Product and Service Descriptions That Answer Questions

Why so many descriptions answer no questions

“Our high-quality hall heating system impresses with innovative technology and maximum efficiency.” This sentence appears in similar form on thousands of websites. It answers not a single question a prospect has: for what size of hall? With which fuel? What does it cost to run per year? What must be in place on site? Anyone who does not find answers to such questions calls, writes an email or closes the page.

This guide therefore answers the question: What must a product or service description contain to answer a business customer’s questions, and how do you write it? We first clarify the terms and a model from the psychology of persuasion that explains how closely buyers read. Then we look at findings on descriptions, weigh two views on length and tone and build a description in seven answers.

Every adjective in a description is a question the text has not answered.

Terms: description, service page and diagnosticity

A product or service description is the text that explains a single offer: in the catalog, the online shop, the data sheet, the proposal or as the core section of a service page. It is smaller than the service page, which the article on the service page describes with the eight-question framework. There it is above all the answer to the questions of what is included and how it works. Because the same description is reused in many places, it pays to write it carefully once.

One term helps distinguish good from poor descriptions: diagnosticity. A piece of information is diagnostic when it helps the reader make a decision, that is, distinguish between offers or recognize whether an offer suits them. “High-quality” is not diagnostic, because every provider writes it. “For halls of 800 to 3,000 square meters, ceiling height from six meters” is diagnostic, because it allows a decision.

Drawing on information economics, the article on the service page also distinguishes three kinds of attributes: search attributes can be checked before purchase, experience attributes only afterward, credence attributes often never with certainty. A description can state search attributes directly and translate the other two into verifiable details, such as guarantees, standards or experience values from projects.

The thinking model: two routes to persuasion

In 1986 Richard Petty and John Cacioppo presented the elaboration likelihood model. It describes two routes by which people are persuaded by a message. On the central route they examine the arguments carefully and change their minds only if the arguments hold. On the peripheral route they rely on cues at the margin, such as an attractive picture, a well-known brand or simply the length of a text. Which route is taken depends on two conditions: whether the reader is motivated, meaning the topic matters to them, and whether they are able to judge the arguments.

For business customers the situation is usually clear. A buyer procuring a system for 40,000 euros is motivated, because money and their reputation in the company are at stake. An engineer reading the technical data is able to judge them. Both read via the central route. For them, arguments count, and promotional language does not merely work weakly but looks suspicious, because it exposes the lack of arguments.

What studies show about descriptions

Sollfrank, Feldner and Kim (2026) studied how descriptions work when shopping via voice assistant, that is, entirely without images. In an online experiment with 391 participants from the United States, these heard descriptions of two household products from fictitious brands, either rich in vivid imagery or factually neutral, and with five or ten attributes. Vivid descriptions that let the listener picture something increased brand stimulation and, through it, attitude, purchase intention and actual choice, and listeners remembered the brand better. The authors deliberately stayed below twelve attributes, because earlier research had found overload from that number on. The study concerns inexpensive consumer goods and voice assistants; its lesson for B2B is limited but clear: concrete imaginability helps, and the number of attributes has a limit.

Wang, Ru and Ye (2026) show a second side in four online experiments. When sellers on second-hand platforms stated that their description had been generated by artificial intelligence, purchase intention fell. The reason lay in perceived diagnosticity: readers considered the description less informative and trusted the seller less. The effect was stronger for experience goods than for search goods and weaker when the seller had reviewed the text. It did not appear in markets for new goods.

Both findings fit the model. Readers ask themselves whether a description helps them decide. Concrete, imaginable details help. Details whose origin or care is doubtful help less, and for services that can only be judged after purchase, this doubt weighs more heavily.

Short or detailed, factual or vivid? Two views

The first view calls for concise descriptions. Nobody reads long texts online, visitors skim, and every additional attribute dilutes the message. This is supported by the overload threshold that Sollfrank and colleagues mention, and by the experience that nobody reads descriptions with twenty bullet points to the end.

The second view calls for complete descriptions. On the central route, the reader wants to know everything that affects their decision. If a detail is missing, they have to ask or compare with a provider who states it. A concise description that omits the one decisive limit is not short but incomplete.

I consider an arrangement on two levels to be sound: concise and vivid at the top, complete and factual below. The first two to three sentences say, in words one can picture, what the offer is for and for whom. Below them follow the remaining answers in a fixed order, as a list or table that can be looked up directly. That way the skimmer gets their answer, and the close reader finds theirs.

The description in seven answers

The model shows what a description must achieve: answer the questions a reader asks on the central route, in the order in which they ask them. I call the structure the description in seven answers.

1. What for?

   The problem or use in one sentence, in the customer’s words: “When the hall heating fails in winter, order picking comes to a halt.”

2. For whom, and for whom not?

   Size, industry, technical prerequisites. The lower or upper limit belongs here; it is one of the most diagnostic details.

3. What exactly is included?

   Scope, dimensions, quantities, service steps. Concrete and countable.

4. What is not included?

   Exclusions and extra costs. This line prevents more trouble than any other.

5. What must the customer contribute?

   Access, documents, contact person, preparation on site.

6. What comes out?

   The result, where possible with a number and reference value, and evidence: inspection report, standard, reference.

7. What happens next?

   Delivery time, price or price range, next step with the same wording as everywhere on the website.

For the sixth answer, the article on expressing customer benefit in euros helps. For language, what the article on B2B website copy justifies with processing fluency applies: simple words seem more competent than complicated ones; customers’ technical terms belong in, in-house jargon does not.

How to tell whether a description works

A description works when prospects ask fewer obvious questions and send more suitable inquiries. Both can be counted. The questions come from sales, the inquiries from the measurement chain.

A second touchstone is search queries. Anyone who knows the terms customers use to search for a service can check whether they appear in the description. How to find the right page for the intent behind a search is described in the article on understanding search intent. Questions too extensive for the description belong on the FAQ page, as the article on the FAQ page shows.

The bigger picture

A good description is a form of transparency. It discloses what a customer gets, what not and what they must do themselves. At first glance this costs inquiries, because unsuitable prospects drop out earlier. But it saves conversations, change orders and disappointments, and it attracts exactly the readers who decide via the central route. How this leads specifically to better inquiries is described in the article on fewer but better inquiries.

The limits lie in the evidence. The experiments cited concern consumer goods, voice assistants and second-hand goods, not purchasing by business customers. The numbers in the worked example are freely chosen. The model is robust: those who are motivated and able examine arguments, and a description is only as good as its answers.

Conclusion and what you can do this week

The question at the start was
a product or service description must contain to answer a business customer’s questions. It must give seven answers: what the offer is for, for whom and for whom not, what is included, what is not, what the customer must contribute, what comes out and what happens next. Because buyers and specialists read via the central route, concrete, diagnostic details count, and promotional adjectives tend to do harm. At the top stand two to three vivid sentences, below them the complete answers in a fixed order. A description is finished when, after reading it, a customer only has questions about their own case.

• Print the description of your most important service and mark every adjective without evidence.
• Check which of the seven answers are missing, especially “for whom not” and “what is not included.”
• Ask sales for the three most frequent follow-up questions about this service.
• Rewrite the first two sentences so that a customer can picture the use.
• Use the revised version identically in the proposal, data sheet and website.