A subject line that says "3 things I learned" tells the reader exactly what they're getting. A subject line that says "something I learned the hard way" tells them nothing—except that opening it might be worth their time. That gap between known and unknown is where the interesting psychology lives. So the question worth asking: if we strip the count out of an email's promise, does the reward actually feel bigger, or does it just feel vaguer? And if it feels bigger, by how much?

The itch that a number scratches

Behavioral research has a useful word for the state a vague promise creates: anticipatory uncertainty. When you don't know what's inside, your brain allocates attention to finding out. When you do know—"5 tips," "3 templates"—the reward is pre-priced. You can decide in advance whether it's worth your click, and often you decide it isn't.

This is close cousin to what B.F. Skinner found with variable-ratio reinforcement: unpredictable rewards produce more persistent engagement than predictable ones. Email inboxes are, whether we like it or not, a reinforcement environment. The sender who reveals the exact payload is handing the reader a calculator. The sender who withholds it is handing them a question.

Where the 22% figure comes from

The "22% harder" framing echoes a pattern that shows up repeatedly in split-test write-ups and in academic work on curiosity gaps (Loewenstein's information-gap theory is the classic reference—people feel a deprivation when they sense missing information, and that deprivation motivates the click). In practice, when marketers A/B test counted versus uncounted promise lines, uncounted versions frequently win on open rate by double-digit percentages, with ~20% being a common landing zone. Treat the exact number as directional, not gospel—your list, your sender reputation, your audience's mood that Tuesday all move it.

But the direction is consistent enough to plan around: the count is a spoiler.

The catch: vagueness has a floor

Here's where it gets more interesting, and where a lot of well-meaning advice goes wrong. Uncounted promises work because they create a curiosity gap. They backfire when the gap looks like a void.

"Big news inside" creates no gap—it creates suspicion. The reader has been burned by that line before. The reward has to feel nameable but unnamed: "the one onboarding email I'd cut first" works because the reader can imagine the category without knowing the answer. "You won't believe this" fails because it promises a reaction rather than information.

Kahneman's loss aversion is relevant here too. A counted email says "here are 3 things." An uncounted one says "here's a thing you might be missing." The second frames non-opening as a small loss—a piece of information you'll continue not to have. Losses loom larger than equivalent gains, and that asymmetry is doing quiet work in the subject line.

What to test next

Pick two emails from your next send. On one, keep the count. On the other, rewrite the promise so it names a category but hides the payload—"the mistake that cost us a quarter's pipeline" instead of "4 pipeline mistakes." Send both to slices of the same segment, same day, and watch opens and click-to-open. If opens rise but clicks fall, you've learned something more useful than a headline number: you've learned whether the curiosity gap is being paid off inside the email or just exploited at the door.

Then run it again next month. Reward schedules that feel fresh in March can feel like a formula by June—which is itself the most interesting thing about variable rewards. They only stay variable if you keep varying them.