Wednesday was the single best day of my 20-year career.
I don't say that lightly. I've bootstrapped Groove from zero to $5M ARR, built and sold companies, and had some genuinely great days in this business.
But Day 2 at SaaStr was different. It validated our product, offer and positioning like nothing else I could’ve imagined. By the time we sat down for dinner on night two, we'd booked over 60 demos that day with ICP aligned mid-market companies. ACVs of $30K, $40K, $50K who are truly hoping we can liberate them from Zendesk.
Over three days, we booked 120+ demos total. And on paper the investment pays back if even two or three of those close.
But what made it an all-time day was everything around it.
I want to walk you through all of it and cover what we learned (and what I'd do differently).
If you're thinking about investing in a conference booth, running founder-led sales, or trying to figure out how to position an AI product in a crowded market, there's something in here for you.
The setup: why we bet $100K on SaaStr
Let me give you the context on the bet.
Helply is the AI-Native support platform built specifically for B2B tech companies.
We'd been building for nearly 12 months w/ five full-time engineers, well over a million dollars in development. Already had 250 active paying customers in the pilot.
The product was ready. But we needed real-world validation at scale and to sit across from support leaders at target market companies and hear them tell us whether this thing was going to work.
So we put $100K into SaaStr.
We splurged for the main entrance booth which was prime real estate. It was literally the first booth everyone saw walking in. The sun was hitting it in the morning like a spotlight.
You couldn't miss us.
The team: me, Tom, and Jared at the booth running demos and conversations. My wife (who came last minute because I didn't trust the first agency we'd hired…more on that in a second) and an agency rep out front, greeting people, handing out swag, and pulling foot traffic in.
We wore hawaiian shirts, had an 8ft panda mascot, stickers. Oh yeah, and I brought 16 protestors.
Not to mention, we had an offer I was betting would be irresistible.
The offer was zero seats. Only pay for AI outcomes. No per-agent pricing or charging you three or four different ways (for seats, AI agents, outcomes, and analytics like our competition does).
In a market where Zendesk is charging per seat, our newer competitors are now layering seats plus AI add-ons plus outcome fees, and everyone is nickel-and-diming support teams.
We walked in with the simplest, most aggressive offer on the floor.
Here’s how it played out.
Day 1: Chaos and finding our groove
Day 1 was shaky if I’m being honest.
The morning started with the protesters.
We’d organized a group to demonstrate on the opposite side of the venue, a guerrilla marketing play. Tom handled the logistics, and thank God for that, because the people we’d hired were… let’s just say they were… basically Craigslisters.
A few of them couldn’t even walk straight. Tom organized them like a boss. He was cool, collected, got it done. There was a captain among the group who was solid and acted as a liaison. But it was messy.
Security was already annoyed with us before the conference even started.
Meanwhile, at the booth, the rush hit before we had our SOPs dialed in.
I had to run out for a photo op, ran back in, and suddenly there were 20+ people swarming the booth. It was me, Jared, my wife, and the woman from the agency and we hadn’t figured out our system yet.
What’s my priority here?
Am I having long conversations?
Booking demos?
Doing product showcases?
All three?
The first hour I improvised.
Here’s what happened: The SOP worked itself out organically, but we bled a little while it did. We probably missed five genuinely good leads in that first hour of figuring things out. Five potential $30K-$50K ACV deals that walked up while we were still finding our footing.
One of our new competitors (whose booth was right behind ours) came up before the conference started, being buddy-buddy and passive aggressive.
“Oh yeah, great booth setup. You’ll probably do well with those Hawaiian shirts.” I didn’t care. I just thought “you’ll see what happens.”
By the afternoon, we’d found our rhythm and started booking demos. Day 1 wasn’t a disaster, but it wasn’t the machine it would become.
Day 1 lesson:
If you’re doing a conference booth, do a dry run of your SOP before the floor opens. Decide in advance: what’s the single action we want from every interaction? For us, it became crystal clear by Day 2… book the demo. Not “scan their badge,” or “have a great conversation,” or “take a photo of their card and follow up later.”
Book the demo right there.
Day 2: The best day of my career
Day 2 started at 8 AM. The conference ran until 7 PM with happy hour after. It was known to be the busiest day. And this time, we were ready.
No distractions.
We had no protesters to manage. Tom, Jared, and I were all at the booth from the start. My wife and the agency rep were positioned perfectly out front. We knew exactly what was coming, and we knew exactly what to do.
The SOP was locked in.
Here’s what it looked like:
The agency rep and my wife greeted people on the floor, made conversation, and funneled anyone with even a spark of interest toward the booth. When they got to us, we had a simple qualifying conversation, what are you using now, what’s your biggest pain point, how big is your team? If there was fit, we didn’t linger.
We booked the demo right there on the spot.
