Analytics and audits
Optimisation
of conversion rates (CRO)
Before you pay for more traffic, check how many of your current visitors get lost on the way to contacting you. We find the points where they give up, remove the obstacles and measure the effect of every change.
What conversion optimisation involves
A conversion is the moment a visitor does what they came to the site for: sends a form, calls or places an order. Optimisation means removing the obstacles between arriving on the site and that moment.
We base changes on data, not gut feeling, and first we check that the measurement is reliable. A common mistake is rebuilding the site to the owner's taste when the data points to a completely different place where people give up.
Where we learn what to improve
- The analytics funnel: where the biggest drop between steps occurs
- Session recordings: what people do before they leave
- Click and scroll maps: what they see and what they never reach
- Internal search: what they looked for and whether they found it
- Forms: at which field users stop filling them in
- Conversations with customer service: what customers ask when they could not find the answer on the site
The last source is easy to overlook. If customer service answers the same question every week, the answer is missing from a visible place on the site. Some people with the same question give up without calling.
What we improve most often
The obstacles below can usually be removed without rebuilding the site.
- A form with fields that could be collected in the first conversation
- Phone number hidden in the footer instead of the header
- Contact button visible only after a long scroll
- No information on lead time for an urgent service
- Consent banner covering the content on mobile
Testing: one change at a time
We measure every change before and after. With higher traffic we run an A/B test: some visitors see version A, others see version B.
With lower traffic we work sequentially: we change one thing at a time and compare the period before the change with the period after it, so the effect can be attributed to its cause.
How much you can realistically gain
We do not promise a specific increase in sales, because it depends on the starting point. On a site with obvious obstacles, such as a hidden phone number or a form that loses data, the first fixes can make a clear difference. On a polished site progress is slower and requires testing with high traffic.
That is why we first check the starting point, and then tell you what can be improved.
How much traffic a test needs to make sense
The result of an A/B test depends on the amount of data. With few visits, the difference between versions is down to chance and may reverse the following month. We say this at the start, because a test on too small a sample does not give a reliable result.
What counts is the number of target events in each variant, not just the number of visits. A shop with a few orders a day waits weeks for a test result. A service company with a dozen or so enquiries a month usually cannot collect enough data before the season or the offer changes, and in that case we do not run tests.
With low traffic there is still plenty to improve. Instead of testing button variants, we remove obstacles that are known to do harm: a missing phone number, a form with eight fields, slow loading on mobile.
The number of target events on the tested page over recent months. It determines whether the test will give a result at all and how long it will take.
The contact form: fields, errors, confirmation
In service companies the form is often the weakest point of the journey. It is created when the site is built and rarely revisited, yet this is exactly where some people give up.
Every field should have a reason. Questions about company name, tax ID (NIP), job title and “how did you hear about us” look harmless, but they make the form longer, and you can easily collect this information in the first conversation.
The second issue is error messages. A form that clears the entered data after you click “send” and shows a generic “an error occurred” loses the customer. We check every error path, including an unticked consent box at the bottom of the form that cannot be seen on a phone.
- Only the fields needed at first contact
- Live validation, not only after sending
- Clear error messages next to the specific field
- Entered data kept after a failed submission
- Visible confirmation after sending, not just a redirect to another page
- Automatic reply to the customer's inbox
Trust signals: what stops people from sending an enquiry
Besides convenience, the risk the visitor perceives matters. To them, sending an enquiry may mean that a salesperson will call straight away, while they have not made up their mind yet. Specific information on the site can reduce that concern.
We check whether the site says the first consultation is free, whether it shows the photo and name of the person who will reply, a real response time instead of “we will be in touch soon” and signed reviews instead of anonymous quotes.
In shops the same mechanism applies to returns and delivery. A visible returns policy and a realistic dispatch time reduce the risk of buying before the customer adds a product to the basket.
Separately on mobile, because things look different there
A large share of traffic comes from phones, while decisions about how the site looks are usually made in front of a large monitor. The gap only shows when you go through the whole journey on a phone, on a slower connection and in bright light.
Things that only show up there: a contact button hidden by the consent banner, a phone number that cannot be tapped, an email field with a keyboard lacking the @ sign, a table that scrolls sideways and a section important to the decision placed several screens down.
What we approach with caution
Some solutions recommended as ways to raise conversion can do harm. A discount pop-up may increase newsletter sign-ups and at the same time get in the way of buying, because it covers the product at the moment of decision. An “only 2 left” counter shown on every visit loses credibility and undermines trust in the shop.
The same goes for live chat on the site. It helps if someone on the other end replies straight away. If the answer arrives two days later, it is better to switch it off, because a broken promise of quick contact puts customers off.
Working together
How we work together, step by step
Measurement and diagnosis
We check that the data is reliable and find the biggest drop in the journey.
Quick fixes
We implement the changes with the best ratio of effect to effort.
Testing and development
We test further hypotheses, measuring each change separately.
Questions
Frequently asked questions
Diagnosis, yes; A/B tests, no. With low traffic we implement fixes that are known to help, and we often recommend working on visibility first.
Usually not. Most changes are fixes within the existing layout: a shorter form, a moved button, added information on lead time.
An audit is a one-off diagnosis. Conversion optimisation is a cycle: change, measurement, conclusion, next change. See also online shop audit.
We price it individually. The cost depends on the number of pages and journeys to analyse, the data available, the number of changes to implement and whether we run A/B tests. The initial assessment is free.
Related services
What else may be useful
Your city
Conversion optimisation in your city
For each of these cities we have a separate page describing the local market and how we work with companies there.
Let's find where you are losing enquiries
Give us your website address and approximate monthly traffic. We will send back an initial assessment and suggestions for the first fixes.