Ten hooks in ten hours: testing hooks against a live audience

A hook is the first line: the one sentence on an ad tile, the subject line, the headline above the fold, the opening of a post. It is the only copy most people ever read, and testing hooks is the fastest test in marketing because a hook is one line long. This month's "vibe marketing" round-ups describe founders turning an idea into ten variants before lunch (mean.ceo has the summary). Fine. The variants are the easy part. The hard part is putting them in front of strangers and reading the reaction without lying to yourself about the sample. This is a way to do ten in ten hours that survives that.

Why ten, and which ten

Ten is not a magic number; it is roughly the count of distinct hook formats, and one of each stops you writing the same line nine times with different adjectives. The formats, with an example for an invoicing tool:

1. The outcome. "Get paid nine days sooner."
2. The enemy. "Stop chasing invoices."
3. The number. "Invoices out in 40 seconds."
4. The question. "Still emailing PDFs?"
5. The who-it-is-for. "Invoicing for freelance designers."
6. The contrast. "Invoicing without the accounting software."
7. The time. "Set up before your coffee cools."
8. The price. "Free until your tenth invoice."
9. The proof. "Used by 300 studios." — only if it is true and you can show it.
10. The plain description. "Send invoices and get paid."

Number ten is the control. Every test needs the boring line in it, because the boring line wins more often than anyone likes, and a test without it cannot tell you that. The longer catalogue of formats, with real tile lines, is in one-line pitch examples.

What "a live audience" means in one day

A hook test needs strangers who can act. Followers are not strangers; they will click your worst line out of goodwill. Your inbox is not strangers either. So for each hour you need a place where people who do not know you read one line and either do something or do not.

The options that work inside a day, honestly compared:

Your own email list, split. A subject-line split on a list of any size is the cleanest hook test that exists, because the audience is held constant and only the line changes. It is also not strangers, so treat a win as "my existing readers prefer this", not "the market does".

A community thread. Post a question, not an advert, and read which phrasing gets replies. Slow to read, and the rules of most communities forbid running ten of them.

A paid slot with a public count. BidSurvivor — ours, so weigh it accordingly — shows one brand per twelve-hour slot with one line of copy read off the brand's site, and publishes the card opens and click-throughs on the brand's own page. That is a live audience of strangers with a readout you can point at. It is also one line per slot, which matters for the next section.

A landing-page headline swap. Hold the traffic source constant, change the headline every hour, and log the click on the call to action per hour. This is where the ten-in-ten actually happens, whatever the source of the traffic. The B2B landing page examples show the layout that makes a headline swap a one-line edit.

The sample-size problem, stated plainly

Here is the sum nobody puts in the vibe-marketing threads. Suppose a traffic source sends, say, 40 people in twelve hours — a hypothetical for the arithmetic, not a figure from anywhere. Across ten hourly headlines that is roughly four people per hook. Four people cannot tell one line from another. Three of them clicking on the number line and none on the question line is a coin landing the same way three times.

So ten hooks in ten hours is not a measurement. It is a screening round. You are not looking for the best line; you are looking for any line that gets any reaction at all from strangers, and any line that gets none. A hook that four strangers ignore in a row is not proven dead, but it has earned its place at the back of the queue. A hook that two of four people act on has earned a full slot of its own.

The screening round produces a shortlist of two or three. Those go into a proper test: one line, one full twelve-hour slot each, everything else identical, the way the one-day test walkthrough sets it up. Two lines across two full slots is still a small sample, but it is a sample of one variable, and a clear winner will show. A narrow winner will not, and you should say so in your notes rather than pick the one you liked.

How people actually read a hook

The reason the first two words of a line carry most of the weight is not folklore. The Nielsen Norman Group has been eye-tracking readers since the nineties and the finding has not moved: people scan rather than read, in an F-shaped pattern, and the first words of a line get the fixation the rest of the line does not (NN/g, "How Users Read on the Web"). On an ad tile, where the line sits under a logo and beside eleven other things, that effect is stronger, not weaker.

Practical consequence: front-load. "Get paid nine days sooner" scans as Get paid. "Our platform helps you get paid nine days sooner" scans as Our platform, and Our platform is nobody's problem. Rewrite every one of your ten so that the first two words are the reader's problem or the reader's outcome, then look at the list again. Half of it will have collapsed into duplicates, and the remaining five are your actual test.

The hour-by-hour, for the record

• Hour 0. Write ten, in the ten formats, including the plain one. Front-load each. Cut duplicates.
• Hours 1 to 8. One line per hour on the landing page headline, traffic source held constant. Log clicks on the call to action per hour, and the hour of day, because 14:00 and 22:00 are not the same audience.
• Hour 9. Shortlist anything with a reaction. Bin nothing permanently; write down why each line dropped.
• Hour 10. Book the next full slot for the leading line, and the one after for the runner-up. On our board that can be up to seven days ahead, and an empty slot is $0.

One honest caution to end on. A hook that wins on a board where nobody was looking for you is a hook that works cold. That is a useful property and a rare one, but it is not the same as a hook that wins on a search results page where the reader already wants what you sell. Test cold hooks cold, warm hooks warm, and do not transfer the result from one to the other without saying so.

---

Sources. Reading patterns: Nielsen Norman Group, "How Users Read on the Web". This month's "vibe marketing" coverage: mean.ceo. Board rules: how BidSurvivor works.