventory is more potential sales. Warehouses don’t punish you for holding dead stock.
An email list behaves like a fruit tree. Every branch — every subscriber — draws from the same finite root system: your sender reputation. Living branches turn that supply into fruit. Dead branches don’t just yield nothing; they keep drawing supply, and the tree allocates water and light away from the branches that could actually produce. Any gardener will tell you the counterintuitive truth that took me years of audits to internalise about email: pruning is how you get more fruit, not less tree.
Here’s the mechanism without the bark. Gmail, Outlook and Yahoo decide inbox-versus-spam largely on engagement rates: what fraction of your recent sends get opened, clicked, replied to — versus deleted unread or marked as spam. That’s a ratio, and every disengaged address you keep mailing sits in the denominator. Send to 30,000 people of whom 9,000 ever engage, and you’re presenting Gmail a 30% engagement story. Send to the 9,000, and the same humans, the same behaviour, reads as a 100% engaged sender. The providers respond by inboxing more of your mail — including to people in the middle of the engagement spectrum who were drifting away. The engaged branches get more light.
And it compounds in the other direction too. Old, abandoned inboxes get recycled by providers into spam traps — addresses whose only job is to catch senders who don’t clean their lists. Dead branches attract exactly the pests you’d expect.
The math founders actually respond to
Take a concrete middle-of-the-road account: 30,000 profiles, 9,000 genuinely engaged, campaigns twice a week. The unpruned version sends 60,000 emails a week to tell a 30% story. Suppose degraded inbox placement costs you even 15% of the opens you’d otherwise get from your engaged segment — a conservative figure for a reputation-damaged sender. That’s your best customers not seeing your best sends. On a list where email drives, say, €40k/month, a 15% haircut on the engaged core is €4–6k/month of invisible loss — invisible because it never appears as a line item, only as “email feels less effective than it used to.”
Against that, what does the dead weight contribute? Run the report: revenue attributed to profiles with no open, click, or site visit in the last 180 days. In every account I’ve audited, it rounds to approximately nothing. You’re holding an asset that yields ~0 and charges you 10–15% interest on your productive assets. Klaviyo also bills by active profile count, so you’re paying storage on the dead stock too — real money, though it’s the smallest of the three costs.
The exact flow
The segment (the pruning line). Profiles where all of the following are true over the window: received at least 8–10 emails (so silence is a verdict, not small sample size), opened zero, clicked zero, placed zero orders, and no active-on-site event. Window: 120–180 days for most DTC purchase cycles; stretch it for genuinely seasonal or long-replenishment products. Two guards that matter: exclude profiles created in the last 60 days (new subscribers deserve the welcome series’ full run before judgment), and if Apple’s privacy opens make your open data mushy, weight the definition toward clicks and site activity — the signals that can’t be machine-generated on your behalf.
The last-chance sequence. Two, at most three emails, spaced about a week apart. Plain, honest, un-designed: “We haven’t seen you in a while — want to keep hearing from us?” One clear yes-link. No parade of discounts; a subscriber who only pulses back to life for 20% off is a discount alarm clock, not a retained customer. Any click or open exits them from the flow back into normal sending.
The cut. No response after the sequence: suppress. Suppression, not deletion — the history stays, and if they ever purchase again they can be un-suppressed with their profile intact. The branch is pruned, not erased from the tree’s memory.
One operational note from a build I did on a WooCommerce store: make suppression an automated end-step of the flow, not a monthly manual export-and-suppress ritual. Manual pruning schedules have a way of becoming quarterly, then annual, then archaeological.
When not to run it
Honesty requires the disqualifiers. Skip or soften the sunset flow if: your list is under a few thousand (the reputation math barely applies and every address is knowable); your sending has been so infrequent that “disengaged” mostly measures your absence rather than their disinterest — fix cadence first, then judge; or you’re about to run a re-permission campaign for compliance reasons anyway, which supersedes it. And never sunset based on a window shorter than your natural purchase cycle; suppressing someone mid-consideration is genuine money-burning, the thing founders feared all along.
The take
Every account I’ve worked on that resisted the sunset flow was optimising the visible number (list size) at the expense of the number that pays (revenue per send, engaged reach). Prune on a schedule, automatically, with an honest last chance. The tree gets smaller. The harvest gets bigger. That trade is the whole job of retention.