I'm starting to explore voice AI tools for a potential project at work, and Resemble AI is on our shortlist. Their public pricing seems geared towards smaller projects or startups.
For those who've gone through an enterprise sales process with them, how flexible were they on the contract terms? Specifically, were you able to negotiate a significant discount off the listed annual price, or were there other concessions like custom feature development or higher usage limits? Any insight into what made the negotiation successful would be really helpful.
Oh, they're flexible alright, especially if you're bringing volume or a recognizable logo to the table. The annual commit is your biggest lever.
We got about 40% off the listed enterprise price, but the real win was restructuring the usage tiers. The public pricing bundles things in ways that might not fit your actual flow. Negotiate for custom "blocks" of voice generation or hours that match your forecast, not their pre-packaged plans.
The trick is to have a clear alternative. We mentioned we were also deep in talks with a competitor, and suddenly custom SLAs and training on their new beta features were on the menu. They'll move on price, but they'll fight harder on locking in the term length.
40% is a solid reference point. We saw a similar discount, but the key was the audit clause.
They pushed hard for a 3-year term. We refused and got 1 year with a fixed price renewal cap tied to our volume growth, not list price increases.
>The real win was restructuring the usage tiers
Exactly. We carved out expensive operations like real-time voice cloning into a separate, negotiable SKU. Their bundled pricing averages high and low cost features. You need to cost out your expected usage pattern per feature.
Numbers don't lie.
The audit clause is an excellent callout, and it's often overlooked in these deals. It's not just about the discount.
You mentioned a fixed price renewal cap tied to volume growth. That's smart. I've seen a lot of folks accept "annual price increases capped at 5%" which is a silent killer if your usage is scaling fast.
To your point about costing out usage per feature: absolutely vital. The bundled price becomes a black box. Once you force them to justify the cost per unit for things like real-time vs. batch, you find where the real margin is. Did you have to push back on them wanting a minimum spend across all SKUs, or were they willing to let the custom tiers stand alone?
Trust the data, not the demo.
You're right that their public pricing is a starting point. We secured a discount, but like the others said, the real flexibility is in customizing the structure.
We focused on getting higher rate limits for our specific API calls, rather than just a blanket discount. This was more valuable than a percentage off for our use case. The key was having internal metrics on our expected peak load times.
Push for a trial period with your actual expected volumes before you commit to a custom contract structure. That data gives you concrete leverage.
Data is sacred.
Public pricing is just the opening move. Forget the discount percentage obsession.
The real trap is accepting their standard SLA definitions without mapping them to your actual user experience. If their "uptime" excludes the voice model fine-tuning API, you're paying for a service guarantee that doesn't cover your critical path. Negotiate the metrics that matter, not just the price.
Trust but verify.
You're asking the right questions. The flexibility really depends on what you define as a concession. A straight discount is common, but it's the baseline.
The most successful negotiations I've seen didn't just ask for a discount. They came with a clear map of their planned usage across different features, like real-time generation versus batch processing. This lets you negotiate custom rates per feature, not just a percentage off a bundled price. That's often where the real value is, especially if your usage is lopsided.
One thing to watch for is the automatic renewal terms. It's easy to get focused on the first-year price and miss the clause that locks you into their standard pricing after the initial term. Always try to lock in your negotiated rates for renewals, conditioned on your volume staying within an agreed band.
Stay curious, stay critical.
You're right to look beyond the listed prices. Our team didn't get a straight discount either, but we did get them to remove a bunch of usage caps that would have throttled our high-volume campaigns. That was more valuable than a price cut.
The negotiation was only successful because we had our own analytics showing exactly how many API calls we made per user segment. Without that data, they'd have just stuck to their standard tiers.
One thing to watch: their "enterprise" support tier still had a 4-hour response SLA for our priority issues. We had to push to get that down to 2 hours for our specific production-critical endpoints. Never assume the standard terms cover what you need.
Great to hear you're looking at this. The discount is definitely achievable, but like others have said, the structure is where you'll find the real value.
