Mathew FitzGerald has rolled out call intelligence before. At a previous company, he brought in one of the big platforms, and it flopped. Not because the insights were bad, but because they showed up too late. By the time a call was transcribed and the follow-up was ready to send, the rep had already moved on to the next conversation. The tool became one more thing people were supposed to check and quietly stopped checking.
So when Mathew, now Global Revenue Enablement at Foleon, started looking at Airspeed, he came in skeptical. He was the one building the business case, running the evaluation, and ultimately telling his CRO which platform to bet on. He wanted proof, not a demo.
He got it with a stopwatch.
The 87 second moment
Mathew literally sat there and timed it. How long from the end of a call to a usable transcript inside the system? With the legacy platform he’d used before, the average came out to three minutes and thirty seven seconds. With Airspeed, it was eighty seven seconds. And the follow-up email was ready the instant the transcript landed.
“That was the real wow moment, and that was when I realized it would work,” he says. The number is, in his words, etched into his memory.
That speed is not a vanity metric. It changed how his reps actually run their day. Before, calendars were booked solid from nine to five, back to back, no room to breathe. Now reps build in five, ten, and fifteen minute buffers between calls. In that little window, the transcript is done, the follow-up is drafted, and the prep for the next call is ready to go. The whole team just started moving faster, because the tool finally kept pace with them instead of lagging a step behind.
From a pile of transcripts to a system that acts
Here’s the thing Foleon was really trying to solve. They already had recordings. They already had transcripts, thousands and thousands of them, a genuinely amazing pool of data. The problem was that they couldn’t do anything with it. You could drop a transcript into ChatGPT or Gemini and get a follow-up email out the other end, sure, but that was the ceiling. Turning thousands of conversations into something you could actually learn from was a heavy data exercise nobody had the hours for.
What Mathew wanted was simple to say and hard to get: which discovery conversations, which customer pains, which questions, tend to land as closed won versus closed lost? Because if you know that, you can speak to those exact pains in onboarding, and you can speak to them in your marketing. That data was sitting right there, locked inside the transcripts, completely out of reach.
Airspeed opened it up. And once the Foleon team saw what was possible, they didn’t stop at one use case. They now have something like 145 agents running across the business, spanning customer success, new business, existing business, and the internal teams who spun up their own versions too. This is not a call recorder. It’s the system Foleon’s humans and agents work from together.
The agents that earn their keep
A few favorites, in Mathew’s words.
The multi-threading coach. Foleon’s revenue leadership dug into the most common factors separating deals that close from deals that don’t, and multi-threading kept coming up as one of the clearest signals. You can pull contact counts from your CRM, that part isn’t clever. The clever part is that every time a rep had a chance to multi-thread and let it slip, Airspeed catches it. A prospect says “I’ll need to run this past Steve,” and in the heat of a live call that can go in one ear and out the other. Airspeed flags it in the weekly coaching report: you never identified who Steve was, and you never looped him in. It’s a little sales coach sitting in every single conversation, quietly nudging deals forward.
The new-rep accelerator. A rep named Andy joined the US team and, about a month in, set the company record for fastest deal cycle, marketing qualified lead to closed won in eleven days. Mathew got curious about how someone that new was operating at that level, so he looked at how Andy was using Airspeed. Turns out Andy had built a weekly agent that scans the calls of Foleon’s top seller, pulling out patterns, objections, and how she multi-threads. Get the right information into someone’s hands, and they can act on it fast.
The product feedback loop. Sales teams are famously bad at proactively feeding feature requests to product, usually you only hear about it when a deal is already lost. So Foleon’s product team runs Airspeed across the month’s customer conversations, focused on the areas where they know they’re softer against competitors, and rolls the repeated requests straight into the roadmap.
The marketing realignment. This one’s a personal favorite. Foleon had a US industry where cost per click had exploded to the point it was hard to justify the spend. Instead of guessing, the marketing team used Airspeed to pull twelve months of customer conversations in that industry, with actual quotes, on how those customers describe their problems. They dropped it all into Claude and compared it to the copy in the ads they were running. Surprise, surprise: over twelve months, the ad copy had slowly drifted away from the language real customers use. The hypothesis now is that realigning to how customers actually talk brings that acquisition cost back down.
What it frees people up to do
Foleon’s business case landed at roughly fifty four hours saved per rep, per month, mostly from follow-ups firing automatically on triggers and call prep running itself. The interesting question is what you do with that time. For Foleon, the answer is redeployment: point reps at revenue-generating activity and watch the output climb. The reps who used to hand-fill CRM fields between calls now spend that time selling.
In fact, Airspeed got so accurate at populating the CRM that Foleon has the opposite problem most teams have. Reps trust it so completely that they’ve stopped touching deal admin, and now the team is working out what they actually need to force people to fill in by hand. That’s a good problem to have.
Mathew’s advice for anyone rolling this out
Do it incrementally, piece by piece, and let your trust build as you go. Start with the early win: call coaching, set up against the factors that actually decide your deals. Then move into CRM deal admin once you see how accurate the data is. Then get a bit more aggressive about what else you automate, whether that’s your customer success tooling or anything else eating your team’s hours.
“It’s laughably easy to implement,” Mathew says. “You can trust Airspeed. It’s really good.”
Want to see what 145 agents could look like for your team? Book a demo and we’ll walk through the first three worth building. You can also read the full Foleon customer story.
Frequently asked questions
How fast is Airspeed compared to legacy call intelligence platforms?
Foleon timed it. From the end of a call to a usable transcript inside the system, the legacy platform they had used before averaged three minutes and thirty seven seconds. Airspeed averaged eighty seven seconds, with the follow-up email ready the instant the transcript landed.
Why does transcript speed matter for sales reps?
Because reps move on. If the transcript and the follow-up arrive after the next call has started, the work never gets done. At Foleon, reps now build five, ten, and fifteen minute buffers between calls, and the recap, the follow-up, and the prep for the next conversation are all ready inside that window.
What can a revenue team do with thousands of existing call transcripts?
More than draft follow-ups. Foleon uses Airspeed to find which discovery conversations, customer pains, and questions correlate with closed won versus closed lost, then feeds that language back into onboarding, product roadmap, and ad copy.
How many AI agents does a team like Foleon actually run?
Around 145, spanning new business, existing business, customer success, and internal teams who built their own. Mathew FitzGerald's advice is to get there incrementally: start with call coaching set against the factors that decide your deals, then move into CRM deal admin, then automate wider.
