GTM Navigator: Implementing AI in Your GTM Strategy

GTM Navigator is our ongoing series where we break down the essential components of nailing the right go-to-market strategy.

In this episode, Lauren Houpt, Investor at Fin Capital, sits down with Jessica Pely, Co-Founder and CEO of Loyee.ai, to explore how AI is transforming go-to-market models, from automation to precision targeting, and how startups can thoughtfully integrate AI to drive meaningful outcomes rather than noise.

Drawing on Jessica’s experience as a fintech founder and technical builder, the conversation breaks down topics such as how to evaluate whether AI truly adds value to your GTM strategy, what “precision over volume” means in practice, build vs. buy decisions for startups and enterprises, and more!

Key Questions Discussed:

• How has AI shifted within go-to-market strategies over the past few years?
• How can startups assess whether AI is a real value driver or just a distraction?
• What does “precision over volume” look like in action?
• How should founders prioritize AI internally across sales, marketing, and success teams?
• What level of data infrastructure should be in place before adopting AI?
• How can teams ensure AI integrations stay adaptable as technology evolves?
• How should companies think about build vs. buy when it comes to AI tools?
• What are the most common mistakes startups make when layering AI into GTM?

Transcription (edited for clarity):

Lauren Houpt (00:00): Welcome to Go-to-Market Navigator by Fin Capital. This is our ongoing series where we break down the essential components of nailing your right, go-to-market strategy. I'm Lauren, an investor here on the Fin team and excited to be joined by Jessica Pely today. She is the co-founder and CEO of Loyee.ai, and we're going to chat about all things AI within your go-to-market strategy - how AI is reshaping go-to-market strategies, some best practices to think about, and some fun things to keep an eye on for your tech stack. Thanks for joining us today.

Lauren Houpt (00:48): Well, thank you Jess, so much for joining us. I'm super excited to chat today about how to implement AI into your go-to-market strategy. To begin, we'd love to hear a little bit more about your background, what you are building, and then we can take the conversation from there.

Jessica Pely (00:51): Thank you for having me here today, Lauren and team. I'm super excited. My background is in finance, so I'm a financial economist and founded a fintech company. I got my PhD and was working mainly on topics like how to sell to hedge funds and banks. Then, I founded a FinTech company called MetaWealth. It was like a fixed income product for crypto. With that in place, we were looking to optimize our own go-to-market strategy. Since I was the technical co-founder, I faced a lot of struggles doing that. So why is that? For example, you have so many banks, so many hedge funds that you can sell to, but at the end of the day, I wasn't interested in the 10 thousands that are out in the US, for example, but I was interested in the ones that have crypto exposure, that have a risk management department, but that are too small or are thinly stretched. I developed a tool that helped us prioritize those banks and hedge funds. Fast forward, the FinTech company shutdown unfortunately, but this new company arrived, which is now a go-to market tool, an AI tool called Loyee, and we work a lot with fintechs - helping them with their go-to market strategy.

Lauren Houpt (02:07): Amazing, thank you. That makes total sense. I guess to just start off the conversation, I would love to hear your perspective on how you've seen AI shift over the past few years, whether that's in your go-to-market strategy or not.

Jessica Pely (02:29): Obviously before the LLMs, AI was always a hot topic, and I was working a lot on machine learning and AI. But it was never sexy to anyone to be honest – at the time, not to investors, not to companies, because it just required a lot of investment. However, with the LLMs and OpenAI and so on launching, the appetite just increased tremendously. It became more accessible and now it's about finding the right way to implement it. For one, you have obviously the bottoms up approach. Everybody's using ChatGPT for example, to help out with coding, with go-to market, but now aligning that with the goals of the whole company and streamlining that as a full process end to end, I think that's the challenge. And that's where a lot of companies now have their efforts focused, I would say.

Lauren Houpt (03:23): Definitely, and to really dive into our first question, a lot of our viewers here are going to be startup, they’re going to be a range from seed to building to all the way up to scaling. That said, curious how startups should start to evaluate whether AI truly adds value to their go-to-market versus just being a distraction or a buzzword?

