Welcome to a Fort Knox update. I am John Fort here this time with Amit Bendov, the CEO and co-founder of Gong. It's been a while, it's good to see you. AI for sales, for just about everything, is very much top of mind. I was just having earnings conversations with Snowflake, which reported today, way up, talking about AI. We've got Elastic coming up after the bell with some strong anticipations there. What are you seeing in sales and revenue generation when it comes to AI, especially with this annual report that you're putting out for the first time?
Well, first, these are exciting times. We're happy to see it. You and I have been speaking for the last few years about how AI will transform the world, and in sales specifically. So it's great that it's finally happening. We've run a study of hundreds of companies and we're seeing two good things. First, we've seen on average companies are seeing 19% growth year over year, which is very exciting to signal that the market is starting to improve in general. But the interesting bit is that companies that are using AI have seen a 29% increase versus 19, so almost 50% better growth, which is very exciting. The adoption is the fastest thing that we have seen. People are asking, well, what's the impact? Everybody's talking about AI, but what value does it create? So definitely, we have the evidence. I'm calling from California around their annual user conference. We have companies like Workday and LinkedIn and Reddit all reporting. One of our customers, for example, HST Newspapers, reported they've increased the win rate by 31% with the use of AI. So real companies getting real benefits and growing faster than the market.
Well, it's the whole thing. For example, Reddit mentioned that they use insights automatically surfaced by AI to fine-tune their messaging, their sales motion. They would be in the dark if they just relied on CRM; they have no idea what's going on between their customer-facing people and their customers. AI is now transforming the organization. We announced today AI Briefs, for example. This is a small example that's costing companies dearly. Every day, companies transition accounts between account managers, right? Hundreds. And usually, it's not a great experience for customers. You get a new person that doesn't know them, and in the worst case, might lose the business or at least create that negative experience. Now, AI will prepare the new person: here's our entire history, here's what you should be knowing, why they bought, what's important to them, who are the key people. If I want to do this manually, it would take me hours. People would spend a whole team. AI now, 30 seconds, you got this, you're ready for the meeting as if you've been working them for ages.
So you talked about the shortcomings in CRM, and I take it that means how much information the system has really captured that's going to be useful for you going forward. So shortcomings in the structured data that's really in there. What is the benefit of unstructured data to the sales process, and how does AI work to really mine that out and put it in front of the user? You just gave one way there with those briefs, but in a way that's actionable and useful.
Well, that's a great example because first, what's better about the unstructured data? It's 100 times bigger than what's in a CRM. So there's a lot more. When you have a meeting like we're talking right now, an hour conversation with a customer, there'll be some 6,000 words being exchanged, give or take. What ends up in the CRM is somebody manually going and filling in forms: you know, what's the amount, what's the close date, and they might add some notes like, oh, it was a great meeting, they're excited, this is it. So 30 words. So think about the loss. This is very lossy. Everything is in people's heads, it's in email, it's in dinners that they've had with customers. None of that even makes it to CRM. And with AI, you don't lose anything. Everything is captured. We spoke, and then you could always go back, immediately after or in real time, or maybe a month after or a year after, and all of there is automatically transformed to structured data, nor reports. You could trigger, for example, you identify customer expressed needs, now this could trigger a workflow that will create an upsell motion for these customers, all by AI. So huge, huge difference and totally a ballgame.
This is going to challenge, perhaps swamp some moats, won't it? Because right now there's this dust-up between Salesforce and Microsoft. Marc Benioff's been taking shots at Microsoft over Copilot and agents. Microsoft saying, wait a minute, what our Copilots and agents are doing is letting people get information about a customer, prepare for an interaction without actually having to open up their CRM. I mean, when some of these moats were based on structured data over a period of time that made it difficult to move off of a particular system, to what degree is AI and the ability to make use of unstructured data that's not locked in any one particular system, to what extent is that going to break down some of those moats and barriers and defensive positions?
Well, that's the most important thing, and you touched upon it. It's the completeness of the data. Because if you want AI to work, it needs to have complete information. Think about AI trying to diagnose a patient, but it doesn't have all your records. It doesn't know what you do and who you are and your age. So if you need to make a customer decision, you need to have all the information. If you rely on what's in the CRM, which isn't very good and it's obviously subjective, you're almost better off not having AI at all. So it is either you have complete information and you can run proper AI, or you have partial information and it's almost best not to use AI at all. So the key would be capturing everything. Just like a Waymo car, right? I mean, it needs to know everything that's going on. If it doesn't understand that there's a motorcycle or a child crossing the street, this is disastrous. So responsible AI means that you have to use complete data, and data is the key. Capturing all customer conversation, also structured data. The structured data is important, but the customer conversation data is key to making AI work.
How much of a problem is sloppy data governance? This is another issue that's coming up lately, where maybe the permissions aren't set correctly on, you know, salary data or whatever else in the system. And sure, maybe somebody normally wouldn't be able to go in password-wise and access that, but with AI having access to a lot of data and people being able to query it, well, maybe they can pull out something that they otherwise wouldn't have been able to open a door to and access directly. Is there a lot of data remediation that needs to take place in order for customers to be able to get the best use out of this technology? And if so, how long is that going to take?
Absolutely. I mean, that's a major issue for companies. And it's not just slapping something on top like ChatGPT and, hey, we could read your customer. Because now people can ask questions like, well, is any of my employees getting a job offer, or is interviewing? So this is huge privacy issues. And people think, oh, you're going to prompt your way to AI. No, you need a framework that controls data access, control, legal compliance, HR, medical information. All of that needs to be part of a framework. You can't. Some people fantasize that you could do everything with one of the LLMs. No, you need an application that will control the data and make sure that it's used in a proper way.
In your annual report, you state pretty directly that the companies that are getting the best use out of AI are not replacing people with it, they are extending people with it. But I wonder, at a certain point, does it make it less necessary either to add that additional person or to add those support people that used to be necessary in order for that effective salesperson to build the relationships and frame the product in the way that she or he was able to in the past?
