IIDigital MarketingGuide

Testimonials and References: Proving Trust Without Exaggerating

Every provider says its customers are satisfied. People only believe it when they hear it from the customers themselves, with names, numbers and a sentence about what did not go smoothly.

Envelopes and pen on paper
Photo: Towfiqu barbhuiya / Unsplash

Key takeaways

  • A testimonial works when it is verifiable: full name, position, company and a concrete result.
  • The proof ladder ranks evidence from weak to strong: anonymous quote, quote with name, review on a third-party platform, case study with numbers, reference call.
  • Flawless looks suspicious. A study with real purchase data found that purchase probability rises only up to about 4.2 to 4.5 stars and then falls again.
  • Testimonials belong where doubt arises: next to the offer, at the form, in the proposal letter.

Why does nobody believe “Our customers are satisfied” anymore?

Your website has a box with three quotes: “Very competent and reliable. M. S., Hamburg.” Next to it, the logos of five customers, without explanation. You know the quotes are real. The visitor does not, and has learned to skip such boxes because they look the same on every website.

The problem is not that testimonials do not work. The problem is that weak testimonials prove nothing. A prospect in business-to-business sales faces a decision whose result they will only see after the purchase. They look for signs that others in their situation have had good experiences. Anonymous praise does not provide that sign.

Turning claims into proof is one item in the checklist A website that generates inquiries. This article goes deeper and asks: how do you turn testimonials into proof that a prospect believes? To answer it, we first clarify what makes a reference credible, then introduce the proof ladder and ask why five stars look suspicious. After that we turn to how to collect good testimonials, where they belong, and what is legally permitted.

Praise without a name is a claim you put into a stranger’s mouth.

What makes a reference credible? Depth beats volume

Boyd, Sese and Tillmanns (2023) studied how the design of references works in business-to-business markets. They distinguish two decisions. Depth describes how richly a reference informs: a written case study, a video or a face-to-face conversation, such as a customer’s talk at a trade show. Breadth describes how many customers appear together in one reference.

Their reasoning: a deep reference costs the supplier more because the customer invests time and reputation. That is precisely why it is a credible signal. At the same time it shows more of what the collaboration really looked like, so a prospect can better judge whether it fits their situation. Many customers in one reference, by contrast, seem less credible, because each individual commits less and goes into less detail.

The authors tested this in two ways: through investors’ reactions to references published by listed suppliers, and through an experiment with business buyers. For the buyers, purchase intentions depended on two things: credibility and fit. Investors and buyers valued depth alike, while their assessments of breadth diverged. So there is no recipe for every situation, but there is a clear direction.

That the effort pays off in sales is supported by a study by Terho and Jalkala (2017). They developed a measure for the systematic use of references internally, for example for training and proposal work, and externally towards prospects. Companies that used references in this planned way had better sales results.

The proof ladder: from weak to strong

From these findings I derive a simple model, the proof ladder. Each step is more credible than the one before and takes more effort. You do not have to be at the top everywhere. But you should know which step your proof is on today.

  1. Step 1: Anonymous quote

    “Very satisfied. M. S.” Costs nothing, proves almost nothing. Usable only as a supplement.

  2. Step 2: Quote with full name

    Name, position, company, ideally a photo. The reader could call the person. That alone increases credibility.

  3. Step 3: Review on a third-party platform

    A review in the business profile or on an industry portal. You do not control it, so people are more likely to believe it.

  4. Step 4: Case study with numbers

    Starting point, approach, result, with the customer’s permission and at least one number. Shows how the collaboration really went.

  5. Step 5: Reference call

    A customer is willing to talk to a prospect on the phone or speak at an event. The deepest and most credible form.

Effort per piece of proof at each step of the proof ladder
  • Quote with name1 hours
  • Requesting a review0.5 hours
  • Case study with numbers6 hours
  • Preparing a reference call2 hours

Worked example with freely chosen numbers for the effort on your side. For a reference call the customer carries the larger share, so use it sparingly.

