ServiceTitan, Inc. Class A Common Stock Pantheon 2026

If you will remember, we went through one major technology transition a decade ago together. Cloud changed how modern businesses operated. Most other industries jumped on the innovation, but the trades were initially a bit behind the curve. Many of you partnered with us to build ServiceTitan. You grew, we grew. As with all major tech transformations, those that adapted won, and those that did not fell behind. Now, the AI race is upon us, but this time, two things are different. First, you are a different leader. AI is still early, but you are already leading from the front. You are using voice agents, AI dispatch, AI coaching, ServiceTitan solutions, third-party solutions. Last time, we had to convince the industry to catch up. This time, the trades are in a much better position to win.

The second thing that is different is the scale is much bigger. The big tech companies invested $1 trillion over 10 years to build out cloud infrastructure. They are investing $1 trillion this year alone for the AI build-out. It is projected to be more than $10 trillion over the coming decade. That is 10 times bigger than cloud. Why? What do they see, and how do we prepare for it? The AI maturity model gives us a sneak peek into the future. Before AI, in the old world, humans did everything in a business. They analyzed the data, they came up with the insights, they decided what to do, they did the work. Now, the new world. Level 1 was ask AI. It is Siri on your iPhone, it is Atlas in ServiceTitan. Humans ask, AI answers. How much did we sell last month? Why did conversion rates drop?

Level 2 allowed us to automate individual workflows. A voice agent that answers calls and books appointments, a dispatch agent that sends the right text to the right jobs. Powerful stuff. Indeed, many of you are already here at level 2, whether you are using our pro products or AI third-party solutions, or a combination of the two. Is this what $10 trillion of AI build-out is all about? Is the end state a better chatbot? Is the Holy Grail a voice agent for every business? No. Do not get me wrong, these solutions are certainly very valuable, just not $10 trillion valuable on their own. They are the beginning of the transition. They are not the destination. The magnitude of this investment only makes sense if AI can completely revolutionize how we operate a business. A business that runs itself, and a business that gets better every day.

That is what this is all about. Of course, to get there are a couple more levels we have to cross. Level 3, where we have far more automated workflows in the business. Level 4, where those automations become coordinated. Finally, level 5, where the system learns and improves itself after every decision. Do not let the sequential numbers fool you. The levels might go 1, 2, 3, 4, 5, but the value jumps exponentially at each step. A business that runs itself and gets better each day is far more valuable than one with just a few automated workflows. So what does this mean for a contracting business? What does it actually look like to be a level 5 contracting business? The mechanics, of course, will be different for residential versus commercial, for service versus construction. Here, I am going to use residential service as the example.

For my commercial and construction friends, you will see your journey in the commercial keynote and in Vyas' keynote later today. For exteriors and property care, yours are in your trade-specific breakouts. I encourage you to watch for the patterns. The workflows might be different, but the problems and lessons underneath them are not. That's what the next 20 minutes are all about. At a high level, this is what a residential service business looks like. You generate demand, book the appointment, sell the work, do the work, manage inventory and payroll, and then you analyze and optimize to repeat the loop again. This isn't just your workflow, it's also your business model. How many leads you generate, what percentage of them you turn into booked appointments, your average ticket, these three things determine your revenue.

Marketing spend, office labor, field labor, and so on, they determine your cost. How well you optimize this loop determines your profit. This might seem very simple on paper, but under the hood, there is a lot going on. Just think about all the things you need to do in order to generate lots of leads. You have to be great at the highly competitive world of paid ads. You have to be great at email marketing to your existing customers. You have to immediately respond to leads that come in through the lead marketplaces. You have to be great at having an active social presence. You have to build and maintain a stellar reputation of reviews, and soon you're going to have to figure out how to compete in the new world of ChatGPT ads.