No scanning badges, “let’s connect after the conference”, lengthy product walkthroughs at the booth.
The result: roughly 60 to 65 demos booked in a single day.
I was sitting at the booth like a boss. People were lining up and companies coming from Zendesk, complaining about Zendesk. Companies coming from the newer competition, complaining about them (right in front of that companies booth now less). And I’m just locking them in.
Demo. Demo. Demo. Demo.
These weren't small businesses kicking tires. These were mid-market to upper-SMB companies ($10M+ ARR, $30M+ ARR). Big support teams who were desperately looking for a solution. ACVs of $30K, $40K, $50K.
Some big boys in there too, deals that could individually pay back the entire booth investment many times over.
After months of building, spending, and wondering whether the market was truly going to be there, this was the moment. 1,000% validated.
The offer and positioning was right. And our product was ready to compete.
The objections that kept coming up (and exactly how we handled them)
Across 100+ conversations, three objections came up over and over, and how we handled them became some of the most valuable reps of the entire conference.
Objection 1: “What counts as a successful AI outcome?”
This was the most common one, and it usually came from more technical buyers, support leaders at mid-market companies, $10M+ ARR.
They wanted to know: how do we define “success” when the AI handles a ticket? What if it resolves something badly? What if we’re paying for outcomes that aren’t actually good outcomes?
Here’s how I handled it:
The framing that worked best was the 90/10 working relationship. I’d say something like:
“Our job is to put the checks and balances in place so you can always audit your AI outcomes, see exactly what happened, how the resolution was derived, and whether it was successful. Will 100% of outcomes be perfect? No. That’s almost impossible. But the collective outcomes, let’s say 90%, from an ROI perspective will massively outweigh the cost of the 10% that might need human review. And here’s the thing: you’re paying zero seats. So even at 90%, the value you’re getting far exceeds the cost. And for the 10% that doesn’t land? Jump in our Slack. Call us. We’ll work with you, extra credits, refunds, whatever it takes to make it right. This isn’t a ‘buy and we push you off to the side’ situation. This is a working relationship.”
That framing resonated every single time. The buyers who asked this question were sophisticated enough to know that no AI is 100% accurate. They needed confidence that we’d be accountable, transparent, and in it with them.
Pro tip: If you’re selling an AI product, don’t hide from the accuracy question. Lean into it. The 90/10 framing works because it’s honest, it pre-empts the “but what about the edge cases” follow-up, and it immediately shifts the conversation from “is this perfect?” to “is this worth it?” The answer to the second question is almost always yes.
Objection 2: “My CFO needs predictable pricing”
This one came up several times, and it’s a legitimate concern. If you’re paying per outcome and ticket volume spikes, the CFO can’t forecast costs. That’s a real blocker for mid-market companies with proper financial planning.
Here’s the booth script that worked:
“We also offer bundled pricing with annual contracts. We look at your account, estimate your expected outcome volume, and lock you into one set price for the year. Your CFO gets a predictable baseline. If there’s a huge unexpected spike, like your ticket volume suddenly doubles, we’ll reach out and work with you on what that looks like. But the baseline is locked in, and you’re not paying for seats on top of it. Just outcomes.”
The key insight: annual contract bundling gives the CFO the predictability they need while keeping the pricing model outcome-based. You’re not reverting to per-seat pricing. You’re just packaging the outcomes into a predictable number.
I referenced how Intercom handles this with Fin. A buddy of mine locked in with Fin at his $30-40M ARR company for about $12K per year on an annual contract. If volume spiked dramatically, the AE would reach out, but not in a punitive way. They’d show the ROI: “Look, your tickets went up 2x. That’s unsustainable at the current price. But look at the value, we just saved you three headcount.” And the response was always: “Yeah, no problem. You’re not just raising the price on me. You’re working with me.”
Objection 3: “How are you different from the new competition / other AI support tools?”
This one I loved. A few times, people walked up and just straight-up asked about our newest AI native competitors. My favorite response:
“Do you like paying three ways? Four ways? Per seat, then lock into a contract, then pay for AI add-ons, then pay for outcomes on top of that? That’s not us. We give you the outcomes. Zero seats. One price. That’s it.”
That landed every single time. People would nod and say: “Yeah, you’re right. I don’t want to pay three ways.”
The lesson here goes beyond Helply. In a crowded market, the simplest offer wins. When a buyer is evaluating five AI support tools that all claim to do the same thing, the one that can explain its pricing in one sentence has a massive advantage.
The Zendesk goldmine
I want to call this out specifically because it was one of the strongest signals from the entire conference.
Almost every single customer who came to our booth and used Zendesk was unhappy. Not mildly dissatisfied. Actively frustrated. Looking for an exit.
