1.4M views, under 2% signups. Adding a question doubled it.
The client is the biggest plant-based supplement brand in Mexico. Their signup popup was getting serious traffic. Over 1.4 million views in eight months. It was converting under 2%.
We rebuilt it in April 2026. By July the submit rate was running between 4% and 5%, and it held there.
The change that did most of the work is the opposite of the advice you have probably read.
Why your popup gets traffic but no signups
Most popups fail for one of three reasons. It fires before the visitor knows what your store sells. It asks for something the visitor has no reason to give yet. Or it offers a discount on a catalogue the visitor has not understood.
This brand had all three.
Their traffic was not the problem. 1.4 million people saw the form. The form was the problem.
What a good popup conversion rate actually looks like
Klaviyo's benchmark tool rated their 2.68% as "Fair." That number is the average across the whole period, including the months before we touched it.
Rough working range for ecommerce:
Under 2%: something is broken
2% to 3%: functional, not optimised
4% to 6%: well built
Above 6%: strong offer plus strong targeting
They moved from roughly 1.5% to a sustained 4% to 5%. On 1.4 million views that difference is tens of thousands of profiles.
What we found when we looked at their old popup
The old form asked for an email address and a birthday, on one screen, in exchange for a mystery discount.
Two problems.
The birthday field is a typing task. You are asking someone to enter three numbers before they have decided they want anything from you.
And the offer was disconnected from the catalogue. This brand sells a wide range of SKUs across protein, vitamins, and creatine. A first-time visitor lands on a homepage with a lot of products and no idea where to start. A discount does not help someone who does not yet know what they want.
The change that everyone gets backwards
Standard advice is to strip fields. Fewer fields, less friction, higher submit rate.
We added a question.
The first screen of the new popup does not ask for an email. It asks:
Tienes un Regalo Sorpresa
Para obtenerlo dinos ¿Qué te interesa más?
With six tappable answers:
Vitaminas diarias
Proteína diaria
Proteína para mujer
Proteína deportiva
Creatina
Todavía no lo sé, sigo decidiendo
The email request comes after.
Three things happen here.
The first interaction is a tap, not typing. Tapping a button is close to zero effort.
The question is useful to the visitor. Someone landing on a store with a large catalogue is confused. This question sorts them. It reads as help, not as data collection.
And once someone has answered a question about themselves, they finish. Step two completion on the old form ran around 40%. On the new form it runs at 98% and has held there for months.
That last number is the one to pay attention to. The submit rate did not double because more people started the form. It doubled because the people who started it finished.
The part your competitors miss: every answer is now a segment. Someone who taps "Proteína para mujer" gets a different welcome flow than someone who taps "Creatina." You are not just collecting emails. You are collecting emails that already told you what to sell them. This is zero-party data, and a popup is the cheapest place to collect it.
The "Todavía no lo sé" option matters too. It gives the undecided visitor a way through instead of a reason to close the popup.
Why timing the popup matters more than the offer
The old form fired almost immediately. That catches people who bounce, which is why the view count was so high and the submit rate so low.
We delayed it to 15 seconds. The form now shows to people who have stayed long enough to look at something.
View count dropped. Submit rate went up. That trade is almost always worth taking, because a popup shown to a bouncing visitor costs you nothing but a worse denominator.
What we tested that worked, but not as well
The mystery discount angle performed. "Tienes un Descuento Sorpresa" beat the original and we ran it for a while.
The gift plus question version beat it. So we discontinued the mystery discount.
Intrigue works. Intrigue plus usefulness works better. The question version gives the visitor something in the moment, before any email is exchanged.
What this looked like in revenue
Between 1 January and 16 August 2026, the signup form drove MX$8.5M in Klaviyo-attributed revenue across 38,000 submits, on a 5-day attribution window.
Two things to be straight about.
The form was off between 8 February and 18 March. You can see the flat line in the data. That gap is not a performance dip, it is a period with no form running.
And the June revenue peak overlaps with Hot Sale, Mexico's largest annual sales event. The rebuilt form was live and running at 4% to 5% through it. The event brought traffic. The form converted it.
The rate improvement is the durable result. It held through July and August with no sale event supporting it.
How to apply this to your own store
If your catalogue is large enough that a first-time visitor could be confused, do this:
Make the first screen a question, not a field. Ask what they are looking for. Tappable options only.
Give five or six options plus an "I'm still deciding" escape.
Ask for the email on screen two, after they have answered.
Delay the trigger to around 15 seconds. Fire it after the visitor has shown some intent.
Route each answer to its own welcome flow. If you skip this step you have collected data you are not using.
Check your step-level completion rate, not just your overall submit rate. Overall submit rate tells you something is wrong. Step-level tells you where.
If your popup is under 2% and getting real traffic, the fix is usually not a bigger discount.