To have high booking rates, think about all the things you need to do there. Your CSRs have to handle each call perfectly. You have to figure out a solution for off-hours and overflow. You have to convert customers who prefer to text or those who prefer to book on your website or through Google. You have to prioritize capacity for high-value opportunities, and you have to nurture these appointments so customers don't cancel. Then in the field to have great average tickets, you have to constantly reshuffle your dispatch board to get the right tech to the right job. Your techs have to be great at presenting those good, better, best options. They have to offer financing on everything. You have to follow up on every unsold estimate, and you have to coach technicians to improve performance every day.

That's 18 things you have to be great at every day. For two decades, we have been talking about generating leads, optimizing booking rates, and increasing average tickets. We're still talking about generating leads, optimizing booking rates, and increasing average tickets. Your teams know these things matter. We have simply been fighting two unbeatable forces, time and speed. How can a marketer optimize your ad spend when capacity changes hourly? How can a CSR respond to texts while on the phone with a customer? How can you review every recording to coach CSRs and to coach technicians every day? There simply isn't enough time. In an increasingly digital world, human speed is simply not fast enough. The only way to be great at everything is to automate it. This has always been our vision.

Owners and leaders focused on outcomes, technicians focused on serving customers, ServiceTitan increasingly automating the complexity in between. AI is making possible what would have otherwise taken us another 2 decades to build. Last Pantheon, we partnered with 50 pioneering customers to build the self-running, self-optimizing business in the trades. I want to share with you this journey, I want to share the lessons we learned going through it. First, achieving level 3 required that we automate more workflows in the business. So we built an agent for each of the workflows I just talked about. We started where missed opportunities were most evident. That was appointment booking, particularly calls arriving off hours, calls arriving when CSRs were already tied up. We built a voice agent to handle those calls, one you could set up in minutes instead of weeks.

It was okay at first, then it got really good over the ensuing months. We wanted it to book as well as a good CSR. We couldn't rely on the booking rates inside ServiceTitan because as you all know, those are very easy to game. So we built a voice intelligence agent that'll review every call to determine if it was a bookable lead. Today, our voice agent books as well as a good CSR. Good enough that when a CSR leaves, many customers choose to let the voice agent handle the overflow instead of scrambling to find the replacement. Some even let the voice agent handle 100% of their phone calls, with CSR standing by to handle escalations that need a human touch.

Customers asked us if we could extend this technology to help CSRs improve their own booking rates, so we expanded voice intelligence to grade how well CSRs followed your call process and coach them to increase those booking rates. Next, customers saw an uptick in text messages. They asked us to automate this channel because response times were often long, customers would end up calling competitors. So we built a text agent that immediately responds and more naturally and professionally than any human has time to type out. Online volume increased, too, so we upgraded Scheduling Pro to improve its conversion rates, we built a web chat agent for customers who wanted to book through chat. Interestingly, we initially thought these web chats would cannibalize Scheduling Pro bookings. It turns out they captured net new appointments that contractors were previously missing.

Now, with booking covered, we turned our attention to demand. For paid ads, Marketing Pro's Ads Optimizer already helped Google improve targeting to increase ROAS. We made it even better. We built support for multi-touch attribution. We built support for Meta Ads, then we built an agent that experiments with new ads as well. For email, we had Marketing Pro's email, but customers were underutilizing it because it required manual work. So we built an agent to automate many of the campaigns you'd typically run, cross-selling services, seasonal campaigns, weather events, so on and so forth. We then built a speed-to-lead agent to immediately respond to those leads you get from Yelp, from Angi, from Thumbtack, from Facebook. Then we created a social publishing agent that automatically posts content to the major networks for you. With demand solved, our next focus was average ticket.

With more jobs on the board, it became even more important to assign the right tech to the right job because you ultimately have a finite number of technicians. We first upgraded Dispatch Pro to consider every data point about the job, the ad that generated the lead, what was said on the call, everything we know about the homeowner, the home, the equipment. Dispatch Pro started making more intelligent decisions. Next, we built an agent that automates those good better options, good better best options for the technicians and the comfort advisors with integrated financing in each proposal. We create an agent to automate estimate follow-ups, and then we rebuilt our sales coaching agent to be native inside of ServiceTitan. No more switching apps, no more forgotten recordings, and greater engagement from technicians.