Here's a question worth asking before you commit a few years of business records to any receipts app: if you stopped paying tomorrow, what would you actually keep?
For most expense apps, the honest answer is "a CSV, maybe, if their export works." Your receipts — the actual images, the thing a tax officer might one day want to see — live in the vendor's database, behind their login, in their format. You can look at them as long as you're a paying customer. The day you churn, or they get acquired, or they sunset the product, your access to your own records gets complicated. That's a strange arrangement for documents the law requires you to keep for years.
This is the one design decision I feel strongest about, so let me make the case plainly.
What "the vendor holds your data" actually costs
It rarely feels like a problem on day one. The costs show up later:
- Lock-in. Once a few thousand receipts live inside an app, leaving means trusting their export to be complete and usable. Often it isn't — you get a zip of cryptically-named files and a spreadsheet that doesn't quite line up. The friction of leaving quietly becomes a reason to keep paying.
- Longevity risk. Software shuts down. Startups fold, products get discontinued, companies pivot. If the receipts you're required to retain for six or seven years live only in an app, your records are betting on that company outliving your retention window. Plenty don't.
- Privacy. Your receipts are a detailed map of your business — who you pay, what you buy, how much, how often. When that sits in a vendor's database, you're trusting their security and their data practices with a genuinely sensitive picture of your operation.
None of these is a dealbreaker on its own. Together, they add up to a simple discomfort: you don't actually control the records you're responsible for.
Why receipts in particular need to outlive the app
Plenty of software holds your data and it's fine — you don't lose sleep over where your to-do app stores its tasks. Receipts are different for one concrete reason: you have to keep them for years, and someone may ask to see them long after the fact.
Tax authorities can come back well after a financial year closes — retention requirements run to six or seven years depending on where you are, and an audit or query can land in year three or four. The receipt you need on that day might be from an app you stopped using two years ago. If your records live somewhere you own, that's a non-event. If they lived in a vendor's database you've since left, it's a scramble — assuming the data is recoverable at all.
Records you're legally required to retain should not depend on a subscription staying active. That's the whole argument, really.
What "you own it" should actually mean
Ownership isn't a marketing word here; it has a specific, testable meaning. Your receipts are genuinely yours when:
- The files live in storage you control — your own Google Drive, your account, your folder.
- They're readable without the app. A receipt image is just an image; you should be able to open it in Drive, on your phone, in ten years, with no software in between.
- They're organised in a way that makes sense on its own — foldered by company and year, named consistently — so the structure survives even if the app that created it doesn't.
- Leaving costs you nothing, because there's nothing to extract. The records were never held hostage in the first place.
This is exactly why Starlog writes every receipt straight into your own Google Drive, in a clean per-company, per-year tree. The app is how you capture and organise; it isn't where your records live. If you ever stopped using it, your receipts would sit exactly where they already are — in your Drive, yours, untouched. That's the point: an expense tool should make your records easier to keep, not become the thing you're afraid to leave.
The honest trade-off
There's a real counter-argument, and I'd rather state it than pretend it away. Vendor-held databases can do things a folder of files can't, at least not as slickly — instant structured search across years, fancy analytics, tight integrations that assume the data is in their system. If those capabilities are what you're optimising for, a closed database has advantages.
But for the core job — keeping the records you're required to keep, in a form you'll still control years from now — ownership beats features. You can always layer search and reporting on top of files you own. You can't easily get back data you've locked inside someone else's system. When the requirement is retention for years, bet on the thing that doesn't depend on a company staying in business.