Hook bank

Social proof hooks that let real customers open the ad

A social proof hook opens the ad with evidence that other people already chose the product: a review theme counted across hundreds of reviews, a customer's own words, a reorder rate, a waitlist or a press pick. The viewer borrows their confidence. It only works when the proof is real, recent and specific, so the writing starts with your review export and order data, not a blank page.

Social proof hooks

Hooks
63
Categories
7
Updated

When to use them

Social proof hooks open with what other people say, buy or do, so the viewer borrows their confidence. Every rating, number and quote must be real and current.

  • Young or unfamiliar brands, where the first doubt is whether anyone else has bought this.
  • Products judged on experience, like scent, comfort, fit or taste, which a customer describes better than a brand can.
  • Gifts, where the buyer wants to know the person opening it will be pleased.
  • People who have seen the brand once and are still deciding whether it's any good.
  • Restocks, waitlists and long-running bestsellers, where buying behaviour is the evidence.
  • Not brands without real reviews yet. Use a story or education hook until you have them.

35 social proof hooks

Fill every [blank] with something true and specific about your product.

  1. [Number] reviews mention [specific detail]. Here's one.

    A count of reviews on one theme shows the praise isn't cherry-picked from a single fan.

    For example: 212 reviews mention the Sudsbury oat bar not going soggy in the dish. Here's one.

  2. "[Phrase customers repeat]" came up [number] times in our reviews.

    Repetition across many customers is stronger evidence than any one glowing quote.

    For example: "Smells like a garden after rain" came up 41 times in our Velmira reviews.

  3. We asked [number] customers what they'd tell a friend. Top answer: [answer].

    Framing a survey as advice to a friend draws out plain, believable language instead of marketing words.

    For example: We asked 300 Quirewell customers what they'd tell a friend. Top answer: "get the dotted one".

  4. Our most common complaint about [product], and what we changed.

    Leading with a fixed weakness signals honesty and settles the doubt a careful buyer was about to raise.

    For example: Our most common complaint about the Quirewell planner was the stiff spine. Here's the new lay-flat binding.

  5. Why [product] has [rating] stars, not 5.

    Explaining the missing stars makes the rating easier to believe and answers the main objection in the same breath.

    For example: Why the Duvra duvet has 4.6 stars, not 5: some sleepers find it too warm for midsummer.

  6. [First name] has reordered [number] times. We asked why.

    One named repeat buyer is behaviour proof, and asking why promises their real words next.

    For example: Meera has reordered Milepair trail socks six times. We asked why.

  7. Waitlist: [number] names. Back on [date].

    A real waitlist shows demand as a number and gives a date to act on without inventing urgency.

    For example: Waitlist: 2,300 names. The Tinlark tea hamper is back on 2 December.

  8. [Number] of every [number] first orders turn into a second order within [timeframe].

    A repeat-purchase rate proves people still liked the product after the box was opened.

    For example: 4 of every 10 first Leafturn boxes turn into a second box within two months.

  9. The [product] [city] orders most.

    Local proof narrows other people down to people nearby, which feels closer and easier to trust.

    For example: The hamper Bengaluru orders most: Tinlark's tea and biscuit tin.

  10. Unfiltered customer photos of [product] after [timeframe].

    Unedited photos from buyers show the product in real homes, where studio shots can't follow.

    For example: Unfiltered customer photos of the Reedwove laundry basket after a year of daily use.

  11. [Number] people tagged [product] this [period]. Here are [number] of them.

    Tagged posts show customers chose to share the product unprompted, a stronger signal than a quote you asked for.

    For example: 640 people tagged their Velmira bottle this month. Here are nine of them, shared with permission.

  12. The [product] our team buys with their own staff discount.

    It names the insider connection openly, which keeps staff enthusiasm honest rather than disguised.

    For example: The notebook the Quirewell team buys with their own staff discount: the A5 grid.

  13. Customers named it before we did: [nickname].

    A nickname coined by buyers is proof of affection that sounds nothing like ad copy.

    For example: Customers named them before we did: the forever socks. Milepair, since 2023.

  14. Read our [number]-star reviews first.

    Sending people to the critical reviews shows confidence and lets you answer the strongest objection on your own terms.

    For example: Read our 3-star reviews first. Most say the Sudsbury bar is small, so here's how long one lasts.

