Revenue per recipient, why last-click attribution robs email of the credit, the holdout test that gives the honest answer, and the cohort view that shows you exactly when subscribers go quiet.
Cyril Nie
Author
Checking your open rate and moving on isn’t analytics. That’s reading a scoreboard without understanding the game. Real measurement connects email to money.
The single most useful money metric. An email to 10,000 people that makes $5,000 earns $0.50 per person. An email to 50,000 people that makes $8,000 earns only $0.16. The smaller send was three times more efficient. Also watch revenue per email sent over time — if it keeps falling while you send more, you’re tiring your list out. Some brands cut sends by 30% and made the same money.
Your email platform will claim more revenue than Google Analytics does. Both are half-right. Platforms use generous windows (opened in the last 5 days? credit to email!). Analytics gives credit to the last click, so email that started the sale gets nothing. The truth sits in between. Use platform numbers to compare your own emails against each other, and analytics numbers to compare email against other channels.
How you give out credit decides who wins the argument. In plain words:
Last click gives all the credit to the final step. This is the standard almost everywhere, and it robs email every time. If someone reads your email, then Googles your name to buy, Google gets the sale.
First click gives all the credit to the first step. It shows you what got attention, and ignores everything that helped someone decide.
Even split shares the credit across every step. It sounds fair, but it treats a random ad the same as the email that closed the deal.
First-and-last gives 40% to the first step, 40% to the last, and shares 20% in between. A sensible choice for most businesses.
Recent-weighted gives more credit to whatever happened close to the sale. Best when your sales take months.
Computer-worked-out figures out the real weights from your own data — but it needs a few hundred sales a month to mean anything.
With last click, the email that did the convincing scores zero and search takes the sale.
One more thing worth knowing. People who hear from you by email and on social and on your site buy about 50% more than people who only see you in one place. Email is usually the thread tying those visits together. Last click almost never shows it.
None of this works if your links are untagged. Add a tag to every link in every email: where it came from (your email tool), what it is (email), the campaign name, and which link it was. That last one tells you whether the big button or the footer link does the work. Then pick one way of naming things and write it down. “welcome-series” in one email and “Welcome_Series” in another show up as two campaigns, and your numbers quietly stop making sense.
Want the real answer? Randomly leave 5–10% of a segment out of a campaign. After 7 days, compare buyers in the emailed group vs. the left-out group. The gap is email’s true effect. Expect two findings: email really does work — and your dashboard was exaggerating. For cart reminders, 30–50% of “recovered” carts would have come back anyway. Both facts help you invest smarter. Run one of these monthly or quarterly.
Simple math: average revenue per subscriber per month × average months they stick around. If a subscriber brings $2.50 a month and stays 14 months, they’re worth $35. Now you know what you can pay to get one. Track this by source — organic subscribers might be worth $42 while paid-social ones are worth $18. Businesses that check this often move 40% of their ad budget afterward. Aim to get at least $3 of value for every $1 spent acquiring. Below that you’re overpaying; above $5 you’re probably under-investing in growth.
Count what you keep, not what you add: (new signups − unsubscribes − bounces − complaints) ÷ your whole list, every month. This matters because lists shrink 22–25% a year all by themselves. People change addresses, change jobs, lose interest. You need about 2% new every month just to stay level. Healthy growth is 10–20% a month under 5,000 people, 5–10% up to 25,000, and 1–3% once you pass 100,000.
Track the people leaving too, and track where they came from. People who signed up through paid ads usually leave 2–3 times faster than people who found you on their own. If more than 1% leave each month, something is wrong: your content, how often you send, or the promise you made at signup. Also watch how much of your list clicked or opened in the last 90 days. Under 30% means you are paying to store people who aren’t listening.
Averages hide the trend. Instead, group people by the month they joined. Then watch each group’s clicks over their first six months. Two things show up. First, whether you are getting better or worse. If people who joined in January clicked twice as much in their first month as people who joined in June, something changed — where your traffic comes from, your welcome emails, or your content. Second, your drop-off point — when people usually go quiet. For most programs it lands between month two and month four. Once you know that month, you can send your win-back emails just before it, instead of long after everyone has gone.
Two things to see here. The later group starts lower, so something changed. And both groups drop off a cliff around month three. Send your win-back emails before that.
Track revenue per recipient per send, then compare against platform and list costs. For a true figure, run a holdout: withhold email from a random slice of your list and compare their revenue to everyone else’s. The gap is what email actually caused.
Excluding a random 5 to 10 percent of your list from a campaign or flow, then comparing their revenue with the treated group. It is the only method that measures incremental impact rather than reassigning credit between channels.
Because last-click attribution assigns the sale to whatever the customer touched last, which is often a search for your brand name prompted by your email. Email also lifts sales among people who never open it, and no click-based model can see that at all.
Discover more insights on similar topics
The complete 2026 guide to email marketing: 16 chapters on list growth, segmentation, automation, deliverability, testing, attribution and AI. Built from 908 sources and 4,798 insights.
Email returns about $36 for every $1 spent — roughly 30 times the next best channel. Here is why it keeps winning, and the six parts every working program is built from.
Apple inflates open rates and Gmail’s scanners click your links. Here are the email metrics that survive scrutiny, the thresholds worth holding yourself to, and why your own last 90 days beat any industry average.