Abishek Viswanathan ran product at Qualtrics and Apollo, co-founded Monaco and doesn’t believe vendor claims. Which is awkward, because we’re a vendor and Monaco runs on Radar. So we’ll just shut up and let him explain how that happened.
You ran product at Qualtrics, then Apollo. What made you want to do it all again with Monaco?
Qualtrics was the first time I saw the go-to-market stack up close. What seemed very logical and simple to me was untrue on the stack side. Everything was disparate and disconnected. A CRM, a couple of data vendors, a sequencer, a call recorder, a forecasting tool, all separate. Something as simple as “hey guys, let’s do a mid-market push, what do we need to get done?” turned into a project.
And the AI wave changed how you saw that problem?
All these systems pre-2020 were built for humans to get the context on what accounts they’re reaching out to. The speed and precision at which you could do outreach was hamstrung by the human ability to comprehend data. Humans work 40 hours a week and agents work 24/7, they comprehend data much faster. So we wanted to build a platform that machines and agents could get the most context out of, but give decisions to the end user to approve.
A lot of AI SDR companies are racing to full autopilot. You’re not?
We’re sequencing it deliberately. A lot of AI companies are sprinting to the autonomous bit and skipping the steps in between. We want to get there too, we just don’t think you get there in one fell swoop.
Most of those tools are also chasing volume. What’s your problem with that?
A lot of these AI SDRs focus on volume and speed, because email and all these channels are cheap. That’s the fastest thing and the laziest thing to do, to be honest. We are at the other end of the spectrum where we focus on an extremely small cohort of prospects with really strong signals.
Okay, can you explain “strong signals” to someone who hasn’t read your pitch deck?
A signal on an account is in a two-by-two grid. One scale is conversion and the other scale is coverage. If you look at your entire TAM, there is not one signal that is high coverage and high conversion, that just simply doesn’t exist. But there ARE lots of high conversion, low coverage signals like somebody being in your network, somebody signing up for a free trial, somebody visiting your website or visiting the pricing page. We want to expose all of that in one single place.
You brought Radar in while you were still in stealth. Why was knowing who’s on the website that important, that early?
Somebody coming to a website is a great signal for intent and you need a way to make use of that signal.
“We do want to make sure it’s white label and easy for our end customers, so they don’t have to manage 30 different vendors.”
Why not build the identification yourself? You’ve got the engineers.
Well, you’re always chasing a long tail. We want to build as many signals as possible and we are completely okay partnering with companies to give us the right signals. But we do want to make sure it’s white label and easy for our end customers so they don’t have to manage 30 different vendors for the signal.
You didn’t take our word for anything. What did you do instead?
We did a bunch of tests, of course. We did blind tests, and we also did inside-the-VPN, outside-the-VPN type of tests.
“Snitcher is obviously at the top of the pack in terms of account accuracy.”
The data we care about a lot is startups in the Bay Area because that’s where we start growing, they have the biggest pain and the accuracy in that segment is really good.
Be honest: how painful was getting Radar live?
Honestly? Not at all, it was really fast. The setup and deployment on our side was only 3 days.
“It doesn’t impact load times or performance on the website, which we rigorously tested. Some of the other vendors impacted load times significantly, which was painful.”
What did your engineers say? They’re usually the harshest critics.
“The developer experience was top notch — what we would expect from another Bay Area tech startup.”
The APIs were extremely easy to use and the documentation was really good. We had a hiccup initially with logging events, but once we got through, it was fairly straightforward.
“We have activated new customers on this. We can immediately see the logs come through and we can verify for them.”
Dream scenario: what does this look like when it’s fully agentic?
The agent looks at all the signals at the account level and determines which one is going to be the highest conversion but the fun part is agents can now combine them. If you know somebody within an account and they also visited the website, “hey, glad to be connected, I also noticed there are some folks from your company on our site with this specific use case, would love to chat.” When you have two high conversion signals joining together, that’s a really great problem to have.
Last one. Anything else you want to snitch on?
Only that the whole game is two things: getting customers and closing them. We tend to pick partners who do one thing extremely well instead of thirty things badly.
Radar passed his tests. It’s ready for yours.
