สำเนาบทสนทนา 🔗
speaker-0: Hi everyone and welcome back to another episode of Broadband Breakdown. Today we have Ty Sorensen here. He's Vice President of Sales and Marketing at ISPN. Ty, welcome.
speaker-1: Thank you having me. How are you? I'm great.
speaker-0: Good. Now tell us about where we're at currently.
speaker-1: We are at NCCA Fall Conference twenty twenty six in Nashville. Yes.
speaker-0: Or at the Gaylord, it's lovely here. So we've been listening to sessions. I feel like AI is always an omnipresent theme, right? And you and I have talked a little bit about this. Tell me a little bit about kind of your background and what you're interested in and kind of what you do at ISP.
speaker-1: Of course. So yeah, in the vice president of sales and marketing. So I lead our our go-to-market at ISPN. My background, I've been in sales my entire life since I was a teenager, back starting in like retail sales, and then of course got more into the traditional type sales I'm in now. When I was in my mid-20s, I took a minute to like reflect and think about what I wanted to actually do with a career, and I decided I either wanted to start my own company or I wanted to lead a company. As I thought about that. I assessed what kind of skill set I need to be able to do that. And then I put in motion a lot of different career changes and moves to to go learn all of those skill sets. So sales worked very, very close with marketing, customer success, customer support, built-out customer support teams. I think the big pivot that was beneficial for me was getting into revenue operations, really learning the infrastructure, data and analytics. process design and really from, you know, lead acquisition through adoption and customer retention, how to build infrastructure and architecture to ensure that you have seamless handoffs between departments. The left hand knows what the right hand's doing. I was very fortunate at at one of the companies I was at to be sent back to school at University of Oregon. nice. For data and analytics. So I learned how to code in Python and SQL and build APIs and A lot of that from an execution standpoint feels kind of useless these days with AI, but I find it extremely helpful to understand how everything works and fits together from a platform standpoint, from a build standpoint. And and that's been really helpful in my career. And spent a lot of time working with finance teams at all the all the companies I've been in, learning F P and A and how to read a P and L an income statement and you know you Looking at the unit economics of the business so that you can make decisions around where to invest, where to pull back, how to drive the business forward and ensure you're growing. Right. 'Cause that's always the the goal. That's
speaker-0: goal. Yeah. How did you start? I mean, when you're basically unpeeling the layers, right, and trying to understand from a business standpoint, how can we insert AI with purpose and how can we do it intentionally, but where it's maximizing everything you're doing and then for your customers, walk me through, was there an AI audit or what did you all do?
speaker-1: Yeah, exactly. We started to tackle it by ourselves, I think, a couple years ago as AI became more readily available and everybody was starting to use it. And then at some point we said, hey, let's bring in a consultant. Okay. Someone that knows this technology inside and out. And the first thing we did is we had them come in and give a little bit of like a test to every employee in our organization. And the two outcomes were of of that were ideas and areas in which we can improve the business. Nice. But also who shows real proficiency and interest and intrigue with AI and those were the people that we pulled in to build an AI working group and an AI council. Okay.
speaker-0: Okay, fabulous.
speaker-1: That we run at ISPN with members of the executive team all the way down to technicians. Okay. so so every every member of of staff on there, there's about
speaker-0: Talk to more about your your committee. I mean, how often do you all meet and you know, I don't know, I feel like everybody rolls out AI things and it's kind of go use it or go figure it out. I mean, do you share tips and tricks or how are you getting the word out and how do you utilize the committee and then get convey that message out to the broader company?
speaker-1: Yeah, so first thing we did is we built a charter. Okay. Right. We wanted to align, make sure we're all speaking the same language, set expectations on what this working group was intended to do. once we got there, we started doing a bi-weekly meeting. So every other week we get together for an hour. That has been a little bit fluid and sometimes it's talking about a specific project we're working on. We always have a B HAG, right? Like a big, hangry, audacious goal that we're trying to go after and tackle with AI. But often some meetings we're even just talking about headlines in the news and are we are we making sure we have guardrails in place to protect the business, to protect ourselves, always having having a contingency plan if we need to roll something back. Sharing tips and tricks. one of the best things we did is we created a Slack channel. I love that. Everybody in the organization, like truly everybody, and just said, Hey, this is our channel. If you have a fun thing you're working on, if you have a tip a tip or a trick that you want to throw in there, throw it in there. And you'd be surprised at the contribution you get and where it comes from. It's it's not always leadership, it's it's the people that are doing the work every day taking a lot of initiative and saying, Hey, I tried to build this thing and it actually worked. And we've taken some of those prototypes and put them into production to help better that that we sourced and You have to have guardrails and make sure that you don't just give everybody free reign to vibe coding. Right. Right, 'cause
Load full transcript
that can get a little bit out of control. It's hard to have governance over that. Right. But with the proper guardrails and and communication channels, I I think it can be executed.
