Quick tips
- Count the views before you change the product.
- Show it to one stranger and time five seconds.
- Change one thing a week, not four at once.
Three weeks ago you put it out. The photographs are good, the description says what the thing is, and the page has been visited. What has not happened is an order, and the tab with the logo file in it has been open since Tuesday.
Put the logo down. Zero sales is not a verdict on the work or on you, and it is not a mystery either. It is one of a small number of ordinary, fixable situations, and there is an order to check them in that stops you changing four things at once and learning nothing from any of them. Twenty minutes of counting will tell you more than a weekend of redesigning. That is the good news buried in an empty sales report. What you are looking at is data, and data is exactly the thing you did not have last month when you were guessing.
Why is nobody buying from my shop yet?
It is almost always one of four things, and they go in this order: too few people saw it, the wrong people saw it, the right people could not tell what it was in five seconds, or they understood it perfectly and the price did not match what they thought they were getting. Test them in that order, because the first one makes the other three impossible to measure.
- Volume. If forty people have seen the page, you have no information at all. Zero out of forty is a completely normal result for something that works.
- Audience. Plenty of views, all from the wrong place. A post that traveled among people who will never buy is traffic, not demand.
- Comprehension. People arrive, cannot tell what it is or who it is for, and leave. This is the most common real cause and the easiest one to fix.
- Price against value. They understood it, they wanted it, and at that number they said no. This is the last one to test and it is the first one everybody jumps to.
Notice what is not on the list. Your logo, your font, your colors, whether the name is perfect. Those are the things you can change at midnight without talking to a single human being, which is exactly why they get changed first and exactly why it rarely helps.
How do I tell a traffic problem from a product problem?
Count the views before you change anything. Under a few hundred visits the honest answer is that you do not have a product problem yet, you have a nobody knows problem, and the fix is telling more people rather than rebuilding the thing.
Run the arithmetic with an ordinary number. If two visitors in a hundred buy, then a hundred visits gets you two orders and fifty visits gets you one on a good week. Zero at fifty views is not a signal. It is the absence of one. Treating it as a signal is how people end up redesigning a page that was fine.
So write down two figures for last week: how many people saw it, and how many did anything at all, which might be an order but might also be a message, a save or a follow. If views are low, stop reading and go back to telling people directly. If views are healthy and nothing is happening, you have a real diagnostic in front of you and the next two sections are yours.
How do I run the five second test on my own page?
Show the page to one person who has never seen it, count five seconds, take it away, and ask them what it was, who it is for and roughly what it cost. Their answer is what your page actually says, whatever you meant it to say.
Jakob Nielsen's research on how long people stay on web pages found that the first ten seconds of a visit are the critical ones, and that the probability of leaving is highest right at the beginning while people triage fast. A page that has not made its case inside that window is usually gone. Five seconds is not an artificially cruel test. It is close to what your page really gets.
You do not need a crowd for this. The Nielsen Norman Group's long standing finding is that testing with five people surfaces around 85 percent of the usability problems in a design, while a single tester finds roughly 31 percent. So that is five friends, five separate five second looks, twenty minutes in total. Write down exactly what each person says rather than the gist. Where three of them say the same wrong thing, they have rewritten your first line for you.
Is my logo the problem?
Almost never. Trust gets built out of things that are far less fun to work on: clear pricing, real photographs, a visible way to reach a human, delivery and returns information, and evidence that other people have been here before.
The Nielsen Norman Group's work on credibility in web design names the factors plainly, and the ones that move people are design quality, telling them up front about costs and terms, keeping content current and accurate, and being connected to the rest of the web. A shop that shows shipping costs before checkout and lists a real address reads as safer than one with a nicer wordmark and no way to ask a question.
The price against value cause hides here too. Baymard Institute's running average across fifty studies puts documented online cart abandonment at 70.22 percent, and the reasons are dominated by things people meet at the last moment, extra costs and unclear terms among them. If people are reaching your checkout and stopping, the problem is often not the number on the product page. It is a number that showed up later.
How many things should I change at once?
One a week. Change four and you learn nothing, because whatever happens next you will not know which change caused it, and you will carry that confusion into every decision after it.
Pick the cause you are testing from the four above, make one change aimed at that cause, and set the date you will look again. A week is usually enough for a small shop to see a difference in comprehension. A change to where your traffic comes from needs two. A price change needs however long your normal buying cycle is, which is longer than you want it to be.
Write the change and the date in the same place. Then leave it alone. The discipline of not touching it is doing real work, because it is the whole difference between running an experiment and stirring.
Decide now what working looks like at week six
Before you close this page, write one sentence: at week six, this is working if a specific thing has happened. Ten orders, or four, or twenty saved items, or three conversations with strangers. Pick the number now while nobody is watching, because the number you invent later, on a bad Sunday, is always the one that makes you feel worst.
That sentence is what stops zero from being a verdict. It turns a feeling into a test with a date on it, and a test with a date is something you can pass, fail, learn from and run again. Six weeks from now you either have your answer or a much sharper question, and both of those are worth more than another logo.
Then go and make more people see it. Almost every version of this ends with the same instruction, and the reason it feels too simple is that it was never a strategy problem. Somebody has to be told, and you are the person who knows exactly who.
Sources
- Nielsen Norman Group, How Long Do Users Stay on Web Pages?
- Nielsen Norman Group, Why You Only Need to Test with 5 Users
- Nielsen Norman Group, Trustworthiness in Web Design: 4 Credibility Factors
- Baymard Institute, 50 Cart Abandonment Rate Statistics 2026