Ad Hook Types: Question vs Stat vs Story (Prompt Picker)

A B2B SaaS that schedules shifts for warehouse teams runs the same ad to the same cold audience for six weeks and the click-through rate never moves off 0.4%. The landing page is fine and the offer is real. The first line of the ad — the hook — is a feature list, so nobody stops to read the second line. In paid acquisition, the hook buys the next three seconds; everything after it is only read if the hook wins.

This picker compares three workhorse hooks — question, stat, and story — and tells you which to use, on which channel, with what prompt. The running example is the same throughout: a warehouse scheduling SaaS selling to operations managers. For hook variants at scale, see the Facebook ad copy workflow, search-focused versions in the Google Ads copy prompt, and a wider set in the AI marketing prompts list.

Hook Type Decision Table

Two questions pick the row. How aware is the reader of the problem — do they already name it, or do you have to name it for them? And how much time do they give you — a search query means intent and a few seconds of attention, while a cold feed means no intent and half a second. High awareness plus intent favors a question. Low awareness plus a value claim favors a stat. Low awareness plus low intent favors a story.

When to Use Each Hook

Question. Use a question when the reader already lives with the problem and will nod at the first line. For warehouse scheduling, “Does your shift handover still run on a whiteboard?” lands on an operations manager who has felt that pain. The question works on search because the searcher typed the problem, and on retargeting because your earlier visit planted it. It fails on a totally cold audience that has never labeled the problem as a problem.

Stat. Use a stat when you can make one number carry the offer. “Warehouses lose 11 hours a week to manual shift planning” stops a skeptical operations lead better than any adjective, provided the number is specific and traceable. Stats perform on LinkedIn and in cold B2B feeds where the reader is rational, comparison-minded, and short on patience. A round, unsourced number does the opposite and reads as marketing noise.

Story. Use a story when the product is easier to feel than to specify, or when the feed is entertainment-first. “At 4:40 a.m., Dana rebuilt tomorrow’s shifts from a printed sheet” makes a scheduling tool human. Stories shine on Instagram, TikTok, and Meta feeds where attention is passive. They fall flat in front of a buyer who arrived with a checklist of requirements and wants the answer, not the anecdote.

Prompt Snippets by Hook

Question hook

Write 5 Facebook ad variations for a warehouse shift-scheduling SaaS.
Hook type: question.
Rules:
- Open with one question that names the exact daily pain of an operations manager.
- Keep the question under 12 words and avoid "Are you tired of...".
- Follow with two lines that hint at the fix, not the full pitch.
- End with a single call to action to start a free trial.
- No emojis, no exclamation marks. Each variation under 60 words.
Audience: operations managers at 50-500 person warehouses.

Stat hook

Write 5 LinkedIn ad variations for a warehouse shift-scheduling SaaS.
Hook type: statistic.
Rules:
- Open with one concrete number about lost time, overtime cost, or scheduling errors.
- Attribute the number in parentheses as ["industry average", "customer data"], leaving the exact figure to be replaced later.
- Move straight to the mechanism that fixes it.
- Close with a low-commitment call to action, such as "see a 4-minute demo".
- Each variation under 55 words. No hype adjectives.
Audience: operations and logistics leads on LinkedIn.

Story hook

Write 5 Meta ad variations for a warehouse shift-scheduling SaaS.
Hook type: story.
Rules:
- Open with a named person in a specific moment, time, and place.
- Show the friction in two sentences, using concrete objects, not abstractions.
- Reveal the turning point in one sentence, then one line of proof.
- End with a warm call to action that fits a scrolling feed.
- Each variation under 70 words, plain language, no jargon.

Trade-Offs

Question hooks are cheap and fast but invite a no. If the reader answers “not really”, you have lost them in the first line, so questions demand a genuinely shared pain. Stat hooks are the most credible and the most fragile: one soft number and the whole ad reads as unreliable, which is why sourcing matters more than cleverness. Story hooks build the strongest emotional pull and the slowest purchase; they work for discovery and retargeting nurture far more than for a bottom-funnel demo request.

The practical move is not to choose one forever but to match hook to funnel stage and to test in that order. Cold feeds start with story; a warm audience that already knows the problem gets the question; a comparison-stage audience gets the stat. Run all three against the same offer, give each its own ad set, and read click-through rate at equal spend before you declare a winner. Then take the winning hook and ask the model for ten variations of that single type — narrowing the hook does more for performance than adding adjectives ever will. Keep the losing hooks, too: a hook that fails cold often wins when it is served to the warm list.

Tuning a Hook That Almost Works

Most underperforming ads are not the wrong hook type; they are the right type with a weak first line. Before you abandon a hook, ask the model to rewrite only the opener and leave the body alone. That isolates the variable, so a lift in click-through is attributable to the hook and not to a rewritten offer.

Here is my current warehouse scheduling ad and its measured results: [PASTE AD] and [CTR, impressions, audience].
Hook type to keep: [question / stat / story].
Task: rewrite ONLY the first sentence, five times, and leave every other line unchanged.
Rules:
- Each new opener must be at most 12 words.
- Keep the hook type identical; do not switch to another type.
- Avoid the phrases "are you tired of" and "in today's world".
- Rank the five openers from most to least likely to stop an operations manager.
Return the five openers, a short reason for the ranking, and the unchanged ad beneath each.

Run the five openers as five ads in one ad set, equal budget, and read results after the spend is meaningful rather than after a day. The opener that wins on a cold feed may lose on the warm list, which is why the same test is worth repeating per audience before you scale the winner into the rest of the campaign.

Sources & Notes

Prompts tested with ChatGPT on the GPT-4o and GPT-4.1 models in October 2026; the same instructions work for ad copy in Claude and Gemini. Hook-type categories follow common direct-response practice. The warehouse scheduling SaaS and its audience are an illustrative scenario, and the sample statistics are placeholders to be replaced with sourced figures. No ad platforms or analytics APIs were called; all prompts are copy-paste text. Category mapping follows the guunn.com AI Prompts plan (Marketing = 959).