speaker-0: Love that. And I think it, you know, it applies to so many companies. I think BSPs could adopt a similar strategy, right? They could come up with an AI policy first and foremost. I think some do, we're hearing as we kind of walk around the conference, some don't. Which, you know, if if you don't have that policy set in place. We don't know what the goal is with AI and then departments are being told to use it, but you don't really know why. Right. And I love the AI committee and I love that you have people in leadership and you have your doers and kind of top down. That's so interesting. So what where do you think the opportunity is here for BSPs with AI? Like things that they should plan to execute over the next six to twelve months with AI.
speaker-1: I think I would start with foundation, right? And AI is still a tool that a lot of us are trying to figure out, right? It's it's the first time I think in our generation of having a new tool set that no one has had previously. And so finding a true expert is still sometimes a bit challenging. Right. And it takes a lot of investment to keep up because the stuff is moving at lightning speed. So fast. Yeah. I go back to what Mike Romano said this morning, which is like change is coming. And you have to adapt, you have to evolve, right? there's massive amount of MA going on, increased competition, funding, et cetera. You name it, a lot of things are going to change from policy and and and legislative component. I think that should be an inspiration for AI as well. Is lean into analyzing the processes and procedures that you've been using for years and years without batting an eye at because they work. And start to look, like we said, top down at should we really redesign this process? If we were going to start this business today, how would we build it? And it's most likely different from how it is built today.
speaker-0: That's a great point.
speaker-1: And as you do that and start to build out processes, we now have this new tool that we can put in there, you know, in AI for automation or agenc work or you know, whatever it might be. But I think I would start there, is is before you even put AI in the process, start or in the motion, start with your process. And long term that could lead to complete org redesigns. you know, whatever it might be. But I think that's a good entry point where you kind of naturally find where AI can fit into your existing workflow.
speaker-0: Agree, and I think you need that backing from leadership too. You need the board to be on board. You need your exec team to be on board. Otherwise it falls flat. Agreed. so one thing I heard it we heard real BSTs talking about today how they were utilizing AI and someone said, you know, all this grant money is coming in or the funding. And basically now they can look at their options and kind of run it through AI and just do like a cost analysis before they model out for everything. Yeah. So basically it helps them narrow down and get smarter and then they know their their path forward, right? So you and I are both on the marketing side and you have your sales as well. But talk to me about just the importance. All of our BSPs here that we're talking to, all of our customers, all of these prospects, everybody in the room, like Mike Romano mentioned, they're dealing with unprecedented unprecedented change and I think more competition than they've ever seen before. Talk to me about the relevance of AI now in go to market. What does that mean? Why do people need to start acting and planning now and how will it help them for the future?
speaker-1: Yeah, I think when I look at the entire go-to-market motion, I think what you teed that the example you teed that up with is spot on, right? It allows us to make decisions much more quickly. Yes. To analyze information much more quickly and and make smarter decisions. you know, I think about just 24 months ago, looking at sourcing and attribution and spend and you know, all of these metrics that would take these really complex linked Excel spreadsheets that we would take, you know, days to build to get them built up. And then you could just update them pretty easily. Now you can do that analysis in like two minutes. Yes. Right. Especially with the MCP connectors. Yes. Plugging directly into source data. Now something that we always talk about is never trust, always verify. Right, because we have seen a lot of instances, myself included, where we're moving real fast and you get a that AI slop that comes out. Yes. And whoops, I forgot to check that and it's not accurate. Correct. We don't want to make multimillion dollar decisions on inaccurate data. So I think you have to slow down a little bit and just verify, make sure everything's working correctly. Mm-hmm. But outside of that you still save massive amounts of time. Massive amounts.
speaker-0: Some time.
speaker-1: And and I think that's a a great use of it in GoToMe.
speaker-0: I love that. Anything else you can tell us about what your customers are doing and how they're utilizing, you know, AI and what they're thinking about or what are they coming to you with and saying, hey, we have this problem. Can you help us fix it? And maybe you use AI, maybe you don't.
