There's a cost almost nobody audits until it hurts: the cloud.

It starts cheap. A small plan, a few gigabytes, light traffic. But the software grows, the data piles up, and nobody looks at the bill again until it spikes. By then you're already paying for storage you don't use, poorly written queries that run thousands of times a day, and files that were never compressed.

Industry studies estimate that close to a third of cloud spend is wasted. Not out of bad faith: out of neglect. Servers oversized "just in case", images that weigh ten times what they should, transfer charges nobody saw coming. Each one sounds small. Together they make a budget grow exponentially, not linearly.

The numbers behind a single detail: photos

Take something as simple as the photos uploaded by the companies we work with.

Recently we recompressed 24,197 photos across two of these projects and freed 85 GB of storage. Photos that weighed between 3 and 6 MB ended up at 50 to 150 KB. The same image, with no loss the eye can notice, just without the dead weight.

Now project it forward. A system that takes in 100,000 photos a month piles up 1.2 million in a year:

Unoptimized (~4 MB each): 4.8 TB a year.
Optimized (~150 KB each): 180 GB a year.

That's a difference of more than 4.5 TB a year, in a single file type. That's storage you pay for every month, forever, because you never touched a detail that gets solved once.

Our approach: review every detail

Before adding infrastructure, we optimize what already exists.

That means getting into the code and asking ourselves uncomfortable questions: does this query need to pull 10,000 records or just 20? Does this image have to weigh 4 MB, or can we compress it to 150 KB without anyone noticing? Does this process run every minute, or is every hour enough? Are we storing data nobody queries anymore?

We optimize two things at once: the code and the budget. Efficient code isn't just faster for the people who use it every day, it's also cheaper to run. Every query you avoid, every file you compress, every process you tune translates into a smaller bill at the end of the month.

Why we build on Supabase

We make much of this work possible with Supabase, and that's no accident.

It's real Postgres. Open, standard SQL. It doesn't lock you into proprietary technology: if one day you want to move, your data is yours.
Everything in one place. Database, authentication, storage, real-time and server functions. Fewer loose pieces means less surface where money leaks.
Predictable pricing. You know what you pay and why. No surprises in the fine print.
Security from the ground up. Row-level access control is built in, so each company protects its data without patches on top.

We love working with Supabase because it lets us do what we do best: tune every detail without fighting the tool.

What we're working on now

Right now we're doing exactly this with several of the companies we work with, among them Marcas Puebla, Geometría Vehicular and Servicios en su Compañía. Reviewing their code, their storage and their infrastructure so that every peso they invest in the cloud works for the business, not against it.

An honest question

Is your cloud storage optimized? Or have you never checked?

If you don't know the answer, you're probably overpaying. We run an analysis of your infrastructure and tell you, with numbers, where you're spending too much and how much you can recover. No commitment.

Write to us. We'll review your cloud before the bill forces you to.