The most memorable conversation: a support leader from a $30M+ ARR company with a team of 60 agents. He walked up and said almost verbatim:
“Zendesk is restraining me from adding more of my team. I can’t get my whole company on the platform. It’s not even close to the value it could deliver if I had everyone on it.”
Then he said: “The zero-seats model is really intriguing based on the outcomes.”
Then he booked a demo.
That was it. No long pitch. No elaborate product walkthrough. The zero-seats offer solved his core problem (Zendesk’s per-seat pricing was literally preventing him from scaling his team), and he was ready to move.
This isn’t one anecdote. This was a pattern. Zendesk customer after Zendesk customer, complaining about the same things: per-seat pricing that constrains growth, a product that feels like a dinosaur, and no path to an AI-native support experience.
The takeaway for founders: If you’re entering a market with an entrenched incumbent, go to the conferences where their customers gather. Don’t just exhibit, listen. The pain points will tell you exactly how to position your product. We didn’t have to convince anyone that Zendesk had problems. They were telling us unprompted.
The panda incident (yes, this is relevant)
I have to tell this story because it was one of the best brand moments I’ve ever been part of, and there’s a real lesson in it.
We had a person in a panda costume as part of our booth presence. Day 2, security came over almost immediately: “Nah, dude, the panda’s got to go. That’s free advertising on the floor.” They were already annoyed with us from the protester stunt on Day 1. “You guys again? Really?”
The panda got kicked out.
But I asked security: “How about happy hour? Four to six?” They thought about it. “All right. He’s in.”
So I knew what was coming. Four o’clock, the panda comes back.
Here’s where it gets perfect. Another company had sponsored a 20-person drum line that was ripping down the main hallway toward the bar. The drum line passes our area. The panda falls in behind them. And now the panda is dancing with this massive drum line, parading down the hallway, right past Pylon’s booth, and dancing into the bar.
People were lining up to take selfies with the panda. Waiting in line. At the bar. At a SaaS conference.
Tom was there with the team, running around, making sure we captured the moment. It was the best organic advertising stunt I’ve ever pulled off. Zero cost beyond the costume. Maximum visibility. And the kind of moment that people talk about for weeks after the conference.
The lesson: At conferences, you’re not just competing for demos. You’re competing for attention and memory. Everyone has a booth. Everyone has swag. The companies that create genuine moments, things that are funny, surprising, human, are the ones people remember and talk about. The panda didn’t directly book us demos. But it made Helply the most memorable brand on the floor. And in a market where everyone’s pitching the same AI capabilities, being memorable is a distribution advantage.
The bigger picture: what SaaStr validated about our positioning
Beyond the demos and the stories, SaaStr confirmed something strategic that I want to share because it applies to anyone building in an AI-saturated market.
The vertical B2B thesis
Helply isn’t a horizontal AI chatbot.
We’re going deep, extremely deep, into B2B tech support. Mid-market companies, $1M to $50M ARR. That’s our world. And the reason nobody can touch us in that vertical is specificity.
Here’s what we’re building that a horizontal competitor can’t match:
- Deterministic workflows built specifically for B2B support use cases, not generic conversation flows
- Revenue and product intelligence surfaced from support conversations, churn signals, upsell opportunities, feature requests. Zendesk doesn’t do this. Pylon doesn’t do this.
- B2B-specific channels like Slack, Whatsapp, and Discord, where B2B support actually happens (not just email and chat widgets)
- Support-to-revenue conversion, turning the support function from a cost center into a revenue engine
When I explained this at the booth, the reaction was consistent: “Nobody else is doing this.” And they’re right. Nobody else is going this deep into the B2B vertical because it’s not a horizontal play. You can’t build this with generic AI.
This was also a major theme at SaaStr beyond just our conversations. Jason Lemkin’s sessions touched on build vs. buy (verdict: you’re crazy if you build your own support platform), outcome-based pricing was a recurring session topic, and turning support into revenue was being discussed across multiple stages
Even Mintlify, a documentation management tool, just announced outcome-based pricing last week. The market is moving in our direction.
Day 3: The bonus round
The last day was shorter and lighter. But we came in with the same energy and the same SOP.
Another 30-35 demos booked.
That brought us to ~120 total demos across three days. From a $100K investment. If even two or three of those deals close, and some of the companies in our pipeline are significant, the investment pays back multiple times over.
After 12 months of building, well over a million dollars in development, and all the uncertainty that comes with launching into a crowded AI market, I know exactly where we stand.
We have an irresistible offer. Zero seats, outcome-based pricing, a product that goes deep into B2B support, and a positioning that resonates with exactly the buyers we want.
Now it’s about distribution.
Consistently getting in front of more of these buyers, across more channels, every single week. The conference validated the product and the offer.
Our next chapter is all about scaling our reach.
That starts now.