Why more info lifted VidaXL's account creation nearly 40%

The registration page was leaking. VidaXL's B2B buyers were getting to the signup step and leaving before they created an account, and the analytics could show exactly where it happened. What the analytics could not show was why.

Sapna Kashyap, CRO Lead at VidaXL, did not start by cutting fields. She started by asking. A survey of the users who dropped off came back with a clear pattern: before the final signup, buyers wanted to understand what they were getting. The benefits, the service, the technicalities of the subscription. Many of them had never seen that information at all.

The test that followed became the biggest winner of her career, and it ran directly against one of the most common instincts in conversion work.

VidaXL is an e-commerce company operating in more than 30 countries, with a B2B segment alongside its consumer business. Sapna runs its experimentation program, has handled more than 155 experiments and an annual roadmap, and owns the quality and governance of everything that goes live. Her story on The Experimentation Edge covers a surprising win, a humbling loss, and the unglamorous design work that makes both of them trustworthy.

The thesis: good experiments are decided long before launch, by the questions a team asks and the evidence it gathers first.

A company that tests every decision

At VidaXL, experimentation is not a CRO side project. Product teams, development teams and business stakeholders all route their ideas through it.

"If they have some kind of idea, any different business decision they wanted to take, they just first want to validate from the experiment and the A/B test," Sapna said.

That culture has an executive sponsor. The webshop manager is data driven and invested in the program, and wants to see the company keep growing toward experimentation maturity. Sapna now leads a maturity program built around three things: the mindset, the process and the tools. Part of it is teaching. Part of it is improving communication with stakeholders so that new people understand why the company tests before it implements.

The lesson from this beat: an experimentation culture becomes real when the business, not only the CRO team, asks for the test.

The 38.7% winner

The B2B flow worked like this. A buyer lands on the site, checks the pricing section for service details and what the subscription includes, then moves to registration. That was the intended path. The data showed that a lot of buyers skipped it and arrived directly on the registration page, without ever seeing the subscription information.

The objective was to increase account creation. The survey pointed to the missing context. So the hypothesis was specific: if the registration flow featured more subscription information, more accounts would be created.

"It led to 38.7% of the uplift," Sapna said. "The learning was that users wanted to be confident for the last time."

That number is worth pausing on. Most conversion playbooks treat every extra line on a signup page as a cost. The team shared that worry going in. Sapna said the assumption itself was that it would add cognitive overload. The design answer was to add the information in a way that was not too much, but still served the purpose.

For a B2B buyer about to commit a business to a subscription, the missing details were not decoration. They were the reason to say yes. Removing uncertainty mattered more than removing words.

The lesson: when users leave a page, find out what question they could not answer before deciding what to cut.

The payment method everyone was sure about

Not every test at VidaXL confirms what people expect, and Sapna is candid about the one that did the opposite.

The company's history said that localizing country by country worked. So when the team tested a localized payment method across its markets, product, business and CRO all believed it would win. The historical data seemed to say so.

It did not. The experiment went negative. Buyers kept choosing credit cards and PayPal over the local option.

The difficult part came next. A loss against that much conviction does not read as a result. It reads as a broken dashboard. Helping stakeholders accept that the payment method should not ship was, in Sapna's words, a really difficult nut to crack.

The team cross-verified every source it could: Stripe data, other sources of truth, all of it reconciled before the result was shared. Then they walked the business through the evidence until it held.

"Whatever we have learned is that not every time something we localize is going to win," Sapna said.

In the end the loss did something a win could not. Sapna called it an awakening moment for everyone, a reminder of why the company validates before it implements.

The lesson: the tests that challenge your strongest priors are the ones that protect the business, and the ones that build trust in the program.

The 70% that happens before launch

Both stories depend on what happens before an experiment ever goes live, and that is where Sapna's team has changed the most.

It starts with research. Even when an idea comes from a product manager, the data and the evidence have to be there. Then comes the objective, the problem statement and the intent behind the hypothesis, all agreed with the stakeholder.

Guardrails are part of that conversation. If a change lifts conversion, there is a chance it lifts the return rate too. The team decides up front which metrics must not move in the wrong direction, so the primary metric can be pursued without hurting the business somewhere else.

Segmentation is planned rather than discovered. VidaXL segments by traffic source, such as Google Ads versus direct traffic, and by device. A change that wins on desktop is not assumed to win on mobile. If it only works for one, it ships for one. If it works for neither, it ships for none.

Data quality gets checked too, because the analysis depends on it. When the preparation is right, Sapna says, the analysis at the end is the smaller part. "It's like 70% of the work for an experiment is going to be done before it goes live."

When a test does lose, the process keeps going. The team looks at more data points and behavioral analysis to understand why users did not respond, then iterates and tests it a different way. In a lot of cases, the next version is the one that wins.

After five to seven years of testing, the biggest shift she describes is cultural. Most experiments are not going to win, she said, but each one is a learning, and that is the reason to test.

The lesson: a program matures when it stops measuring itself by wins and starts measuring the quality of what it learns.

The pattern

Four moves from VidaXL's program that any team can borrow.

1. Ask why before you cut. Pair the analytics that show where users leave with a survey that shows why, then write the hypothesis against the real objection.
2. Treat friction and reassurance differently. Extra information that answers the buyer's last question can lift conversion, as it did by 38.7% here.
3. Test your strongest priors and verify the surprises. When a sure win loses, reconcile the data across sources before you share it, then let the evidence carry the argument.
4. Do 70% of the work before launch. Research, objectives, guardrail metrics, device and traffic segments and data quality are decided up front, so the analysis at the end is short and trusted.

Full conversation with Sapna Kashyap, CRO Lead at VidaXL, on The Experimentation Edge.

🎧 Listen to the full episode →