How to Create a Category through a "Pull", not "Push"

A fireside chat with Barr Moses at Monte Carlo

Many founders are enamored with the idea of category creation, a motion that we were laser-focused on at Gainsight. But not all categories are created with the same ease. I was psyched to discuss this topic with Barr Moses, whom I worked closely with for many years before she started her own company, Monte Carlo -- which recently announced its Series B funding from Redpoint and GGV (plus me :) ) and also its recognition as a 2021 Enterprise Tech Top 30 company

In my interview with her, she talks about the incredibly creative mechanisms that she and her team have invented for detecting whether your customers are "pulling" you, versus you "pushing" them. She also discusses how she did things completely "out of order" to achieve tremendous product-market fit. Barr is a contrarian in the best sense, and that allowed her to create the emerging category of Data Downtime.

Can you tell us the origin story of Monte Carlo? 

It started when I was at Gainsight. You may remember this was back in 2016, and I was responsible for GonG, which is short for "Gainsight on Gainsight" -- this was our own internal instance of Gainsight that we used to manage our customers, since of course we drank our own champagne. We were using our own data to try to better understand our customers. The hardest thing about this was gaining trust in the numbers. I remember going into a room to meet with you and Will Robins to map out what our data looks like, where data often breaks, and what are the main reasons, then trying to figure out better ways for us to be able to trust that data. I remember thinking to myself how crazy it is that we have to get in a room to manually do this. We'd wake up to reports that were incorrect, or people emailing us to ask "Can I use this number? Is it up to date?" As a company that was really data-driven, it was quite frustrating to not be able to have the tools to be certain that you can trust your data and use it. Fast forwarding to my time after Gainsight, I started Monte Carlo with the goal of helping organizations build trust in their data. 

What new software category are you creating?

There were tons of other companies that had the same problem we had at Gainsight. I interviewed over 150 data teams, primarily asking them, "what's keeping you up at night?" The problem of lack of trust in data came up so often that I gave it a name: Data Downtime. Monte Carlo was founded in order to help companies become more data-driven by eliminating Data Downtime. 

We believe the way to solve this problem is by borrowing concepts from software engineering -- specifically, observability. Data can break due to millions of different reasons. Every organization has their own unique data and every company thinks they're a snowflake. But if you look at the analogous problem of Application Downtime, applications can similarly break for a million reasons, but there's still a consistent framework and consistent way to measure and to understand the health of your application with off-the-shelf solutions like New Relic, AppDynamics, and Data Dog. 

We adopted those concepts and applied them to data, so we actually created what we believe are the right frameworks for Data Observability. We created a kind of "New Relic for data teams." Just like any engineering team has something like New Relic in order to manage the health of their applications and infrastructure, any data team should have something like Monte Carlo to help to manage the health of their data.

When you're creating a new category, how do you balance having empathy for customers -- which helps you deeply understand their pain point -- with promoting your own point of view on how the market should evolve?

When we talk to our customers, the pain point is very well understood. There isn't anyone in data who hasn't experienced something like Data Downtime. Where we are challenging our customers is with the question, is there a better way to solve Data Downtime? 

Right now people are manually validating reports -- taking the numbers and having six different people look at them before those numbers are reported to leadership. That's definitely not scalable and is crazy to be doing in 2021. 

When we started out, we had a lot of visionary theories, but we needed to test those theories in practice. So it was critical to get a product into the hands of our customers as soon as possible. We iterated with our design partners. After a while our partners offered to pay us -- we didn't even ask them first. That was a very strong signal that there was a connection between our vision and the right solution for the pain point. 

It sounds like in creating a category, you never felt you were pushing a rock uphill -- which can sometimes be true for category creators -- but rather you were riding a wave, or fostering the growth of a wave that already had legs. Were there any other indicators that you looked at to validate product-market fit?

I'm thinking a lot about how much we're pushing versus being pulled. In the early days, you don't have a ton of data, so you're looking primarily for anecdotal evidence of value with your handful of design partners. We track something that we call "Hell Yeah! Moments." For example, let's say you're on a call with a customer, demoing a new feature or showing them something in their environment, and their reaction is: "Whoa, that's amazing." You can tell that person is incredibly excited. 

Our team is measured by how many Hell Yeah! Moments they can create for customers. For example, we'll point to minute 49 of a recorded call to show the excitement in the customer's reaction. Or the customer might show their reaction in a Slack message. We collect those moments and associate them with use cases, to help us evolve the product. And we share those moments across the team to celebrate.

For prospects, we look at what we call a "Jump-Off-Your-Seat" score. Once we've spoken with a prospect about their Data Downtime experience and then describe Monte Carlo's solution, we ask, does this resonate with you? You can see if they're jumping off their seat or if their response is more muted: "Okay, sounds interesting." 

One example of a high Jump-Off-Your-Seat score is when prospects send us materials and decks about why they need a solution like Monte Carlo. They're selling us! We share calls with high Jump-Off-Your-Seat scores with everyone on the team. And we track them so we know what parts of the product or what use cases are resonating.

How do you know when it's time to open up the floodgates to a broader set of customers? 

I'm in favor of focusing on what makes the most sense for your customers at the time. We had paying customers and a business before we even had a website. That wasn't exactly intentional. We literally had more customers who wanted to work with us than we could handle, so we de-prioritized the website until later. 

Going even further back in time, we were talking and writing about the problem and the category even before the company was incorporated. One time we decided to give a talk at a large data conference. I was sure that there were only going to be four people in attendance. 

But I was blown away by the reaction. There was a huge audience. Tons of people came up to me after and told me how excited they were about the movement. They were psyched that there was finally a name -- Data Downtime -- for the pain that they'd been feeling all along.


If you’d like to learn more about Monte Carlo’s approach to eliminating data downtime, reach out to Barr and the MC team:

If you know of any data teams who could benefit from data observability, you can share this post with them:


If you'd like to hear from more founders like Barr about once a week, sign up for my newsletter: