Alright, let's talk about the elephant in the room with Iris.ai: their pricing model seems to actively punish growth. We just doubled our research team from 10 to 20, and the quote we got back felt like a parody.
It's not just the raw increase—that's expected. It's the *structure*. The per-user cost barely budged downward with volume, and the "scaling" tiers feel like they were designed by someone who's never had to justify a SaaS budget to a finance committee. The jump from our current plan to the next one essentially forces you to buy a suite of tools (the "Researcher Workspace" bundles) that half the team might not even need, just to get the document processing limits required for the other half.
So my question is: is anyone else hitting this wall?
* Are you just swallowing the cost and hoping the ROI materializes?
* Have you found success negotiating custom terms based on actual usage metrics rather than headcount?
* Or have you started looking at exit strategies, moving to a more granular, credit-based system elsewhere?
I'm skeptical of the "more seats, more value" argument here. Doubling the team doesn't mean we're extracting double the value from the platform, especially when a lot of the core AI processing is a shared resource. The pricing feels like it's built on a simplistic, old-school SaaS model, not on the actual value logic of an AI research tool. Feels like we're being charged for the privilege of scaling, not for the output.
Your free trial ends today.
You've identified the core issue perfectly. The disconnect between headcount growth and actual value extraction is a classic problem with per-user pricing for research tools, but it's especially acute here because Iris.ai's primary resource constraint isn't user logins, it's document processing volume. Tying the two together directly creates the exact scenario you describe, where you're forced to buy redundant 'seats' to get the document capacity your power users need.
We faced the same dilemma last year. We successfully negotiated a custom contract that decoupled the two, establishing a base fee for a pool of document processing credits and then a much lower, nominal fee for 'light' users who only needed access to review filtered outputs. It required providing several months of our actual usage logs to prove the disparity, but they were open to it. The key was framing it not as a discount request, but as a move to a model that would scale more predictably for both parties.
If they're unwilling to budge on bundling, that's when you should seriously evaluate the credit-based alternatives. The switching cost is non-trivial, but the long-term budget predictability often outweighs it.
Data is the new oil – but only if refined