Jessica Pely (03:43): Yeah, I mean every company's different, right? And every go-to market motion is different, especially depending on stage. What I always try to summarize is AI can help with three things, and the easiest is to start with it can automate repetitive tasks. For one, I would say any small company, even seed stage, can automate those repetitive tasks. Then you have, hey, AI can bring you true insights into what you're doing. I would say that's the next step, number two. Once you automate something that is repetitive, then you can add the next layer on top, which provides the right insights and helps youto make better decisions. And the last, but I would say that's when you really scale is automate the decisions, after you trained it, you're confident that it's working, it's producing the right insights. So now, hey, let's help us make decisions for us with us. For an earlier stage company, they should really focus on automating the repetitive task because the teams are also smaller and not worry so much about the insights and how it feeds into the decisions. For the later stage companies that usually have already a solid base of people, processes and tech, they can really start with the insights while training the AI to eventually also automate decisions.

Lauren Houpt (05:11): I think in our previous conversation, the one prepping for this conversation, one thing that stuck out to me that you had called out too was that “precision over volume” piece as well. I would love it if you could touch on that a little bit too and just add some more context around precision over volume.

Jessica Pely (05:29): I would say the precision over volume gets especially important with AI because what you want to avoid is that you leverage AI to create MOS spam. So that's pretty much it. You can really use and leverage machine learning models and AI to really figure out who you should be going after, for example, and why, who you should be focusing on and why. So it's way more about efficiency than spamming, and I think that's very important that when you start implementing AI that you really focus on, okay, let's not do more and more and more, but create value with less and let's understand what is the precision, what is the less that we need to focus on? Because then you can leverage AI for that and still have the human in the loop. We work with many fintechs, especially in the B2B space.

One example is payments providers, they can sell to anyone in the market, to all businesses, to all freelancers. They could use AI to write so many emails, but there is a constraint, right? You cannot send millions of millions of emails, you cannot send millions of millions LinkedIn connect requests, and the fintech space is still very human focus; you want to build relationships. Now you can use AI to really help you figure out in the huge market in the world, who are my next best prospects I should build relationships with? I think bringing that mindset to the table when evaluating AI is important.

Lauren Houpt (07:08): That was perfect – the extra context helped a lot. I think it leads to another question; now that startups say they have implemented some AI, they're understanding how to use it in their go-to-market strategy. Now, how should founders try to put together a criteria on how to prioritize AI internally? Whether it's through their sales teams, whether it's through marketing or customer success, what criteria should founders start to think about when it comes to AI?

Jessica Pely (07:35): What we see working very well are three key pillars, one is to focus on your north star metric. What are you focusing on this quarter of this year? That can be different obviously from startup to startup. For some startups it might be customer success. We really want to improve our churn rates, so let's start with that piece. How can we do that and how can we bring in new technology to help us out there? So that's number one. Number two is really this efficiency aspect, especially if you are a younger company. Where's your team wasting a lot of time on manual tasks, which oftentimes is especially repetitive tasks (because those tasks are easier to automate and easier to train). If you're always doing the same on, let's say top of the funnel side, finding the best prospects, finding the best events to attend whatever it is, that's usually always the same.

It doesn't change much from prospect to prospect. That might be a better way to start. Then the third pillar is looking at your full flow, map it out - how does your funnel look?It's always easier to start at the beginning. What is your ICP, right? How do you prioritize your ICP? Because everything else is dependent on that. Let’s say you are a payments provider and your ICP, you're selling payments for, I don't know, travel bookings and you define that as your ICP, then obviously it should start at the beginning of the funnel because you know your ICP, you want to qualify those leads in, you want to book meetings with those leads, you want to close deals with those leads, you want to onboard them easily and smoothly. It just makes sense to start at the beginning of the workflow for some because everything is dependent on that.

Lauren Houpt (09:36):That's super helpful and I totally agree. I think a question that comes to mind when you talk about the workflow, going back to earlier stage companies, what type of data infrastructure or internal hygiene should they be thinking of or that might be actually required before they start to implement AI to ensure that that AI can be effective in their go-to-market strategy?

Jessica Pely (09:58): It depends always on the goal, but there should be something in place already. I would say you should already have a few customers. With that, you have people, at least the founders who are working on deals, you have some sort of tech stack that you're using to help you out, like a CRM system or an outreach system to automate your LinkedIn messaging and some light process in place so you are clear already. You know which is the best way for us to get a customer, a lead, into the door. So, something should be already in place. Then, AI, think of it as an enabler at the beginning that helps you scale that. Because if you don't know what you want to go after, if you don't know what your goal is, it'll be just very difficult for the AI to do any good job, to be honest and don't expect magic from the AI.