  15. Every review from [month], including the bad ones.

    Showing a complete, unfiltered month removes the suspicion that the reviews in the ad were curated.

    For example: Every Duvra review from September, including the two bad ones.

  16. Our customer inbox, summed up in one line: "[line]"

    Support messages are unprompted, so a line drawn from them reads as overheard rather than requested.

    For example: Our customer inbox, summed up in one line: "When is the lavender bar back?" Sudsbury.

  17. Most [product] buyers picked [variant] this [period].

    Showing which option most buyers chose answers the which-one-should-I-get question with behaviour.

    For example: Most Quirewell buyers picked dotted pages over lined this term.

  18. Picked by [publication or award] for [specific reason].

    Naming why the publication chose it turns a press mention into a reason to buy, not just a logo.

    For example: Picked by Crumbwell magazine's gift guide for the handwritten card in every Tinlark hamper.

  19. Back at [market or fair] for the [number]th year.

    Returning to the same market year after year is proof from organisers and regulars, not just online buyers.

    For example: Back at the Kestrel Lane Craft Fair for the 7th year with Reedwove baskets.

  20. Reviewed by the person who got it as a gift: "[quote]"

    For gift products, the recipient's reaction is the result the buyer is really paying for.

    For example: Reviewed by the person who got it as a gift: "I've used the Tinlark tea tin every morning since."

  21. Questions people asked before buying [product], and what they said after.

    Pairing the doubt before purchase with the answer after it shows the doubt settled by real experience.

    For example: Questions people asked before buying the Duvra duvet, and what they said after the first week.

  22. You've probably seen [product] on your feed. Here's why people post it.

    It acknowledges buzz the viewer may already have noticed and turns it into a concrete reason.

    For example: You've probably seen the Velmira bottle on your feed. Here's why people post it: the glass stopper.

  23. The compliment our customers hear most: "[compliment]"

    For products other people notice, like scent or style, the reactions of strangers are the proof buyers want.

    For example: The compliment Velmira customers hear most: "What are you wearing?"

  24. [Number] [product]s shipped. The first buyer still has theirs.

    Joining a big count to one long-lived first order combines scale and durability in a single line.

    For example: 10,000 Reedwove baskets shipped. The first buyer still uses hers for laundry.

  25. [Number] [uses] in, here's what [first name] says.

    Attaching a usage count to a review tells the viewer the opinion formed after real wear, not on unboxing day.

    For example: 200 kilometres in, here's what Arjun says about the Milepair heel.

  26. Rated [rating] by people who [demanding use].

    Filtering the rating to a demanding group of users makes it relevant to anyone with the same habit.

    For example: Rated 4.8 by people who run more than three times a week. Milepair trail socks.

  27. Stars are easy. This is the detail reviewers mention: [detail].

    Moving past the star rating to one concrete detail gives a sceptical viewer something to check.

    For example: Stars are easy. This is the detail Duvra reviewers mention: no feathers escaping through the seams.

  28. A comment from last week: "[comment]". Our answer: [answer].

    Replying to a real public comment shows the conversation is live and that other people are already interested.

    For example: A comment from last week: "Does it fit A4 sheets?" Our answer: yes, folded once. Quirewell folio.

  29. [Number] people saved [product] to their wishlist this [period].

    Wishlist counts are early demand proof for new products that don't have many reviews yet.

    For example: 830 people saved the Leafturn mystery box to their wishlist this week.

  30. The pair our customers buy together most: [product] and [product].

    Showing what others bundle suggests a choice and lifts the order without a discount.

    For example: The pair Sudsbury customers buy together most: the oat bar and the sisal soap dish.

  31. What [first name] expected, and what arrived.

    Contrasting expectation with experience in a buyer's own words works as a result and a review at once.

    For example: What Kavya expected from a Reedwove basket, and what arrived: "heavier, and it doesn't smell of glue".

  32. [Professionals] who [do the work] keep ordering [product].

    Repeat orders from people who use the category for work signal quality to everyone else.

    For example: Illustrators who sketch every day keep ordering the Quirewell dot-grid pad.

  33. No cherry-picking: our [number] newest reviews of [product].

    Choosing the newest reviews rather than the best ones shows the product is good now, not just at launch.

    For example: No cherry-picking: the five newest reviews of the Duvra duvet.

  34. Stocked by [number] independent [shops] that only reorder what sells.

    Retailer reorders are proof from buyers whose income depends on choosing well.

