The Five Clauses That Quietly Cost Startups the Most
Nobody gets burned by an exotic clause nobody has heard of. Startups get burned by the same five clauses, over and over, because nobody was reading closely enough.
7 min read
In four years of reviewing vendor, customer, and partnership agreements for early-stage companies, I have never once seen a founder get hurt by an unusual clause. Nobody gets ambushed by an obscure force majeure carve-out or a strange choice-of-law provision from a jurisdiction they've never heard of. What actually costs startups money — sometimes a lot of it — is one of about five clause types, appearing in slightly different clothing in almost every contract that crosses a founder's desk.
These clauses aren't dangerous because they're clever. They're dangerous because they're boring enough that people stop reading carefully by the time they get there. Here are the five, in the order I'd triage them.
1. Uncapped indemnification
Indemnification clauses are where one party agrees to cover the other's losses if something goes wrong. That's reasonable in principle. The problem is scope. "Any and all claims, damages, and expenses arising from or related to this Agreement" sounds like standard boilerplate, and it is — which is exactly why it gets skimmed. Without a cap, you have agreed to an open-ended liability that has nothing to do with the size of the deal. I've seen a €40,000 pilot agreement carry an indemnification clause that, read literally, could have exposed the vendor to seven figures in exposure over a dispute that had nothing to do with the actual services performed.
The fix is almost always the same: tie the indemnification cap to fees paid, typically over the trailing 12 months, with carve-outs for gross negligence, willful misconduct, or IP infringement where a higher or uncapped standard is more defensible.
2. Auto-renewal with a short opt-out window
This one doesn't cost you in a dispute — it costs you in inertia. A contract renews automatically unless you cancel within some window before the term ends. When that window is 15 or 30 days, and nobody owns tracking it, you end up locked into another 12 months of a tool you'd already decided to drop, or a vendor rate you meant to renegotiate.
The commercial fix is a 60-day notice window, which gives a reasonable amount of time to actually notice the deadline is coming. The operational fix is separate and just as important: something has to actively track these dates, because good contract language doesn't help if nobody's watching the calendar.
3. No cap on general liability
Distinct from indemnification, a limitation of liability clause caps what either party owes the other for ordinary breach — missed deadlines, service failures, that kind of thing. When this clause is missing entirely, or drafted so broadly it doesn't function as a cap, a single bad outcome on a contract worth a few thousand euros a month can turn into an existential dispute. This is the clause I'd flag hardest for any company without in-house counsel, because its absence is easy to miss — you're looking for something that isn't there.
4. Broad IP assignment language
Common in consulting and contractor agreements: "Contractor assigns all right, title, and interest in all deliverables, including all work made for hire." On its own, fine. The trouble starts when that language is broad enough to sweep in tools, scripts, or frameworks the contractor built before the engagement and reuses across clients. Without a background IP carve-out, a contractor can inadvertently assign away the foundation of their own business, one client contract at a time.
5. Personal data without a data processing agreement
If a contract involves processing personal data — customer records, employee data, anything covered by GDPR — and there's no reference to a Data Processing Agreement or Art. 28 terms, that's not a stylistic gap. For companies operating in or selling into the EU, it's a compliance gap, and it's one of the most common gaps we see, because DPAs are often handled as a separate afterthought rather than a required attachment.
The pattern underneath all five
None of these require a specialist to catch. They require someone to read the whole contract, every time, with the same checklist in mind — which is precisely what's hard to guarantee when review depends on whoever has 20 spare minutes that week. That's the gap we built Conlegie to close: the same five checks (and about 35 others), run consistently, on every contract, regardless of who's available to read it.