Why your launch underperformed while your list grew

A list goes from 4,000 to 11,000 over two years. The next launch makes less than the one before it. Nothing about the offer changed, and the sales page converts at about the same rate.

This is a common shape, and the cause is usually the same. The list grew. The readership did not.

Subscriber count measures every address that ever signed up. Revenue comes from the smaller group still reading. Those two numbers drift apart quietly, and the gap between them is where a launch goes missing.

The number that grew and the number that pays are different numbers

A list is not one audience. It is a stack of cohorts, each arriving at a different time and decaying on its own schedule.

Somewhere in the first month, a share of every cohort stops opening. Addresses go dead as people change jobs. Others never build the habit, because the early sends landed at the wrong moment. What remains is a warm core that keeps reading, sitting inside a total that keeps climbing.

The blended average hides this. One agency audit shows the mechanism cleanly. A brand had been mailing a clean engaged segment of about 30,000 to 35,000 people. Campaigns were then merged into a combined list of more than 85,000, most of the addition inactive.

Click-through on the engaged segment held steady at 1.38%, while the average across all sends fell to 0.72%. Revenue per recipient dropped from $0.14 to $0.05 over 12 months. The engaged readers had not changed at all. The reporting had.

That is the same arithmetic behind a disappointing launch. If you plan against the total, you are forecasting from a denominator full of people who left.

The common launch heuristics make this worse. One to three percent of the list buys. A dollar per subscriber per month. Both treat every address as equivalent. That was not true even then, and it gets less true every month a list grows.

Your open rate is telling you a story machines wrote

Open tracking broke in 2021, and most dashboards still report it as though it did not.

Apple's Mail Privacy Protection preloads the tracking pixel whether or not a person reads the message. Industry measurement now puts machine opens at roughly half of all reported opens. AI summaries add a further layer. A model can render your letter with no person deciding to read it.

So the open rate can rise while readership falls. Owners watch that number climb, conclude the list is healthy, and get surprised at the checkout.

There is a trap on the other side of this too. Sunsetting subscribers by last-open date deletes engaged Apple readers, whose opens the sender cannot see. It keeps disengaged Gmail readers whose pixels fire. Clicks and replies do not have that problem.

For a launch, one number does most of the work. Which numbers to watch month to month has its own answer.

Compute your warm core this afternoon

Your warm core is the count of subscribers who clicked, replied, or bought inside a recent window.

Set the window by how often you send. Weekly senders use 90 days. Monthly senders use 180. Then run two calculations.

First, warm subscribers as a share of the total. Second, revenue from your last launch divided by the warm count rather than the full count.

Suppose a list holds 11,000 people, with 1,900 of them warm. Suppose the last launch produced $38,000. That is $20 per warm subscriber, and $3.45 across the total.

Run the same two numbers against the previous launch. If revenue per warm subscriber held steady while revenue per subscriber fell, the offer worked and the list thinned. That is a different problem from a launch that failed on its merits, and it takes a different fix.

Where the decay comes from

Four causes, and each has its own repair.

Addresses go dead. People change jobs and abandon accounts, and business lists decay faster than consumer ones. Verification and bounce hygiene handle this part.

The habit never formed. A subscriber who misses the first few sends learns to ignore you before deciding anything. The fix is a welcome sequence that arrives promptly and gives something worth the inbox space. It does more for a launch a year later than any subject-line testing.

The source was wrong. Subscribers who arrived for a discount, a giveaway, or a paid cross-promotion often do not remember signing up. Track engagement by where each subscriber came from. A source producing half the average engagement is worth half what its volume suggests.

The letter drifted. Cadence rose, substance fell, or the subject moved away from what people subscribed for. This one shows up as a slow slide across every cohort at once. That pattern distinguishes it from the other three.

What to do before the next launch

Define engagement on clicks and replies. Opens go in the report as a trend line and nothing more.

Build three segments: engaged, at risk, and inactive. The thresholds follow your cadence, and the middle group is the one worth working.

Run one re-engagement sequence to the at-risk and inactive segments, then suppress the people who do not respond. Suppression is not deletion, and you can hold the addresses without mailing them.

Mail the warm core more often and the cold segment rarely. Mailbox providers score you on how the people you mail behave. Dead weight in every send costs you placement with the readers who wanted you.

Then plan the launch against the warm number. The total belongs on no forecast.

Two limits on all of this. A warm core does not rescue an offer the list did not want. No segmentation fixes a price the audience was not prepared for. And pruning will shrink the number on your dashboard. That is the point, though it rarely feels that way the day you do it.

The list you can reach is the list you have

A subscriber count is a record of past interest. Some of it is current, and the rest is history that still shows up in the total.

The launch already told you which was which. It sold to the people who were still reading. The rest had already gone.