    For example: Stocked by 48 independent bookshops that only reorder what sells: the Leafturn reading journal.

  35. Real customer, unpaid, unedited: [first name] on [product].

    Saying the customer wasn't paid tackles the most common suspicion about review-style ads, so it must be literally true.

    For example: Real customer, unpaid, unedited: Rohan on his Milepair socks after a marathon.

By product category

  • Bath and body

    • The scent reviewers ask us never to discontinue: [scent].
    • [Number] reviews mention the bar lasting past [timeframe].
    • The bath set customers reorder before the first one runs out.
    • Unfiltered photos from customers' bathroom shelves.
  • Fragrance

    • The compliment [product] customers hear most: "[compliment]"
    • Sample buyers who came back for the full bottle: [number] of every [number].
    • Reviewers describe it in three words: [word], [word], [word].
    • The scent [number] reviewers say strangers asked them about.
  • Bedding

    • [Number] reviews mention sleeping through a [hot or cold spell].
    • Why our duvet has [rating] stars, not 5.
    • Guests at [number] homestays asked where these sheets came from.
    • Washed weekly for a year: customer photos of the same set.
  • Books

    • The book our subscribers rated highest this year.
    • Book clubs in [number] cities picked this for [month].
    • Reviews from readers who say they rarely read [genre].
    • Independent bookshops keep reordering it. Here's why.
  • Gift hampers

    • Reviewed by the person who opened it, not the person who paid.
    • The hamper [city] offices ordered most for [festival].
    • [Number] repeat business orders since last [festival].
    • Photos recipients sent back, shared with their permission.
  • Sports gear

    • [Number] kilometres in, what runners say about the [part].
    • The [product] our local club bought for the whole squad.
    • Customers named them before we did: the [nickname].
    • Rated [rating] by people who train [number] times a week.
  • Handmade crafts

    • Back at [market] for the [number]th year.
    • [Number] repeat buyers from our very first market stall.
    • Photos from customers' homes, [number] years after they bought.
    • Every piece is signed. Here are the notes buyers sent back.

Mistakes to avoid

  • Numbers without a date

    "Thousands of happy customers" with no date can't be checked and may already be stale. Put a month or a range on every count, and update it when the figure moves.

  • Quotes the brand rewrote

    Trim a review for length, but never change what it says. Polished enthusiasm sounds like marketing, and it turns into a testimonial the customer never actually gave.

  • Praise with no reason in it

    "Love it!" with five stars tells the viewer nothing. Pick reviews with one concrete detail, like the soap that doesn't go soggy or the sock that survived a marathon.

  • A face that isn't the reviewer

    A real quote beside a stock or AI-made face implies that person said it. Use a plain quote card, the product, or the reviewer's own photo with their permission.

  • Proof aimed at the viewer

    "Other runners over 50 love these" implies the viewer's age, and Meta's personal attributes policy says ads must not assert or imply traits like age, health or financial hardship. Say who buys it, such as "runners over 50 reorder these most", not who is watching.

What a social proof hook does

Nobody wants to be the first person to try a brand they've never heard of. A social proof hook answers that hesitation in the opening line by reporting that other people already took the risk and were glad they did. It doesn't argue that the product is good. It shows that buyers decided it was, and lets the viewer borrow their judgement.

That makes it the most fragile hook in the bank. A curiosity hook can be a little playful, and a result hook can round a figure. A social proof hook is worth nothing the moment it isn't true, because the truth is the whole offer. So the craft is less about clever wording and more about finding proof that is real, recent and specific, then saying it plainly. The social proof glossary entry covers the wider idea.

Five kinds of proof and the doubt each one answers

Pick the kind of proof that answers your product's biggest doubt, not the one that sounds most impressive.

Kind of proof What it looks like The doubt it answers
Volume Review counts, themes counted across reviews Is anyone actually buying this?
Voice A customer's own words What is it like to use?
Behaviour Reorders, waitlists, sell-outs, bundles Is it still good after the first one?
Peers A city, a club, a profession Does it suit someone who uses it the way I would?
Authority Press picks, awards, shops that stock it Is this a legitimate brand?

A scent or a duvet is judged on experience, so voice and behaviour usually carry it. A new gift brand fights doubts about legitimacy, so authority and volume do more of the work.

