New registration form: Minha Oi

User flow and usability issues: adjustments that have improved our conversion rate

The problem

Low performance of the registration form. The conversion rate was around 11%.

Understanding the problem

In order to better understand what was happening, we analyzed the user flow.

User Flow of Minha Oi

First user flow, registration form – Minha Oi

Apparently it was a simple 5 steps flow. However, when we look a little deeper at the conversion funnel, a couple things caught our attention:
1. The number of users abandoning the process in the course of the steps;
2. The unnecessary complexity of the flow.

Conversion rate showing how many users we had at each step

User flow in numbers, step by step, register form – Minha Oi

The conversion rate was 10,26% and we noticed that the large majority of the users were leaving the registration form in the first half of it (68% on the first screen and 51% on second). These screens were the part of the process that we call validation stage. This stage was making the flow more complex and we believed that was the reason users were abandoning the registration form.

Understanding the validation stage: talking with a stakeholder

So before any solution proposals, we decided to talk to the people responsible for including this stage in the flow to understand why it was there.

What is the validation stage?
– The validation stage is the part of the process where we ask the user to validate his identity as a client of our company and owner of, at least, one of our products .

For what purpose does it exists?
– Basically to prevent frauds.
It guarantees that only the actual owner of a CPF can manage his account and personal data in the client area.

How it works?
– At the beginning of the form the user is asked for his CPF number which is a document, like an ID number, that identifies every person in Brazil. So, to prevent frauds (people using other people’s CPF), we ask the user to validate his identity by associating a product that he owns (from our company) with his personal data in our database. There are a few methods for doing this but the most common is by token via SMS.

A few thoughts after the talk

We understood that the validation stage is process required for information security matters and it couldn’t be removed from the flow. However, a few thoughts came to our mind:

1. Is the registration form the right moment (or place) to apply it?
2. Aren’t we preventing users of being registered because of the complexity added by this?
3. Isn’t the validation stage making the cost of interaction higher for users that just want to get registered?

The hypothesis

Raising these and other questions, made us have an idea:

What if we put the validation stage outside the registration form and request it only when needed?

By doing this, we could reduce the barrier for users to register.
Now, once inside the client area, if the user try to do something that is considered more restrict, in matters of information security, we ask him to validate his identity.

New flow proposal

After discussing it better, between flows and screenflows, the idea was shaped and we created the new flow.

User flows and screenflows at the wall

Flows and screenflows to shape the new idea

New user flow of Minha Oi

Proposed user flow, registration form – Minha Oi

The proposal not only puts the validation stage out of the form, but also simplifies the steps that remained in the flow. Another modification is that the user e-mail now requires validation. This wasn’t asked before (in the old flow) but it was a business requirement for the new proposal. This way the company would have a valid digital communication channel with customers.

A brief dip in usability issues

During the flow analysis we noticed some usability gaps on the screens. But we didn’t knew how they were affecting users in the registration process. So before making any progress on this, we wanted to release the new flow, measure it and make a few adjustments if needed. This would give us an idea about how the new flow proposal was impacting the process.

After that, we would launch a usability test, with real users, analyze it, make the adjustments, test again and then release. However, our deadline was very short, we didn’t had enough time so we had to “fill two needs with one deed”.

One of a few usability issues found

Usability issue #10: On the left side, we have the desktop version with the new brand and visual identity (mainly pink). On the right side, we have the mobile version with the old brand and visual identity (purple).

Heuristic evaluation

To help us in the analysis we did a heuristic evaluation, according to 10 usability general principles of Nielsen and Molich, using a google sheet. (You can see it here).
We identified 14 issues classified among average, high and serious.
We run a guerrilla usability test with a few people from other squads (that’s what we could do with the time we had), did the required adjustments on the interface and then released.

Heuristic evaluation sheet containing 14 usability issues

Heuristic evaluation sheet containing 14 usability issues. Click to see the sheet.

The outcome

Next, you’ll see the final interface with the usability fixes made and a comparison between flows and funnels of what we had and what we achieved:

Desktop and mobile version of Minha Oi registration form

Usability issue #10 fix: Both devices are displaying the same interface, now responsive


Comparing conversion on user flows

Old and new user flow, step by step compared.

Comparing conversion on funnels

Old and new conversion funnels compared.

As described before, the old flow wasn’t performing very well and the conversion rate was 10,67%. Now the new flow performed better with conversion rate of 23,23% (more than double).


We saw a better performance of the registration form, comparing both version. Of course, there is always room for improvements and that is what we are doing right now, thinking on how we can do better next time.

Nevertheless the greatest accomplishment we had was that this experiment showed us that whenever we build products (systems, forms, websites etc.) in a user centered culture, it will be good for everybody. It will be good for users, because we reduce the barriers and effort by making things easier and will be also good for companies in achieving the goals.

You can see the final release here.