Why unclear messaging costs you, and how to test for clarity
The commonest reason a page underperforms is that visitors can't tell what's on offer, and the team that wrote it is the last to notice.
Picture a results meeting where the agency is presenting a test on the hero image. Lifestyle photo against product shot, two weeks of traffic, a flat result. Nobody in the room has asked the question that would have mattered more, which is whether a stranger landing on that page could say what the company sells.
The team usually can’t answer it, because they’re the worst people to judge. They’ve read the headline hundreds of times, they know what the product does, and they fill in every gap without noticing there’s a gap. Visitors have none of that context. They arrive from an ad or a search result, give the page a few seconds, and leave if it doesn’t make sense to them. Nielsen Norman Group’s homepage guidance puts it plainly: don’t assume visitors already know your brand.
Unclear messaging is one of the most common conversion problems, and one of the least tested, because it doesn’t feel like a problem from the inside. It also sits underneath everything else in this track. Social proof, urgency and reassurance all assume the visitor has understood the offer. If they haven’t, a test of any of those is measuring confusion with a bit of persuasion on top.
What clear looks like
Section titled “What clear looks like”A clear page lets a visitor answer three questions without effort. What is this? Who is it for? Why would I pick it over the alternatives, including doing nothing? That third question is the value proposition. It’s the reason to choose you, stated in terms the customer cares about, and it should be specific enough that a competitor couldn’t paste it onto their own homepage unchanged.
The usual failures are easy to spot once you look. A headline that’s a brand slogan rather than a description. Copy written in the company’s internal language, so the product name is explained using other product names. A hero section that says “welcome” and nothing else. A generic stock photo that suggests a different business. Each of these makes sense to someone who already knows the answer.
How to test it
Section titled “How to test it”The cheapest check is a five-second test. Show the page to someone in your target audience for a few seconds, take it away, and ask what the company does, who it’s for and what they’d do next. You’re looking for whether their answer matches what you meant, in their words. If several people describe a different business, you’ve found your first test, and it cost you an afternoon.
Five seconds is a convention with little evidence behind the exact number. A Smashing Magazine review of the research behind the method notes that the figure rests mostly on practitioner habit, that busier pages may need closer to ten seconds, and that the method suits single-purpose pages better than ones trying to do several jobs at once. Use it to catch gross misunderstanding. Don’t use a handful of answers to pick between two headlines that differ by a word, because the sample can’t carry that.
Message testing goes a step further. You write three or four versions of the core proposition, each built around a different reason to buy, and show them to people from your audience. Ask what each one means, which reason matters most to them, and what’s missing. The language for these versions should come from customers, from reviews, interviews and support tickets, rather than from the brand guidelines. Then take the finalists into a live A/B test, because what people say they prefer in a survey and what they do on the page often drift apart.
A good clarity test idea names a specific misunderstanding you’ve seen in research. “Visitors think we’re a marketplace when we’re a service, so the headline will say who we serve” is a test you can learn from. “Make the headline punchier” isn’t, because whatever happens you won’t know why.
How value propositions get overclaimed
Section titled “How value propositions get overclaimed”Value propositions drift upwards over time. “Helps you plan” becomes “effortless planning”, then “the UK’s leading planning platform”, then “set up in minutes” when most customers take a week. Each step feels small, and each one gets approved because nobody wants to be the person who made the copy less exciting.
There are two problems with this. The first is regulatory. In the UK the CAP Code requires marketers to hold documentary evidence for any claim a consumer would take as objective, and the ASA treats an unqualified superlative like “best” as a comparison against every competing product. “Leading” and “fastest” need evidence on file before they go live.
The second problem is that testing tends to reward the overclaim. The boldest promise often wins the click and the sign-up, because it’s the most attractive thing on offer. The cost shows up later, in refunds, cancellations, support calls and customers who feel misled, none of which your primary metric will see. This is where guardrail metrics earn their place. If a variant makes a stronger claim, measure what happens after the sale as well as the sale itself.
What to ask an agency
Section titled “What to ask an agency”When a messaging test lands on your desk, a few questions sort the good ones from the guesses. What misunderstanding did the research turn up, and how many people showed it? Who took the five-second test, your target audience or the team’s colleagues? Where did the wording come from? Does any variant promise more than the product delivers, and could we substantiate it if the ASA asked? Which guardrails will tell us if the winner is selling something we don’t have?
A useful habit in results meetings is to ask, before anyone reads the numbers, whether the variant made the offer easier to understand or just louder. If it was louder, check the guardrails before anyone celebrates.
Put it to work
Show your most important landing page to five people in your target audience for a few seconds each, then ask them what the company does, who it's for and what they'd do next. Write their answers down in their own words.
There is no right answer. Write yours first, then compare.
Sources
Shoshin is written by Another Web is Possible. Lessons are licensed CC BY-SA 4.0.Our promise.