VI
The Economics, Without the Gloss
10 One brand is a rounding error
The portfolio creates the operating advantage. Yield decides the prize.
It is easy to make this model sound richer than it is. So take an illustrative brand and follow the money.
The brand has two million dormant addresses. After validation and suppression, 1.4 million are eligible for Recover. A four-Mini sequence sends 5.6 million emails. Five per cent of eligible readers take a verified action and graduate: 70,000 readers. Each receives about 13 Digests a month, and a person engages with 35% of those editions, which gives about 318,500 engaged editions a month.
Now the yield. Suppose each engaged edition offers one ActionAd, 1.5% of them are completed, and the average advertiser payout is $0.40 per action. That is $6 of net revenue per 1,000 engaged editions: about $1,900 a month in total, and about $950 a month to each party. At the operator’s own infrastructure cost of around $0.20 per thousand emails, sending is cheap: roughly $1,100 for the whole Recover sequence and under $200 a month for the Digests.
Delivery is cheap. Audience creation is not. Editorial production, card design, data work, brand onboarding, advertising sales, support and measurement all cost more than the sending, and they come before the revenue. The economic test is recovering the whole investment, not the cost of the emails. As a standalone media property, one brand’s dormant half is a rounding error.
Scale changes the picture, but less than intuition suggests. A hundred such brands would hold about seven million graduated readers and produce about $1.15 million a year in advertising revenue for the operator at the same yield. That is a business line, not a company. If net yield reaches $30 per 1,000 engaged editions, which lead-generation categories may support, the same portfolio produces about $5.7 million a year. Yield, not database size, decides which of these futures arrives.
Figure 5. Illustrative. Per brand: 70,000 graduated readers, 13 editions a month, 35% engaged; operator keeps 50% of net ActionAd revenue.
There is a second number in the illustration, and it matters more to the brand. If one in ten of the 70,000 graduated readers buys within ninety days, at an average order of $30, those customers spend $210,000 in a quarter. Not all of that is incremental; the holdout decides how much. But even a modest incremental share is worth more to the brand than a year of its advertising share. That is why the pitch leads with recovered customers, and why the transaction meter may matter more than the attention meter in many categories.
If the hundredth brand costs as much to operate as the first, there is no factory. That is the test the portfolio must pass, and the operator’s equation shows where. It is simple to write. Contribution equals half of net advertising revenue, plus any transaction share, minus delivery, content, operations and the cost of Mu, with the initial recovery investment repaid over the life of a graduated reader. Four numbers show whether it works.
The first is contribution per 1,000 graduated readers per month. It strips out the size of the database and shows whether a restored reader pays for being kept.
The second is recovery cost per graduate: what it costs, including content and sending, to turn one dormant address into one verified reader.
The third is payback: how many months of contribution it takes to recover that cost. A twelve-month operating term works only if payback falls well inside it.
The fourth, and the one that tests the Bending Spoons thesis directly, is marginal contribution per additional brand. If the factory works, each new brand in a category should cost less to onboard, need less new editorial work and reach advertisers who already buy elsewhere in the portfolio. The tenth travel brand should be cheaper to operate than the second.
These numbers also show where the levers are. Recovery rate determines how many readers the factory gets. Habit Rate determines how long it keeps them. Yield determines what each engaged edition earns. Shared production determines what each edition costs. I set out a similar four-lever equation in How Publishers Can 10X Their Revenue; the same logic applies here, with the dormant half as the starting audience.
The choice of brands follows from the levers. The age of dormancy matters less than four other questions: whether the brand holds legitimate permission, whether its dormant customers are plausibly recoverable, whether advertisers want to reach its category, and whether its world naturally produces a recurring Digest and collectible sets. Jobs, travel, lifestyle, entertainment and some retail categories pass all four. Banking, where third-party advertising and reward currencies face regulatory limits, does not, and needs a compliance-first version of the Digest.
