What a Website Comprehension Test Taught Us About Explaining Friendship

A product can be meaningful to its founders and still sound vague to someone seeing its website for the first time. We learned that while testing how people understood g3t's earlier Paris marketing site. We were trying to communicate an unfamiliar idea: not just discovering an activity, but creating better conditions for seeing the same people again. The difference felt obvious internally. It was not always obvious on the screen.

When the visitor has to assemble the product

Our earlier website put the idea of repetition across several sections: why meeting again matters, how hangouts work, who the community is for and what someone would pay. A person could understand individual sentences yet leave with an incomplete model. The service might look like a local events calendar, a hobby club, a generic friend finder or a paid membership. Those categories are not interchangeable. Each leads a visitor to expect different things after sign-up.

A September 2026 qualitative website review documented this confusion among a small, nonrepresentative set of participants. Because the sessions included different participant backgrounds and some moderated explanations, the findings are directional. They should not be advertised as statistical proof that a certain share of Paris residents misunderstood the page. What interested us was the kind of mistake, not a headline percentage.

The lesson was about sequence

The most useful explanation should answer four questions in the order a new person needs them. Who is this for? What actually happens? Why does it differ from one-off activity discovery? What is the commitment? A paragraph about the psychology of friendship cannot replace a clear description of what someone receives. A list of possible activities cannot demonstrate that a service creates repeated encounters. Copy has to connect the mechanism to the benefit without forcing the reader to infer the product.

Why this gets harder during a product change

By October, our team was also evaluating a different commercial and hosting approach. An old pricing screen or feature description could be accurate for the previous product version and misleading about an upcoming one. That is a special danger for a startup publishing educational content: a blog article can make promises far beyond the source it copied. A current live site is evidence of what is deployed; a prototype and strategy memo are evidence of intended direction. They should not be casually blended.

What we changed in how we evaluate copy

Instead of only asking whether the headline sounds inspiring, we want to know what a visitor can accurately repeat after reading it. Can they explain what happens after joining? Do they know whether a feature is currently available? Do they understand what is optional and what requires commitment? A copy test should invite people to describe the product in their own words before the founder explains it. Otherwise the facilitator may accidentally supply the missing clarity and overestimate the design.

Repetition should not be repetition in the page

There is an irony in a brand focused on repeat encounters: repeating the same promise in five sections can make a page feel slow. Different screens should add different information. The first establishes the value, the next makes the mechanism tangible, and the conversion step explains what the person is being asked to do. Better communication is not louder insistence that friendship takes time. It is a more coherent answer to an unfamiliar practical offer.

A lesson that remains unfinished

The earlier website review gave us hypotheses for improving comprehension. It did not prove that any redesigned page would increase paying membership. That requires real testing with the current product and a suitable audience. For the blog, the lesson is equally important: describe friendship research carefully, and describe g3t's available services only after they can be verified. The reader should never have to guess which claims are educational and which are product promises.

A communication test should preserve the visitor’s first impression

When interviewing someone about an unfamiliar service, explanations from the interviewer can accidentally supply the understanding being tested. A participant may ultimately describe the product correctly because a researcher clarified it halfway through the call, not because the website communicated it. This is especially important when a product combines activities, small-group gatherings and an ambition to help people meet again. Each component is easy to understand alone; their connection may not be.

A stronger test asks for an unaided description before any demonstration, records the point at which the visitor becomes confused and separates what they inferred from what they were told. It should also include people with different backgrounds rather than only those used to reviewing digital products. This approach tells us whether our words are understandable, not whether someone will pay or form a friendship. Our earlier small-sample review helped surface a communications problem, but it cannot substitute for measuring live customer behavior.

Future testing can help distinguish three separate questions. Comprehension asks whether visitors can correctly explain what g3t does. Intent asks whether they want to try it and why. Behavior asks whether they actually attend, reconnect or purchase. A persuasive prototype can do well on the first two and still fail on the third. We should not combine all three into a single 'people love it' claim. Nor should one qualitative participant represent every newcomer to Paris. The public lesson for anyone building an unfamiliar service is to hear the first interpretation before offering the carefully crafted explanation that the team wishes visitors had understood.

Sources and verification

Self-disclosure and liking: a meta-analytic review — Collins and Miller (1994)

How many hours does it take to make a friend? — Hall

References are provided to support the specific research or public context discussed above. Correlation does not establish causation, and no meeting or app can guarantee friendship.

For another starting point in Paris,explore g3t’s current Paris page. Check the live details there before making plans; this article does not imply that an upcoming feature has launched.