For a small business, a good mix is realistic: three to five quotes with full names on step 2, a steady flow of new reviews on step 3, two case studies on step 4 for the most important services, and two or three customers available for a call. How to get reviews continuously and reply to them is covered in the article Google Business Profile.

Why do five stars look suspicious?

Many businesses would prefer to show only praise. Research advises more composure. Maslowska, Malthouse and Bernritter (2017) linked the star ratings customers saw while shopping with actual purchases at three online-only retailers. They studied light bulbs, women’s athletic shoes, hair care and herbal vitamins, that is, consumer products.

4.2 to 4.5stars: purchase probability rose up to this point and fell afterwards
3online retailers with real purchase data
4product categories in the analysis

Source: Maslowska, Malthouse and Bernritter (2017)

The result contradicts the common assumption: more stars do not simply mean more purchases. Purchase probability rose up to about 4.2 to 4.5 stars and declined after that. More reviews did not always lead to more purchases either. The authors explain this with distrust: customers know that providers can influence reviews and deliberately look for critical voices to make sure nothing is being hidden. A few moderate reviews make the good ones look more credible.

The study concerns inexpensive consumer goods and a period of a few weeks in 2014. The basic logic fits business-to-business sales well, though, where buyers read skeptically anyway. For you, this means: do not delete justified criticism, answer it. And leave in your case studies what was difficult and how you solved it.

How do you get good testimonials?

Most satisfied customers are happy to help if you make it easy for them. The worst request is: “Could you write us a few lines?” The customer does not know what to write, postpones it and forgets it. A short conversation with fixed questions works better; you turn it into the text and let the customer approve it.

Four questions for a testimonial
QuestionWhat it is forWhat it becomes in the text
What was the situation before you hired us?Recognition for the readerOne sentence on the starting point
What made you hesitate beforehand?Anticipating other people’s objectionsThe doubt the reader also has
What specifically has changed?Proving the resultA number or a tangible example
Whom would you recommend us to, and why?Showing fitWho the service is meant for

Worked example

The right moment is shortly after a visible success: after acceptance, after the first month with the new system, when the customer thanks you of their own accord. Plan the request as a fixed step in project closure so it does not depend on chance. A side effect is a further argument: in three studies, Liu, Zhang and Chintagunta (2026) found that customers whose reviews were more emotionally colored were afterwards more likely to recommend and buy again, while factual, weighing content tended to have the opposite effect. So in the conversation, do not only ask for numbers, but also how the collaboration felt.

From request to finished testimonial in one year
  1. 30Completed jobs with a request
  2. 18Agreed to a conversation60% of previous stage
  3. 12Text approved67% of previous stage
  4. 5With a number or case study42% of previous stage

Worked example with freely chosen numbers. If you close every job with a request, after one year you will have enough proof for all important services.

Customers who give a testimonial are often also willing to recommend you. How to turn that into a plannable path to new orders is described in the article Referral marketing.

Where do testimonials belong?

Testimonials do not belong on a separate “References” page that hardly anyone opens. They belong where a prospect has doubts. That happens at several points in the sales process.

The right proof in the right place
PlaceThe prospect’s doubtFitting proof
Service pageCan they actually do this?Quote with name from the same industry
Landing page at the formIs it worth inquiring?Short quote on the process: fast, straightforward, honest
After the first conversationDoes this fit our case?Case study with a similar starting point
In the proposalWhat if it goes wrong?Case study with a problem that was solved
Before the decisionCan I ask someone?Offer of a reference call

On a landing page, the strongest proof sits right next to the button. The article Building a landing page explains why. Case studies also work as expert articles that sales sends after the first conversation. How that works is described in the article B2B content marketing.

Testimonials also work in combination with advertising. In two experiments with 317 and 123 participants, Dwesar, Sharma and Tyagi (2025) showed that combining customer reviews with an ad reduced skepticism towards the advertising. Those who saw the reviews first were then more strongly persuaded by the ad. The study used consumer products and lab conditions, but the direction is clear: proof from customers makes your own message more credible.

What is allowed? Approval, authenticity, data protection

Testimonials are advertising. That is why advertising rules apply to them, and because they contain names and often photos, data protection rules too. This is not legal advice. Have your approach checked once by a specialist law firm. A few principles still help in everyday work.