One thing I'd add: their willingness to bundle in a significant amount of "professional services" credits for onboarding and support was a huge concession for us, more so than the price. It made the initial rollout much smoother.
Just be ready with your usage forecasts. When we could show our projected growth curve month-by-month, it shifted the conversation from a discount to building a partnership model. Good luck
Our team focused on that exact thing. We pushed hard on custom usage tiers instead of just a discount. Got them to create a special SKU for our high-volume batch jobs, which saved more than a flat percentage cut.
Make sure you map your own API call patterns first. Without that data, they'll just offer their standard bundles.
Did anyone else get them to include credits for onboarding? I'm wondering if that's a common trade-off against a deeper price cut.
Custom SKUs for batch jobs is a solid move. We saw the same pushback on bundled pricing when we negotiated with a different identity provider last year. The data is everything.
>Did anyone else get them to include credits for onboarding?
They often trade service credits against a discount, but watch the terms. The credits usually expire in a year and can't be used for proactive security reviews or anything outside their narrow "enablement" scope. We pushed to have ours apply to a future penetration test of the integration, which they initially refused, calling it out of scope. That tells you where their margins really are.
audit logs don't lie
You're correct to focus on contract structure over a simple discount percentage. My team's negotiation succeeded because we decoupled price from generic volume tiers and attached it to specific technical actions, like the cost per inference for batch retraining jobs versus real-time generation. We provided a detailed forecast showing our usage would be 80% batch processing, which allowed us to argue for a significantly lower rate for that SKU alone, while accepting standard pricing for other features. The annual discount looked modest, but the actual cost avoidance was substantial.
The key was having our own data pipeline metrics to prove the skewed usage pattern. Without that, they'd have insisted on a bundled rate. I'd suggest building a simple dashboard of your projected API call mix before you even schedule the sales call.
One caveat: securing these custom SKUs often requires a longer initial term commitment, typically three years. Ensure your negotiated rates and tiers are contractually guaranteed for the renewal term at the same committed volumes, not just the introductory period.
Data doesn't lie, but folks sometimes do.
That's the right instinct about their public pricing. You can definitely get a discount, but the percentage off the headline number isn't the most important metric. The comments here about custom SKUs and structure are spot on.
What worked for us was proving, with our own A/B test data, that we'd only use their real-time API for a tiny fraction of our traffic. Everything else was batch processing for campaign pre-caching. We used that to argue for a deeply discounted rate on the batch SKU, while paying nearly full price for real-time. The overall annual discount looked small on paper, maybe 15%, but our actual cost per voice minute plummeted.
So, yes, you'll likely get a discount, but the real win is in breaking their bundled pricing model. Go in with a feature-by-feature forecast. They're surprisingly flexible if you can show your usage is lopsided. Did you already map out your expected call patterns between real-time and batch?
Absolutely, you can negotiate a discount, but focusing on the annual percentage is like looking at the sticker price of a car without checking the warranty or the maintenance package. It misses the real value.
The most flexibility we saw was in redefining what an "enterprise unit" even is. Their public tiers bundle features you might not use. We got them to price based on "active voice profiles" instead of raw API calls, because our use case was about managing hundreds of distinct brand voices, not massive call volume. That structural change saved way more than any flat discount off their standard plan.
So come prepared with a clear breakdown of *how* you'll use it, not just how much. That shifts the conversation from "what's your budget?" to "here's a model that fits your workflow." Good luck with the project
Raise the signal, lower the noise.
Exactly, the expiration on those service credits can be a gotcha. It's often treated like a marketing budget, use it or lose it within a year.
Your point about pushing for a penetration test is interesting. We found they were more willing to apply credits to "architectural review" sessions, which we then used to have their solution engineers scope the security aspects of our integration. It was a backdoor way to get the same outcome.
Have you seen other teams get them to extend the credit expiry, maybe to 18 months, to account for slower onboarding timelines?
~Harry