Think of it like an intern or a junior hire that even as a founder, if you want to delegate something to this junior hire, that's what you can delegate to AI. You still need to coach it, you must train it, you must be patient with it, it can make mistakes. When you're ready to coach, train, and improve, you can get amazing results. But if you're not and you really don't know what you're doing yourself, then I would say, it's a waste of time.

Lauren Houpt (11:32): Yeah, I agree there, and I think it's been fun here at Fin Capital, we've been trying out new things and implementing AI into our own internal systems. We think we have a Director of AI here, Fan Wen, and he's amazing, but it's been fun to learn from my own perspective on what's working and not and to be patient with different models.I totallyagree with you there. Going back to the beginning of the conversation when you gave an overview on how AI has transformed within the past few years, obviously it's going to continue to evolve very rapidly. The question there is how can startups and companies, regardless of scale, really ensure that their AI integrations are going to be able to stay adaptable as AI does rapidly evolve?

Jessica Pely (12:21):The AI evolves, but obviously you as a company evolve. I think it's still important to have the human in the loop, and who is responsible for that? You do it internally, you have someone that is taking care of it. I think it'svery important to have this human in the loop to give feedback, double check things, and continue to improve the systems. Also at the least, every quarter review what you have built, if it's still working or not working, ifit's generating the results and going into the right direction. Really put in a block into the calendar to review. Ask, what does it bring us? Anything? And obviously experiment - never stop. If you implement something and it's working, be hungry for the next because you must keep going as it improves. For that, it's always easiest to have a system where you collect all the data so you have a single source of truth where everything collects. Some people, it might be a database or someone is using Snowflake, others use their CRM, but somewhere where it's all collecting nicely and updating nicely. I’d also say work with APIs and integrations and ensure that the workflows are end-to-end automated because then you can also optimize on the integrations and the workflow; that's something that companies oftentimes forget.

Lauren Houpt (13:47): I think that brings us to another big question that a lot of companies, even large enterprises and banks that we're chatting with on a daily basis have the internal debate around is, what do you build versus what do you buy off the shelf? How should companies, regardless of size, be having those conversations internally? Curious about your perspective there because it can be a tough decision.

Jessica Pely (14:10):Yeah, it's a tough one because I agree with you, right? Do you have a buy versus build culture? I think what we see as well is that enterprises prefer the build culture versus the buy culture, and startups are rather focusing on the buy culture. This is simply because it is costly to develop something, but sometimes it's not even about the cost. How many team members do you need that have to develop something? Do you have the expertise? I mean, why reinvent the wheel when things are already out there? Also, AI needs to be trained on certain models need to bring in and consolidate all expertise. So, it just takes time to build up this knowledge hub. I would say when you start from scratch internally, you might have a great knowledge hub and execute on it fast, but there are so many companies that are doing that on a day in day out basis. They truly understand what is working, not working, and you'remissing out on this experience - it'lloften take you longer. What we see is when you build internally, it just takes you longer. If speed is important, then I would say it's easier maybe to delegate that.

Another big question to ask is what do your teams want to work on? What we see also for startups, delegating it, and instead of building it themselves is because their core engineers and AI data scientists, they don't want to work on internal tooling. They want to work on the core product of that company. They don't want to be working on a sales tool If I ask our engineers to do something for my internal needs, they roll their eyes, right? They're like, we have so many bugs in our product, or hey, we want to focus on the features that help us provide value to our customers. Why should we not care about your problem? I think that's something that as a startup, you're just way more pragmatic.

Lauren Houpt (16:20): Yeah, I totally agree there. That's a great call out. Moving into another question, curious from your perspective and all your experience, what are some of the common mistakes that you see startups make when they are starting to layer AI into their go-to-market process?

Jessica Pely (16:35): A little bit back to what we discussed earlier, they think that AI will solve their problems, but you must solve it first yourself. Once you solve it, then you can delegate it to an AI. I think that is something that we see oftentimes with companies is, for example: okay, we have so many bottlenecks and we want to improve sales conversions, we want to prevent customer churn, we want to book more meetings, we want to have better customers that are easier to onboard. Can AI do it? They think that the AI can now do everything but think of it when you would try to hire someone to help you with that job. You hire people for specific departments, for specific responsibilities, for specific functions - the AI is very similar to that.