Build a proof inventory before you write

Social proof hooks come from data, not imagination. An afternoon with your store exports gives you more usable lines than a day of brainstorming.

  1. Export your reviews and search them for phrases that repeat. Count how many reviews use each one.
  2. Pull behaviour numbers from your store: repeat-order rates, waitlist sign-ups, best-selling pairs. Write down the date range for each.
  3. Read the support inbox for unprompted lines. A question like "when is the lavender back?" is demand in the customer's own voice.
  4. List press mentions, awards and stockists with dates and, where you know it, the reason they chose you.
  5. Flag incentivised reviews. If a review came with a free product or a discount, mark it. The rules further down this page treat hidden incentives and rewards for positive reviews as a problem, so those reviews either carry the disclosure or stay out of the ads.
  6. Record permission. Note which customers agreed to be quoted or to have their photo used, and keep the message.

Give every entry a date, and drop or refresh it when it goes stale. Then add the strongest entries to your brand kit. BulkAds' ad copy only uses the reviews and proof points you've added there, so the inventory becomes the outer limit of what any ad can claim.

A worked example: one soap bar, five proof hooks

Say Sudsbury sells a handmade oat soap bar for $2.55 and a set of three for $6.70. Its inventory after one afternoon: 1,180 reviews averaging 4.7 stars, 212 of them mentioning that the bar doesn't go soggy in the dish, 38 percent of April's first orders followed by a second within 90 days, and 1,900 names on the lavender bar's restock list.

Hook Kind of proof Layout to pair it with
212 reviews mention the bar not going soggy. Here's one. Volume and voice Five-star review
Waitlist: 1,900 names. Lavender is back on 3 November. Behaviour Back in stock
Our customer inbox in one line: "When is the lavender back?" Voice Comment reply
The pair customers buy together most: the oat bar and the sisal dish. Behaviour Bundle
Why the oat bar has 4.7 stars, not 5. Voice, flaw included Reddit post

Two kinds from the table above are missing. Sudsbury has no press picks, stockists or club orders yet, so authority and peers are missing. That's fine. Five hooks from three kinds of proof is still a fair test, and the gaps tell you what to collect next, such as a local shop that would stock the bar.

Notice what stays out of the hooks, too. The set of three works out at about $2.25 a bar against $2.55 for one. That's a real reason to buy, but it's an offer, not proof. Put it in the headline or description and keep the hook doing one job, because a deal squeezed into a review line weakens both.

What the rules say about reviews in ads

As of October 2026, two consumer rules spell this out in detail. In the US, the Federal Trade Commission's rule on consumer reviews and testimonials (opens in a new tab), announced on August 14, 2024 and in effect since October 21, 2024, bans fake reviews and testimonials, including AI-generated ones, that claim to come from people who don't exist or never used the product. It also covers reviews from company insiders that don't clearly disclose the connection, rewards offered on the condition that a review says something positive (or negative), and buying fake followers or views when you knew or should have known they were fake.

In the UK, the Competition and Markets Authority's fake reviews guidance (opens in a new tab), published on April 4, 2025, explains a banned practice added by the Digital Markets, Competition and Consumers Act 2024. It treats fake reviews and concealed incentivised reviews as automatically unfair, and it names suppressing genuine negative reviews and selectively promoting positive ones as ways of publishing reviews misleadingly. That makes a useful test for volume hooks too: a count or a star rating in an ad should come from all your reviews, not a filtered set. Other countries have their own consumer laws, so check the rules wherever you sell.

Meta draws a separate line, about the viewer rather than the reviewer. Its personal attributes policy (opens in a new tab), as of October 2026, says ads must not assert or imply personal attributes such as age, religion, physical or mental health or financial hardship, and its own examples reject lines like "Are you 18 years old?" and "Meet other Buddhists". Peer proof is where this bites. "Other eczema sufferers swear by it" implies the viewer has a medical condition. "The unscented bar our customers reorder most" is about the buyers and the product, which is the safer shape.

AI-made visuals and real reviews

BulkAds builds image ads around your real product photo, so a real review can sit on a clean quote card beside your packshot, or over a UGC-style product scene. What it shouldn't do is give the quote a face. An AI-made person holding a phone next to a real review reads as the reviewer, and that's a testimonial nobody gave. The FTC's own questions and answers on the rule (opens in a new tab), as of October 2026, say there's no blanket ban on AI avatars in marketing, but a testimonial an avatar delivers is still prohibited when it's fake or false. The simple choice is quote cards, product shots, or the customer's own photo with their permission.