  • Every testimonial is genuine and reflects what the customer said. Shortened, yes; reinterpreted, no.
  • The customer has approved the text, their name, their photo and the logo in writing.
  • The approval is filed with a date, and the customer knows they can withdraw it.
  • Numbers in the case study come from the project and are agreed with the customer.
  • Any compensation for testimonials is disclosed or avoided.
  • If you show reviews on your own website, you state whether and how you check that they come from real customers.

What you can do this week

Start with an inventory and then work your way one step up the proof ladder. One hour is enough to begin.

  • Collect all existing testimonials and assign each one a step on the proof ladder.
  • Remove anonymous quotes or ask the customers for approval with their full name.
  • Pick three satisfied customers from the last six months and ask them for a short conversation using the four questions.
  • Plan a case study for the most important service with starting point, doubt, result and one number.
  • Add the request for a testimonial as a fixed step in project closure.
  • Decide which proof goes where on the website, the landing page and in the proposal.

After one quarter, your proof will be one step higher. You will not notice it in clicks but in first conversations in which the prospect says: “I read how it went at company X.” Events are also suitable for encounters in which customers speak for you themselves. More on that in the article Trade shows and events.

In short, a testimonial becomes proof when it can be verified: full name, role, company, and a specific result. The higher a piece of proof stands on the proof ladder, the more weight it carries, and flawless ratings tend to look suspicious rather than convincing. Proof belongs where the prospect's doubt arises.

Frequently asked questions

How do I write a good testimonial?

A good testimonial names the starting point, the doubt before the order, a concrete result and whom the service suits. It carries the full name, position and company. It is easiest to create from a short conversation whose text the customer approves.

Am I allowed to publish testimonials on my website?

Yes, if they are genuine and the customer has approved the text, name and, where applicable, photo and logo. File the approval in writing. For details on advertising law and data protection, you should seek professional advice.

What is the difference between a testimonial and a reference?

A testimonial is usually a short quote. A reference in the broader sense covers all proof in which customers speak for you, up to case studies and reference calls. The more detailed and personal the reference, the more credible it is.

Should I delete negative reviews?

You should leave justified criticism in place and answer it factually. A study with real purchase data shows that flawless ratings tend to arouse suspicion. Only reviews that violate the platform’s rules can be reported.

How many testimonials do I need?

Fewer than many think. Three to five quotes with full names, two case studies for the most important services and a few customers available for a call are enough for a small business. More important than the number is that they fit the industry and situation of the prospects.

When should I ask customers for a testimonial?

Shortly after a visible success, for example after acceptance or when the customer thanks you. Plan the request as a fixed step in project closure. Make it easy for the customer by offering a short conversation and writing the text yourself.

Sources

  1. Boyd, D. E.; Sese, F. J.; Tillmanns, S. (2023): The design of B2B customer references: A signaling theory perspective. Journal of the Academy of Marketing Science, 51(3), 658-674.
  2. Maslowska, E.; Malthouse, E. C.; Bernritter, S. F. (2017): Too good to be true: the role of online reviews’ features in probability to buy. International Journal of Advertising, 36(1), 142-163.
  3. Terho, H.; Jalkala, A. (2017): Customer reference marketing: Conceptualization, measurement and link to selling performance. Industrial Marketing Management, 64, 175-186.
  4. Liu, R.; Zhang, W.; Chintagunta, P. K. (2026): Customers' Review Content and Their Referral and (Re)Purchase Behaviors. Journal of Marketing, 90(3), 127-147.
  5. Dwesar, R.; Sharma, A.; Tyagi, M. (2025): Examining Psychological Mechanisms Underlying Synergistic Communication Effects: The Case of Online Reviews and Display Advertising. Informing Science, 28, 1-28.

All sources in the knowledge base

Reading room of a historic library with galleries, walls of books, and a skylight

Found a mistake or a topic that’s missing?

Feedback, corrections, and topic ideas go straight to the editorial team. We correct openly and with a date.