Numbers as Proof of Trust: Stating Customers, Projects and Years Honestly

Why numbers persuade, and why so many are not accurate

“For over 25 years,” “more than 500 customers,” “over 1,000 projects”: numbers like these appear on almost every B2B website. They are meant to show that the business is experienced and that others trust it. Often nobody in the business knows exactly how they came about. The 500 customers are everyone who ever received an invoice, the 1,000 projects include every minor repair, and the 25 years include the time of the predecessor business.

This analysis therefore answers the question: When do numbers such as the customer count, projects or years in business work as proof of trust, and how does a business state them honestly? We first clarify two models, one on the effect of majorities and one on the credibility of signals. Then we look at what studies show about precise and round numbers, weigh two views and derive a test for every number on the website.

A number does not convince through its size but because the reader sees what was counted.

The thinking models: social proof and signaling theory

The psychologist Robert Cialdini described seven principles of influence, which the article on persuading without tricks puts in context. One of them is social proof: what others do is considered right. Those who are unsure orient themselves by what many have done before them. A customer count condenses this principle into a single figure: this many have already chosen this business. Social proof works most strongly when the others are similar to the reader. A hundred customers from their industry tell them more than a thousand from anywhere.

Whether a number is believed is explained by the second model. According to Michael Spence’s signaling theory, which the article on trust signals on your website describes, a signal is credible when it would be costly or impossible to send for someone without the claimed quality. A number on a website costs nothing. Anyone can write “more than 500 customers.” The number becomes credible only when it can be checked: when the reader sees what was counted, or when names, references and cases support it.

What studies show about precise and round numbers

Schindler and Yalch (2006) studied whether the form of a number changes its credibility. 199 students each compared two ads for a fictitious deodorant. One made a claim in words, “lasts far longer than any other,” the other with a number, either round at 50 percent or precise at 47 or 53 percent. The authors’ reasoning: round numbers are the ones people give when estimating. A precise number therefore suggests that something was counted. The result confirmed the assumption. Precise numbers were judged more accurate and more likely to be based on data, and thus more credible, even by participants who fundamentally distrust advertising.

Xie and Kronrod (2012) found in four experiments on environmental claims that precise numbers make a company appear more competent, but mainly among readers with little distrust of advertising. Highly skeptical readers were less influenced, unless they knew little about the topic. Pena-Marin and Bhargave (2016) showed a countervailing effect: round numbers evoke the idea of stability, and product attributes stated in round numbers were perceived as working longer.

A third finding concerns expectations. Rossi and Schleef (2026) analyzed ratings of films before and after an Oscar nomination. Viewers who rated a film after the nomination gave measurably lower scores than similar viewers had before, and text analyses pointed to disappointment as the reason. The effect was strongest among less experienced users. Awards and large numbers raise expectations, and whoever fails to meet them is judged more strictly.

The studies concern consumers, students and films, not B2B buyers. Their direction is useful nonetheless: numbers work through what readers assume about how they came about and through the expectations they raise.

Precise or round, large or small? Two views

The first view recommends precise numbers. If readers take precise numbers as counted, “487 customers” is more credible than “around 500.” That argues for stating precise values where they exist, along with the date and counting method.

The second view warns against misusing this effect. If precise numbers seem more credible, an estimate can be disguised as a precise number. Schindler and Yalch themselves point out that readers make the false assumption that precise numbers are not estimates. Invented precision exploits exactly this error. Then there is the finding by Rossi and Schleef: whoever raises high expectations with large numbers is later measured more strictly.

I consider one rule right that combines both views: as precise as was counted, and not one digit more. Anyone with a list of all active customers states the precise number with a date. Anyone estimating says “around” or “about” and gives the basis for the estimate. And anyone who can choose between a large and a meaningful number takes the meaningful one: not everyone who ever received an invoice, but those who ordered again last year.

The number test: four questions for every number

The models yield a test that I call the number test. It turns a claim in digits into evidence. Every number on the website, in proposals or in presentations goes through four questions.

1. Unit of reference: what exactly was counted?

   Customers with an order in the last year, contracts, completed projects above a minimum value. The reader must know what one unit is.

2. Period: when, and over what time?

   Date of the figure and period of the count. “Since founding” is a period, but usually not a very meaningful one.

3. Counting method: where does the number come from?

   From accounting, the CRM, a survey with response rate. For surveys, the number of people asked and of responses belongs with it.

4. Comparability: does it tell the reader something about their case?

   Customers from their industry, projects of their size. A smaller, similar number convinces more than a large, unrelated one.

Numbers and names complement each other. On the evidence ladder from the article on customer testimonials and references, a bare number stands no higher than an anonymous quote. Only with references that can be looked up does it become a strong signal. The references page is therefore the natural place where a customer count finds its evidence.

What needs to be considered legally?

Numbers in advertising are not a private matter. Under the German Act against Unfair Competition (UWG), a commercial practice is misleading if it contains untrue information or information likely to deceive, for example about the characteristics of the company, its size or its experience (Section 5 UWG). A customer count reached only with an unusual counting method, or a number of years that includes a predecessor business without saying so, may fall under this. For reviews, since 2022 the following also applies: anyone who makes customer reviews available must state whether and how they ensure the reviews come from actual customers (Section 5b(3) UWG).

Claims of leadership such as “market leader” or “number one in the region” require special caution. They presuppose that the business can prove a clear and lasting lead. Stricter rules have applied to environmental claims since September 27, 2026, as the article on greenwashing describes.

Where numbers work, and where they do harm

Numbers help where a reader wants to quickly judge whether a business is experienced and stable enough: on the home page, the about us page and in proposals. They do harm when they take the place of evidence or when they promise more than the service delivers.

For the about us page, the rule from the article on the about us page applies: every praising adjective is replaced by a verifiable detail. Numbers are often the best choice for this, but only if they pass the number test.

What follows beyond the individual case

Numbers are a form of accountability. A business that says what it counted, when and how, makes itself verifiable. It may forgo the largest possible number but gains one that holds when a buyer asks questions. This is the value at the core of good financial statements too: a figure is worth as much as the transparency of how it came about. Whoever embellishes with numbers may impress in the short term. Whoever explains with numbers builds trust that does not dissolve even when a customer does the math.

The limits lie in the evidence. The experiments concern consumers and students, not buyers, and the finding on disappointment comes from film ratings. The worked example is freely chosen. The models are robust: what others do is considered right, and a signal convinces when it is hard to fake.

Conclusion and next steps

We set out to ask
numbers such as the customer count, projects or years in business work as proof of trust, and how a business states them honestly. They work when the reader sees what was counted, when and how, and when the number says something about their case. Precise numbers seem more credible than round ones, so they may only be as precise as what was counted. Large numbers without explanation are claims in digits, and they raise expectations against which the business is later measured. Stating them honestly means disclosing unit, period, counting method and comparability and supporting the number with names and cases.

• Collect all numbers on the website, in proposals and in presentations.
• Check every number with the four questions of the number test and file the basis with a reference date.
• Switch the customer count to active customers in the last twelve months, or state the counting method.
• Mark round numbers that are estimates as estimates.
• Have leadership claims such as “market leader” reviewed, or drop them.