It's trained for a very specific thing and it's not yet a human that you can retrain to do something else. If you hire an SDR, they can maybe shift over and help in customer success with some things. However, the AI needs way more time to be trained. That’s number one, don't expect the magic to happen. Solve your problems first, and once you understand how to solve them, it's super easy to onboard an AI tool. Number two mistake is chasing something groundbreaking that is new, that you have never done before. Focus on what you're doing, where you had learnings, where you identified things going very well, and try to double down on that using AI instead of doing something completely new what you have never tested before, because then it just gets super expensive and aligned.

Number three is across functions, but also across hierarchies, I think that's the crucial point. Usually, AI helps the more junior folks, but it needs to be aligned with the more senior people and then also within the contrasting functions. Everybody must be aligned. To give you a simple example, and this is usually starting around series A when first functions start to develop, like you mentioned, sales, marketing, customer success when we came in as a tool, and I can share a little bit more what we are doing and just ask the simple questions, what are your top customers? Each department separately or even each person separately, they would produce 10 different answers. Sales and marketing would also be different. Sales says, “oh, the best deals are closed.” Marketing would say, “oh, the best leads we brought to the table are these 10.” Customer Success says “oh, the best customers we work with are these 10” - and they were all different. Aligning, why are they similar, discussing together on why are they the best? What makes them a good prospect, a good customer? That is super, super crucial.

Lauren Houpt (19:37): I would love to hear more. I think the segue is great into just hearing a little bit more about what you're building and how you're solving that problem.

Jessica Pely (19:44): What we are building is Loyee.ai, it is a set of AI research agents on your market, on your prospects, and it finds the best customers for you. Think of it as, instead of reps doing research manually on who they should be going after and why, the AI now does it for you and it's trained basically on your products and on who you have sold to before, it always finds your next best customers. This is working quite nicely because now your sales team or marketing team can really focus their efforts on the best prospects in the market, and because you have limited resources, by focusing on the best prospects you can more quickly penetrate the market and grow faster and way more efficiently.

Lauren Houpt (20:35): Amazing, thank you. That's super helpful. I guess the final fun question, would love to hear about some of the other companies out there that either you all are using internally or that you have heard of, just an example Tech Stack. I would love to hear your perspective on some emerging folks in the space.

Jessica Pely (20:53): On our go-to-market side, we have a bunch of tools. We first had HubSpot as our CRM and have transitioned to Attio. I love the ease and simplicity and that it's AI first. For top of the funnel, we use Loyee.ai, our own tool obviously. We make sure that we only focus on the right companies and the right people. Then this gets into Attio. During our calls we use Spiky for voice recording and sales enablement. What I can also recommend, which we cannot use, but if you are a Salesforce shop, use Momentum because it completely updates your CRM and helps with your ops. I think when you use Attio you do not really need that, but if you are a Salesforce shop, that is super helpful. Then we obviously use a lot of ChatGPT and the APIs from the LLMs.

For contracting, we just onboarded Atlas, so they help us close the contract and feed everything into the CRM. I would say we are also an early-stage company, so that is what we have in place right now beyond those tools like website traffic – to look at who is hopping on our website, trying to anonymize it. We use Bombora to understand who is researching account research or lead generation. We have a bunch of tools that we are using.

Lauren Houpt (22:39): Awesome. I always love hearing what other tech stacks companies are using and where we might overlap, so that's super helpful. Well, Jess, I really appreciate the conversation. To wrap things up, do you mind letting folks know the best way to get in touch with you? Obviously, we will have your LinkedIn, and everything attached here, but if there is an email, if they want to reach out directly, we would love for you to share.

Jessica Pely (23:00): You can either email me at [email protected], or just simply find me on LinkedIn. Something that I also want to highlight is we have an amazing network in New York and in the Bay Area as we are hosting every month, go to market round tables for executives and leaders. If you want to learn from others and network with others around those topics, feel free to join us – they are happening in New York and San Francisco. It’s always a wonderful crowd.

Lauren Houpt (23:32): Awesome, thank you so much. Well appreciate the conversation and thanks for tuning in.

Jessica Pely (23:37): Thank you.