speaker-1: I think a lot of providers we work with a a little over four hundred and fifty providers. Okay. So we have a lot of relationships and conversations daily. A lot of people are doing the basic task automation, those things that I think everyone is doing. Mm-hmm. There are a few though that I think are really innovative and forward facing and trying to do things like working with us on the knock side of the house to assess data and say, Hey, can you help us answer through text prompting or multi-tenant agents? What's
breaking, why is it breaking, and how do we prevent it from breaking in the future? Right. And instead of using data for as a lagging indicator, trying to use it to predict the future and say, you know, hey, if you go replace this equipment in this neighborhood, you're gonna prevent, you know, X amount of dollars lost in truck rolls over the next twelve months.
speaker-0: that and the proac the proactivity instead of the reaction
speaker-1: Exactly. So we're really focused on trying to work with providers to solve those types of problems. And it's challenging. We we haven't crafted it quite yet, but we're getting pretty close and getting closer every day.
speaker-0: So you personally, how have you util or how has AI helped you? Like how do you use it where maybe something recent where you feel like, light bulb, this is it this would have taken me so long, or how has it helped you in your day to day?
speaker-1: In many ways, but the the example that comes to mind is actually pretty basic. And this is something that I think I would advocate to think through as you use it is like you don't always have to build these multi agent complex systems. Sometimes the biggest ROI is something pretty basic. So for me, I travel a ton. I'm in meetings usually nine to five every day, maybe thirty minutes or an hour block. And so I really struggle with keeping up with email. Yes. Slack messages, you know, all of us with all of these different channels are getting bombarded with information and and questions all day. I built we use Claude. I built a skill in Claude that I can run every morning. And it queries my email, Slack, HubSpot, all of the tools that we use. And it basically gives me a breakdown of a snapshot of the last 24 hours, tells me through prompting that I built into it, what I should be prioritizing. who I have forgot to respond to on an email that's like needs to be responded to today. And that once I got it trained and like dialed in, has been an absolute game changer.
speaker-0: I like what you said there. I once you get it trade, I mean there is some there's a still an iterative process, right? AI is not going to be perfect and it's only as good as the prompts that we give it. So how have you worked with your teams on the importance of prompting? Yeah. And the knowledge behind that and how to prompt better.
speaker-1: Yes, so we've built some frameworks. Okay, great. So I think prompting does not come naturally to everyone, right? So we've built frameworks so that there are really templates that you can use to prompt and make sure it has all the right information. The other thing that we've done is really adopt a mindset that if you're building agents, for example, or like this skill that I talked about, right? Mm-hmm. You build it, you launch it, you're gonna find gaps in it. If you Adjust your mindset to think of it as a human employee, just like you would coach and say, hey, this is what we talked about. We kind of missed the mark a little bit on this one part. Let's iterate and improve. And next time, let's get a little bit better at that, right? Treat these AI tools with the same mindset and know that like you'll have to coach them and get them to improve. And then you'll see the inputs or sorry, the outputs get better and better as time goes on.
speaker-0: So I have an important question, or maybe not important, but I think it's fun. Have you had anybody that's very reluctant to adopt AI and how do you handle that?
speaker-1: Yes. Okay. I think we've had a handful of people and I think it's become less and less as as a society we've all kind of said we are using this thing moving forward. Sure.
speaker-0: It's here.
speaker-1: but I think you know, there's reluctance often in creative sides of the house, right? Whether it's could be marketing, could be software engineering, people that write code, saying like, hey, I've div worked on this craft my entire life. And now I'm being told to hand it off to a machine. Yeah. And that's insulting. That's
speaker-0: Yeah.
speaker-1: Yeah. And so I think it's been, you know, take it back to the beginning. I think it's been a little bit of an exercise of let's put all the all the puzzle pieces out on the table. Let's start to sort through it and figure out where this can be a tool that helps you save time. And I know we've had folks, you know, on the marketing team that have saved hours and hours of time in video processing, right? that was a manual process before, then now AI can help that. And it's like boom, five hours back in your day after a shoot to not have to sit in front of a computer and scrub through you know, all the video. Right. So I think it's just finding, you know, trying to number one lead from the front, bring the team along, let them know that you know, this can be a helpful tool and work with them side by side to help them figure out where it can help.
speaker-0: I love that. And just to recap, I love I love the idea of creating a committee and having those i stakeholders from across the company. I haven't really heard of that. I wonder if VSPs are doing that. So I would love to hear, you know, if if we have VSPs out there that are doing this, I want I want to know more about that. thank you so much for all of your time today.
speaker-1: Thank you having me. No, I would just say if anybody has questions or wants to talk through and brainstorm, I'm I'm available. So
speaker-0: Okay. Awesome. Well thank you so much, Ty. Thank you. Appreciate it. Take care.