So I was doing a capacity and cost review for a sales team migration, and the client was all-in on Pipedrive because of their "unlimited contacts" promise on their higher tiers. You know the sales pitch: "Scale without worry, no arbitrary limits, perfect for growing teams!" I'm inherently skeptical of any vendor using the word "unlimited," as it's usually a red flag masquerading as a feature.
Sure enough, after digging through the labyrinthine support documentation—buried far away from the main pricing page—you hit this little gem. The "unlimited" contacts are, in fact, subject to a "fair usage policy" that caps you at 100,000 contact records. Once you breach that, you're asked to upgrade to a custom "Enterprise" plan, which of course involves a sales call, a minimum contract term, and a price that is nowhere listed publicly. This isn't an edge case; for any business with a decently sized list or one that uses contact data for marketing purposes, 100k is a very reachable limit.
This is a classic bait-and-switch tactic dressed up as modern SaaS pricing. It forces you into a completely opaque negotiation at the worst possible moment: when you're already locked in and actively hitting the limit. Your leverage is gone. From an infrastructure planning perspective, this makes forecasting costs and growth trajectories a nightmare. You can't build a reliable model on "unlimited."
The practical gotchas here are numerous:
* **Renewal Shock:** Your next renewal isn't a simple per-seat bump; it's a jump to a completely different, un-priced tier.
* **Data Architecture Workaround:** Teams start deleting old contacts or moving them to external systems, undermining the single source of truth the CRM is supposed to be.
* **Hidden Contract Terms:** The "fair usage" clause is subjective and gives the vendor all the power to define what "fair" is.
If you're evaluating this, you must factor in this hidden ceiling. Your true cost projection should look something like this:
```plaintext
Current Plan: "Advanced" @ $99/user/month
Projected Year 2 Contact Count: 120,000
Result: Forced upgrade to "Enterprise"
Enterprise Cost: Unknown (Sales negotiation, likely 2-3x current)
New Contract Term: Minimum 12 months, likely 24.
```
The lesson, as always, is that there is no such thing as "unlimited" in infrastructure. There is only capacity that is metered visibly or capacity that is metered invisibly, with the latter always being more expensive and risky. You're better off with a vendor that has clear, published overage fees or tiered limits you can actually plan for. This kind of obscurity is just over-engineering your financial risk.
monoliths are not evil
Exactly. This "fair usage" backstop is why I always pull the raw API response for entity counts during vendor evaluations. I've seen the same pattern with "unlimited" storage in BI platforms that throttle query performance after a terabyte.
For Pipedrive specifically, the 100k limit applies to the total contact objects, not just active leads. That includes archived and deleted records if they haven't been purged. So the effective limit hits faster than you'd think if your sales process cycles through a lot of contacts.
Yep, that's the exact moment the project's TCO model goes out the window. Been there.
The real kicker is when you're already using their webhook integrations or have built internal tools around their API. Migrating off becomes a multi-quarter rewrite, so you're stuck paying the enterprise tax.
Always check the /api/v1/limits endpoint during your proof-of-concept, and get *written* confirmation from their sales about any soft caps before signing.
garbage in, garbage out
Oh, you still believe in "unlimited" promises? That's cute. The real issue isn't the hidden limit, it's that anyone plans capacity based on marketing copy instead of the actual API constraints. The moment you base a migration on a sales slogan, you've already lost.
I've seen this exact scenario play out with serverless function invocations and "unlimited" concurrent connections on API gateways. The pattern is identical: a soft cap buried in the docs that becomes a hard business limit the second you approach it. The vendor isn't lying, technically, they're just relying on you not reading the fine print before you're architecturally committed.
If you're not hitting their `/limits` endpoint during design and hammering their pre-sales support with "what happens at 100,001 contacts, in writing" before you move a single record, then the bait-and-switch is on you as much as them.
Yep, the "unlimited" claim is pure marketing theater. The 100k cap is bad, but the opaque enterprise pricing is worse. You're negotiating with zero leverage and no public pricing to anchor to.
My rule: any "unlimited" offer from a SaaS vendor gets a pre-sales question in writing: "At what exact volume do your fair use terms engage, and what are the specific commercial terms at that point?" If they won't answer clearly, walk away.
Trust but verify.
>pull the raw API response for entity counts during vendor evaluations
Good call. Makes me wonder if there's a standardized way to audit a SaaS's true operational limits.
What else do you pull from the API? I've checked for rate limiting headers and concurrent connection caps, but there's always something you miss until a process breaks.
null
The architectural lock-in you've described is the real cost multiplier, more than the hidden cap itself. You're now forced into an enterprise negotiation while every integration webhook and internal reporting script you've built assumes the Pipedrive data model. Migrating off at that scale requires a parallel run, data mapping, and regressing every workflow.
This is why any vendor evaluation now includes a "decommissioning" section in the design doc. You need to model not just the integration cost, but the cost and timeline to extract and replace the service when these terms shift. For a CRM, that means understanding the true schema, mutation rates, and whether you've used any proprietary fields that don't map cleanly.
I've started treating "unlimited" in marketing copy as a directive: immediately go find the rate limiting, concurrency, and total object endpoints in their API. If they don't expose them, that's a different kind of red flag.
infrastructure is code
Absolutely, and that shift from soft cap to hard limit is where the real damage happens. You're cruising along, everything's fine, then suddenly you're getting warnings about "performance degradation" or needing a "consultation" to keep things running. It's never a clean line.
> the moment you base a migration on a sales slogan, you've already lost.
This is so true. I think the problem is that technical folks are brought in too late sometimes. Marketing and sales teams hear "unlimited" and run with it, and by the time we're asked to build the integration, the decision feels set in stone. We have to push back earlier and ask for the ugly details during the vendor demo, not after the contract is signed.
You mentioned serverless and API gateways - saw the same with email sending on a "unlimited" plan. They let you hit 50k/month, then suddenly you're throttled and support says you need a dedicated IP contract. The pattern is everywhere once you start looking for it.
If it's not measurable, it's not marketing.