PostgreSQL
The one your developer asks for. How to tell whether that's a reason or a preference.
PostgreSQL is the one your developer asks for.
Unlike most of the database decisions that reach a boardroom, this one usually arrives as a request rather than a discovery. Someone technical wants to build the new system on Postgres, and they’re keen about it in a way they aren’t keen about the rest of the infrastructure. That enthusiasm is the thing that makes it hard to assess. You can’t tell from the outside whether you’re being given a considered recommendation or a preference with good PR.
The awkward part: they’re probably right. Postgres is the strongest general-purpose database available to you and has been for years. But a correct answer given for a bad reason still costs you money later, because nobody has thought about the parts that bite.
“It’s the most capable open source database”
You’ll be told PostgreSQL is the most capable open source database, and the obvious choice for anything new.
That’s accurate, but on its own it tells you nothing about whether it suits your system.
Nobody can take PostgreSQL away from you
Nobody is in a position to take it away from you, and this is the strongest thing about it. PostgreSQL has never been relicensed, and there is no company that could decide to. It’s developed by a distributed group of contributors and sponsors with no controlling commercial owner. Databases have had their licences changed under businesses that already depended on them. It cannot happen here, because there is no single owner who could make that decision. Few databases can say the same, and it rarely gets mentioned first.
The real cost is operating it, not building on it
Hiring is good rather than abundant, and this is the real cost. Any competent backend developer can work with Postgres. The smaller pool is people who have run it in production: handled a major version upgrade, tuned it under load, recovered one at three in the morning. Building on it is easy. Operating it is where the gap shows, and it’s the gap you’ll be paying for if your team is small.
Its strictness is the feature, not a flaw
It’s stricter than you’re used to. Postgres refuses data that other databases would accept and reshape. In development that reads as friction and slows people down. In production it’s the reason your numbers add up.
Why an untuned PostgreSQL gets blamed for being slow
An untuned PostgreSQL is slow, and then gets blamed for being slow. Its default configuration is deliberately conservative so it will start on almost any hardware. Nobody changing those defaults is the single most common reason a business concludes “we tried Postgres and it wasn’t fast”, when nobody had configured the database.
How to test your developer’s reasoning
- “Are we choosing this for a specific capability we’re going to use, or because it’s the better database in general?” Both are legitimate answers, but the first shows somebody has thought about your system.
- “Who here has run Postgres in production rather than developed against it?” If the answer is nobody, that’s not a veto, it’s a budget line.
- “What happens when a major version goes end of life, and whose job is that?”
- “Are we running default configuration?” Ask again in six months.
When PostgreSQL is the right choice for your business
Most new builds, especially when the data model is the hard part of the system. If your application’s main difficulty lives in relationships between things, in variable-shaped data, in reporting that has to aggregate across a lot of rows, or in search, Postgres handles all of it without bolting on a second system, and keeping everything in one database saves you more than any single feature would.
It’s also the right answer when you expect requirements to grow in directions you can’t currently name. Postgres has more headroom than most businesses ever use, and reaching a limit that forces a migration is rare enough that you can reasonably stop planning for it.
When PostgreSQL is the wrong tool
When you’re running WordPress or WooCommerce. Those platforms are built for MySQL, and putting them anywhere else means fighting the software for no benefit you’ll ever notice.
When nobody available to you can operate it. The best database in the world becomes the worst one the week it needs attention and there’s nobody who knows how. Either budget for that expertise or choose something your team can run.
When the system is simple. If the data is rectangular, the queries are predictable, and the hosting hands you something else for free, Postgres is a fine choice that isn’t buying you anything, and you’re paying a small operational cost for headroom you may never use.
It's my default, and I'd still make you justify it.
PostgreSQL is what I reach for on a new build, and I've never regretted it on a system where the data mattered. But I'd rather you ran MySQL well than Postgres badly, and I've seen the second one happen to businesses who took the recommendation without taking the operational side of it. So the question I'd put back to your developer isn't whether Postgres is better. It's who is going to look after it in eighteen months, and whether that person currently works for you.
Alexander De Sousa · Founder, Digital Royalty · LinkedIn
Compare it against the alternatives
The questions worth asking, answered honestly for PostgreSQL — line it up against any of the alternatives to check the call.
- What it actually is
- Relational database
- Licence
- PostgreSQL Licence — a permissive, BSD-style open source licence.
- Has the licence changed?
- No. The licence has never changed in the project's history, and there is no single owner in a position to change it. This is the strongest position of any database on this page.
- Who controls it
- The PostgreSQL Global Development Group — a distributed community of contributors and sponsoring companies, with no controlling commercial owner. Nobody can unilaterally relicense it or put a feature behind a paywall.
- Can you hire for it
- Good — readily available
- Paid support available
- Yes, from several independent vendors (EDB, Crunchy Data and others) and from every major cloud provider as a managed service. Because no single company owns the project, you can change support supplier without changing database.
Tell us what you need
A few quick questions, then a straight answer from a real person — usually within a few hours.
What happens next
- 1 We reply A real person reads your answers and gets back to you, usually within a few hours.
- 2 Questions or a short call If anything isn't clear, we ask by email or book a short call. Before any call we audit your website and make the quick fixes it finds, free.
- 3 Your fixed quote A clear, fixed price in writing, within 48 hours of any call. There is no obligation to go ahead.
Tell us what you're working on
Whether it's a new site, a platform, or a process that shouldn't be manual any more — we'll tell you honestly if we can help.