Case Study - WU Tests

Hey bubble fam!
I’m here to showcase my case-study of the WU.

Image below for the TLDR guys.

So, I have built a website that contains:
1/ 2 Workflows
2/ 2 buttons
3/ 1 form
4/ 1 popup

This is a very basic site, just for testing the cost.

So, I published my website and promoted it in several facebook groups just to create some live users that could really help with the tests.
The users did not know they were testing the website all they’ve done is filling the form.

I had 30~ real users which filled the form and the cost was 3500WU per that day.
I repeat, 30users filled the form, 3500WU per that day.

So let’s make some quick calculations!
3,500WU (with an average of 30 daily users) per day x 30 days = 105,000WU


Nothing to optimize really.

So please, let me know, why should I keep the use of bubble?
Why should I pursue with it?

Eilon Levy, Bubble dev



That looks very discouraging.

I have checked the numbers on my apps the situation looks grim. on one app , 150 DAU generate 1M WU per month. It makes the entire app totally unfeasible.

The most troubling part is that it seems bubble is charging WU for frontend workflows. Why does that make sense? frontend workflows dont create on bubble servers, don’t they?

I’m afraid if this pricing plan really comes to production many bubble app will be unfeasible.


Could be you have elements having data sources set as a search which incur realtime search WU. Try adding the :static operator to them.


:make static
This operation converts a list that relies on dynamic data, e.g., input’s values in search constraints, and converts it into a list with the current items. Once converted to static, the list will not change if underlying values change. Note that this works only within the context of workflows; it will not change the behavior of repeating groups. Caution: This is an advanced feature.

Sorry might not work as i mentioned it. What you can do to “optimize” is to have any searches saved to a state using a workflow. Then point any elements that need that retrieved data to that satate.

That way you won’t need to use elements that rely on realtime search and you can control when you need to retrieve data.

Eles disseram ao Renato Asse aqui do Brasil, que eles iriam melhorar isso, mas não acredito muito nesse papo, ao ponto que vai diminuir tanto assim.

Parece que o bubble quer recuperar o prejuízo tudo de uma vez.

Voltei a fazer minhas aulas de porque não tem como pagar esses preços.

Também tenho essa impressão.

Hey! You can have 900 people on your website per month for their cheapest new plan! $32/month.

That’s not too bad!

The most you’ll pay for a WU is $0.00018
The best you’ll get for a WU is $0.00008*

I’ve used 1.5M WU this month. ($180/month)

But I’ve used nearly 60X in the past. ($7,200/month)

Oh yeah - - It’s just me. 0 Customers. At < 10% capacity for the $29/month plan.


Problema é a aqui no Brasil que esse valor se multiplica por 5x. Temos governantes corruptos e com isso nossa moeda não vale nada em relação ao dólar.

1 Like

If that it is indeed the case, won’t the log show me the usage as “fetch data”?

As I’ve mentioned, I see the usage under “workflow”.

1 Like