Testing social proof hook variants at volume with BulkAds

Test the kind of proof before you test the wording. Hold each layout steady and change only the evidence on it.

  • The five Sudsbury hooks from the table, each on two layouts (a review card and a UGC-style product shot), make ten ads in one batch, which uses 20 credits.
  • Write the batch instruction around the inventory: "Social proof hooks only. One hook per ad, from these five: [paste them]. Quote reviews exactly as written. No other numbers."
  • Put every quote and count in the brand kit before you generate. The generator is told never to invent reviews, ratings, awards or numbers, so proof that isn't in the kit can't reach an ad.
  • Read the pre-flight report for on-image text. A quote that has lost a word is no longer the customer's quote, so regenerate that element or fix the copy before anything launches.

Launch the batch paused through the bulk ad launcher, then approve it, and its results sync back into BulkAds for you to judge. Or let Autopilot launch the test itself. On the tests it runs, it pauses any ad that reaches your stop line without a sale, and after 48 hours it marks ads with three or more sales at or under your target cost per purchase as winners. In the default ask-first mode it proposes each step up to you, and in hands-off mode it scales within your caps.

Two details matter more for social proof than for other hooks. First, the hook style grouping in creative analytics sorts lines by their wording, so "212 reviews mention" lands under number or stat while a plain quote usually lands elsewhere. That's a rough split between counted proof and voiced proof, not a label for each kind, so sort by purchases and read the five hooks ad by ad. Second, when Autopilot runs the test, each of the four variants it drafts from a winner is built to change one thing, and the proof point is one of the things it can change. A winning review-count ad can come back carrying the waitlist figure in the same layout, which tells you whether the layout or the evidence was doing the work.

Social proof has one advantage at volume: the facts behind it keep growing. When the review count or the waitlist moves, update the proof point in the brand kit, then edit the copy or regenerate the on-image line with the new figure. The concept stays and the evidence gets fresher. When you need something genuinely new rather than a refresh, the concept versus iteration guide helps you tell which one you're facing.

Questions people ask

What is a social proof hook?

A social proof hook opens an ad with evidence that other people already bought and liked the product, such as a counted review theme, a customer quote, a reorder rate, a waitlist or a press pick. It works because the viewer borrows the confidence of earlier buyers, so every number and quote has to be real.

Can a new brand use social proof hooks?

Yes, on a small scale and honestly. Ten specific reviews, a waitlist count or a return visit to the same market are all real proof. If you have none yet, use a [story hook](/hooks/story) or an [education hook](/hooks/education) until reviews arrive, and never write your own.

Can AI write or polish customer reviews for my ads?

It shouldn't write them. As of October 2026, the FTC's rule on reviews, in effect in the US since October 21, 2024, bans AI-generated reviews that claim to come from people who don't exist or never used the product, and the UK's fake reviews guidance treats fake reviews as a banned practice. AI can help you search, count and trim real reviews, and BulkAds only uses the reviews and proof points you've added to your brand kit.

Should I ask permission before quoting a customer in an ad?

It's good practice to ask and keep the reply, especially when you show a name or a photo. A public review feels shared, but an ad puts someone's words in front of strangers they didn't choose. Rules on this vary by country, so check where you sell.

How is a social proof hook different from a testimonial ad?

The hook is the opening line or first frame. A [testimonial ad](/ad-templates/testimonial) is a layout built around one review. A social proof hook can open a testimonial, but it can also open a back-in-stock, bundle or comment-reply ad with a count or a behaviour instead of a quote.

Which social proof hook should I test first?

The one that answers your product's biggest doubt with your strongest real asset. For experience products that's usually a counted review theme or a quote. For new gift or premium brands it's often a press pick or a stockist. Test two or three kinds of proof against each other before rewording any of them.

Keep going

These hooks are free to copy and adapt for your own ads. Product names in the examples are fictional.

BulkAds AI is an independent product and is not affiliated with, endorsed or sponsored by Meta Platforms, Inc. Facebook and Instagram are trademarks of Meta Platforms, Inc.

Your next fifty ads are one link away.

Paste your website, meet your brand kit and see your first batch in minutes. 7 days for $9.99 — cancel anytime.

yourbrand.comMake my ads

See pricingExplore featuresBrowse the library