Trust Signals, Proof and Case Studies

The five questions that produce usable testimonials

Each question targets one of the seven worries from chapter one. Send them in writing rather than asking on a call - the customer will give better answers when they have a few minutes to think than when they're put on the spot. Send no more than three at a time; five is the maximum across two emails. The five questions, in priority order:

One: 'In one or two sentences, how would you describe the problem we solved for you?' (Targets the first worry: will this solve my problem.) Two: 'What's specifically different now compared with before we started working together? A number or a concrete change is most useful.' (Targets the same worry, with measurable proof.) Three: 'How would you describe what we're like to deal with day to day?' (Targets the third worry: will they be pleasant.) Four: 'What would you say to someone in your position six months ago who's deciding whether to spend this kind of money on this kind of help?' (Targets the price-fairness and safety-net worries together.) Five: 'Is there anything we did differently from how you expected, in a good way or a bad way?' (Surfaces the unexpected proof points you'd never have thought to ask about.)

When to ask

Three moments work well. The first is two to four weeks after a clear delivery moment - far enough that the customer can see the outcome, close enough that the experience is fresh. The second is at a milestone in an ongoing relationship - end of the first quarter, end of the first year, after a particularly good piece of work. The third is when the customer spontaneously says something appreciative in another channel ('that report was really helpful, thanks') - reply with thanks, then ask if they'd be willing to put a fuller version in writing for the website. Each of these moments produces better answers than the generic 'how's it going?' check-in.

The moments that don't work: immediately after delivery (too soon for outcomes to be visible), at the end of a renewal cycle (the customer is in admin mode, not appreciation mode), or in response to an awkward situation you're trying to recover from. Asking for proof in any of these moments produces either nothing or something that doesn't read as real. Wait for the right moment instead.

What to do with the answers

Most customers reply with usable raw material that needs light editing rather than rewriting. The temptation to smooth the customer's prose into your house style is strong and almost always wrong. Real customer language - including the small grammatical roughness and the slightly clumsy phrases - is part of what makes the testimonial believable. Smooth it too far and the testimonial reads as written by you, which destroys most of its value.

Three editing rules. First, you can cut. Reply long? Take the strongest two or three sentences. Second, you can reorder. The most useful sentence often comes near the end of a longer reply; bring it to the front. Third, you can add a short attribution if the customer didn't include it - but only with their permission. What you can't do: paraphrase, combine multiple customers' words into one quote, add specifics that the customer didn't mention, or change the tense to make a past observation sound more present.

Consent in writing

Always written consent before publishing. Reply to the customer's email with a draft of how the testimonial will appear (the quote itself, the name, the role, the business name and any photo if you have one). Ask three explicit questions: are you happy with how this is worded, are you happy for it to appear on our website and marketing materials, and would you like anything changed? Wait for the explicit yes. Save the email.

Allow the customer to ask for the testimonial to come down later if their circumstances change (they leave the company, they have a falling-out with the wider business, anything). Honour the request within the week. The cost of removing one testimonial is much smaller than the cost of building a reputation for ignoring customers' wishes about their own words.

Reviews on third-party platforms

Google, Trustpilot, Facebook, sector-specific platforms (Checkatrade, Houzz, Bark, etc.). These do work that on-website testimonials can't - they're verifiable, the visitor doesn't have to trust you to have invented them, and the platform's brand carries some of the trust weight. The cost is that the platform owns the proof and shows it through their own filter; you can't curate which review shows top.

The right approach for most small businesses is both. Gather direct testimonials for placement on your own pages (where you control the placement and matching). Encourage Google and one other relevant platform for the verifiable third-party signal. Don't try to game either - manufactured reviews are usually obvious and the platforms penalise them. The simplest 'encouragement' that works: a short note in the email that goes out two weeks after delivery, saying 'if you'd be willing to leave us a Google review, here's a direct link' with the actual review link in the email. About fifteen to thirty percent of customers will, over time. That's plenty.

Worked customer asks for four businesses

Bookkeeper. Email two weeks after the first quarterly tax planning call: 'Tom, hope the new tax estimate has helped you sleep better. We're putting together some short notes from clients to put on the website - would you be willing to answer two quick questions for me? One: how would you describe what we solved for you compared with the pre-us situation? Two: what would you say to a tradesperson in your position six months ago who was deciding whether to hire a bookkeeper?' Three sentences in reply usually produces a usable two-sentence quote.

Plumber. After a planned annual visit: 'Mrs Patel, thanks for the tea. Would you mind leaving us a quick Google review? It really helps other local homeowners decide. The link is here.' Specific timing (right after a visit when the customer is feeling positive), one specific platform, one direct link. About a third of customers leave a review when asked this way.

Freelance designer. After project sign-off, two weeks later: 'Anna, hope you've had time to live with the new site. I'd love to write a short case study about the project for our website - it would take about thirty minutes of your time on a call, and you'd see and approve everything before it went up. Are you up for that?' Different format (case study not testimonial - chapter three covers the difference) but same gathering principle: ask specifically and at the right moment.

Homewares shop. Automated email three weeks after delivery: 'Hope you're enjoying the bedding. If you have a minute, we'd love your review on the product page - what worked, what didn't, and any photos of it in your bedroom. Here's a direct link.' Specific format, direct link, gentle ask, no follow-up if there's no response. Adds about thirty reviews a quarter to the product pages.

When you can't get a testimonial

Some categories make testimonials hard. Coaching and therapy clients often can't or won't be named. Bookkeeping clients sometimes can't share specific financial outcomes. Senior corporate clients sometimes can't have their company name attached to a small supplier's marketing. The half-version conventions from the storytelling eBook apply: first names only, sector but not company, specific kinds of outcome but not exact figures. Anonymised quotes are weaker than named ones but stronger than no quotes at all - and labelled honestly ('quote from a coaching client, used with permission, name withheld'), they don't damage trust the way fabricated quotes do.

What to do this week

Pick three customers whose work has gone well in the last three months and where you sense they'd happily reply to a short request. Send each of them an email using two of the five questions above. Wait. The replies that come back are the raw material for the testimonials you'll publish in the next chapter. Keep the email replies in the proof-library folder you'll set up in chapter seven.

Now move from short quotes to longer proof. The next chapter, 'Case Studies That Earn Their Place on a Busy Page', covers the structure that turns a real customer story into a two-minute read that actually does conversion work, rather than a twenty-page document nobody opens.