1Mby1M Virtual Accelerator AI Investor Forum: With Yash Hemaraj, General Partner at BGV (Part 4)

Sramana Mitra: Enterprise has always been like that. We have this solution selling methodology that you have to break it down to technical decision makers, economic buyers, champion, and coach. That has always existed.

But if the go-to-market strategy is with a platform and you build something on top of that and start selling to the business users, sell there. Then you go to another business user and sell there. Then you want to penetrate into the IT organization so that that gets pushed to multiple different other business users. That’s that much more complicated, obviously, right?

Yash Hemaraj: Yes, but that’s where you have to land and expand, right? You land with one particular case—a very de-risked initial learning spot where people can start testing and seeing the value of your product. I often talk to some of my portfolio companies about the concept of the value waterfall. It’s about showing what value your product delivers, but you can’t stop there.

You really need to have measurement mechanisms in place to calculate the total value you’re delivering. The magic happens when you can get the customer to recognize and acknowledge the value your product has provided.

In many cases, even though you’re delivering a lot of value, the attribution often goes to the internal team, as in, “Hey, we did all this, and that’s why this value was delivered.” This can diminish the perceived value of the technology provider.

So, we need to focus on showing and attributing the value to your platform and figuring out how much of that value you can start capturing and building on. It’s an art that some entrepreneurs know instinctively, but it’s also a process we need to guide our founders through.

Sramana Mitra: It’s an art, but I think in general, if you have the option of selling to one buyer and really go deep into that organization and have a lot of offerings that can be sold to one buyer, companies grow much faster. Those are much higher velocity companies than if you have to sell to multiple parts of an organization.

Yash Hemaraj: So when I say “sell to the business user,” that’s what I mean—selling to the people whose business you’re directly impacting. But what I’m also suggesting is that there are multiple stakeholders who influence the design.

Sramana Mitra: No, there are two issues here. This is a very interesting conversation, by the way.

Within any enterprise purchase cycle, there are multiple buyers – economic buyer, technical decision maker, internal champions, coaches, etc. So that permits you to make one enterprise sale happen in solving one problem.

Now, what I’m saying is that if you go to market as a platform and you have one application with one business organization, another application with another business organization within the same enterprise, but then you have that whole complexity repeated multiple times across that organization. And that is a slower growth company than one where you can sell more into one purchase cycle.

Yash Hemaraj: Absolutely, that’s where you’ve got to find where the burning platform is. I’ll switch to a vertical application that we have.

Sramana Mitra: Finding a vertical application that has clear ROI that you can sell and then you can add ROI to that by doing more modules and more functionality, et cetera. If you build with that strategy, it’s a much faster growth company.

Yash Hemaraj: Absolutely, I’ll take an example. One of my best performing companies is a company called AI Dash. I invested in them in September 2020. At that time, they were using, traditional AI techniques of computer vision and  machine learning to focus on vegetation management.

What they built leverages satellite data, which is becoming cheaper, more granular, and more frequently available. They don’t operate satellites themselves, but they use the vast amount of satellite data being generated today at much finer, more frequent intervals. They take that data and have built models to predict which trees can grow. What species of tree it is, how fast it is growing, if it needs to be trimmed, and when it needs to be trimmed – that’s where the AI aspects of it come into play.

Instead of trying to serve many different households and use cases, they chose one industry where the problem is acute: utilities. A lot of people advised me against this, saying, “Utilities are hard to sell to. They have long sales cycles. It’s a bad investment.” These are common clichés. But guess what? This company has been growing 100% year-over-year for the last four to five years.

And why are they able to do so? Because it is really a burning platform with wildfires everywhere. One wrong pole, or one tree interacting with transmission lines, can cause massive damage, impacting millions of square miles, customers, businesses, and more. This is a very real, pressing issue.

But what’s important is how they handled it. They didn’t stop at just providing vegetation models to utilities. Had they done so, they wouldn’t have been as successful because many other companies offer similar models.

What they discovered was that in utilities, contracts are given out 6 to 12 months, sometimes even 24 months, in advance to people who cut trees. So, two things were needed: First, not just knowing where trees are today, but predicting where they’ll be in the future. Second, they developed user-friendly mobile and tablet applications for operators to take before-and-after pictures—before and after cutting the trees.

This enables closed-loop learning. If the model is inaccurate—let’s say it’s off by 5%—the data from the before-and-after photos feeds back into improving the model. As a result, they are the only ones in the market today delivering 90%+ accuracy.

Because of this, utilities are buying more. It’s an end-to-end, vertical solution. They didn’t just focus on the model; they also tackled the “boring” parts like building web and mobile applications. This is what we call solution engineering—thinking about systems-level design. That’s what we look at.

So, even if you have the best model, if you don’t think end-to-end, especially in legacy industries, you fail. These industries don’t have engineers like Google or Facebook to piece things together—you have to offer a complete, end-to-end solution. That’s key.

Now you can say, “We are the best intelligent vegetation management solution in the market today.” That gets you in the door and helps win the RFP. Then, they continue delivering value and having those value conversations with their customers. But eventually, you discover new use cases. That’s what I mean by “land and expand.” For example, they moved into encroachment detection. Now, they can apply the same solution to things like faulty pole detection.

There are a lot of other operating use cases within the utilities whose decision makers are adjacent to your decision maker, but you’ve already passed through all the enterprise decision makers once  that enable you to expand to other decision makers adjacent to the business unit that you originally sold to.

That is the way you expand. It’s the Jeffrey Moore’s bowling pin strategy. You basically capture one and then you go to the adjacencies and then you get the entire stack.

This segment is part 4 in the series : 1Mby1M Virtual Accelerator AI Investor Forum: With Yash Hemaraj, General Partner at BGV
1 2 3 4 5 6