Magnify’s Josh Crossman on Incubating a Startup and Bringing AI to the Customer Lifecycle

Magnify Founded & Funded, Josh Crossman and Investor Elisa La Cava on the customer lifecycle

Welcome to Founded & Funded. My name is Coral Garnick Ducken — I’m the digital editor here at Madrona. Today, investor Elisa La Cava is talking with Josh Crossman, CEO of Magnify, which was incubated at Madrona Venture Labs. Josh was actually recruited by Madrona Venture Labs to help launch a business that would bring AI to the customer lifecycle. He was quickly persuaded and signed on in 2021, spinning out of MVL in less than three months with a $6 million seed round.

In today’s episode, Josh and Elisa discussed the incubation process and how Magnify’s intelligent application to improve customer retention, expansion, and adoption is solving one of the biggest pain points of the customer experience — focusing on and understanding the needs of every single user. They also dive into finding product-market fit and the importance of incorporating AI/ML into a product from the very beginning. I’ll go ahead and hand it over to Elisa to dive into all this and so much more.

This transcript was automatically generated and edited for clarity.

Elisa: Well I’m so excited to be here today I’m Elisa La Cava — I’m an investor on the Madrona team, and I have the distinct pleasure of having a conversation with Joshua Crossman. Josh and I have known each other now for a little over a year, since we invested in Magnify’s seed round. And today’s topic is all around very early-stage company building. Josh, you have been in the trenches, literally going from zero to one, building a company from ideation to launch and there are so many incredibly meaty topics we can cover, and I’m excited to dive in. But first I just want to give you a chance to introduce yourself.

Josh: Yeah. Thanks Elisa. So I’m the CEO and Co-founder of Magnify, the operating system for the customer lifecycle, and I’m really excited to be here and — let’s dive into it.

Elisa: So the customer life cycle, that’s a like a big statement, right? Everything, especially in today’s economy, it’s all about how do we make sure our customers are happy. In software that means are they using our product the way that we wanted them to? Can we get more folks to use the product over time? What does it look like years down the road when you want them to continue to be happy, thriving customers and utilizing all of the new products and features that you’re building? Sometimes we call that process post-sales orchestration, which can be a mouthful — how do you think about what that space is?

Josh: Yeah. The whole concept is exactly what you’re talking about. The challenge is that we’ve built these really amazing technologies in the software industry — particularly in B2B software, and enterprise software — but most enterprise software companies have real significant challenges with adoption and retention. And so what Magnify is doing is bringing automation, bringing AI/ML, bringing software into a problem that software faces it in itself. It automates and personalizes that customer life cycle. And fundamentally, that’s about understanding the needs of the user — what are their challenges and how do we automate and connect with them in ways that can transform their experience in really delightful ways? It’s a real shift from the way that traditional enterprise software thinks about working with their customers and working with their users. And we think it’s one the industry’s really ready for.

Elisa: Let’s talk about that shift. What is the status quo? Why does enterprise not get this today, and why is this a huge pain point that is worth solving?

Josh: Let’s start with kind of the current state of the state. Let’s say you buy a piece of business software — what typically happens is that you’re assigned a human being called a customer success manager. And this customer success manager is this wonderful human that is going to work with you over the course of your onboarding, your adoption of the product, ongoing usage, retention, your experience. This is the account manager for you that is really centered on making sure that you get value out of your software product. And I’ve run some of the largest and most highly regarded customer success teams in the industry. And I have nothing but love for customer success.

Elisa: Well, and it’s important, right? You are the voice and the face of that company and that software. You are the point person they go to when something goes wrong.

Josh: That’s right.

Elisa: You know, when they have questions about things they can do better. It’s like, this is the one person on your team that is always a go-to.

Josh: That’s right. Sometimes we call it “The one throat to choke,” when things are going bad. Usually, we like to say it’s your single point of contact that’s here to make sure you are delighted by our piece of software. And so this human being is doing great work, and we’ll call her “Sally CSM” and Sally CSM is talking to you every day maybe every week or maybe every month, depending on the nature of the engagement. The challenge is that Sally’s only talking to Elisa, and Sally has another 15, 20, 30 other accounts that she’s managing. Sally just doesn’t scale. Really the model and the industry has really been relying on that customer success manager working with a single point of contact and then that point of contact in turn driving adoption and usage out into the rest of that business, into that enterprise. And Sally never talks to the thousands of users at Elisa’s company. She can’t. She’s a human being with like finite time and finite capacity. And that just doesn’t scale. You can’t scale the personalization.

And with the rise of product-led growth, with the rise of user adoption, we think there’s a real change needed in the way that we think about it. We’re still going to have CSMs. We still need that. That’s absolutely essential. And we think that we need to start centering on the user. So much of everything we do in, in software and enterprise software is centered to the account level. And we have to get to a place where we’re thinking about every single user, no matter the size of the account, and understanding their needs, understanding their behaviors, understanding where they are and their adoption journey, understanding what are their excited and thrilled, and where they’re frustrated and understand how can we automate and personalize that experience for each and every user.

Elisa: This is exciting. I love this space. I love what you’re building. Let’s talk about what building a day-one startup feels like. You co-founded this company literally day zero to now with product launch. And you also had an interesting path that I want to talk about in how you did that, which was through a venture lab. When you started working with Madrona Venture Labs to build the idea for Magnify, what was your goal?

Josh: So, I’ve been in tech for over 20 years now. I’ve been a senior executive at a bunch of different companies, scaled companies, taken companies public, and as a GM I took a company from single-digit millions to almost a hundred million and profitable. The last one we scaled and then it was acquired by a public software company. So, where I was in my process after we sold the last company, I took some time off. I sort of poked my head up and said, “What am I going to do?” And I was in conversations with a late-stage data observability company doing some really interesting stuff. They’d just taken down a big round and they wanted to bring in sort of a COO/ president type to help run the company and work with the CEO founder. And I thought, oh, that’s great. It’s a perfect kind of role for me and let’s go do that. And then I get this phone call, and this phone call is from Madrona, and I know some of the partners and folks and, and they have this phone call says, “Hey, we’ve got this really interesting idea that we’d love to talk to you about.”

And I think that’s where the conversation with MVL (Madrona Venture Labs) starts. So, MVL has two models. One model is that you are a CEO/founder type that wants to come in and you’ve got this concept, you’ve got this passion. And if it aligns with what MVL is thinking about and how they think about the world, they can help bring you along on that journey. There’s a second type of idea though, that they have, and that is ideas that they’ve been thinking about and saying, Hey, we want to bring AI to finance, or we want to bring AI to travel, or we want to bring AI and just take AI for a second. We want to bring AI to the post-sales world. And so Moderna Venture Labs has been thinking hard about this problem, and then what they say is, I think there’s something here. We need to bring in a co-founder to help solve this problem, to help create the idea. We’ve sort of developed it, we’ve tested it with other companies, depending on where they are in that process can be relatively far along and the idea, and then they’re trying someone to help take it all the way.

And that was what happened to me. So they said, “Hey, we’ve got this thesis around bringing AI automation into the customer experience, into the post-sales world. Sounds like it’d be a fit given your background, Josh, why don’t we talk?” And I said, you know, “Tell me about the company.” And they said, “It’s pre-revenue…”

Elisa: it’s not a company, yet.

Josh: It’s not a company yet, it’s an idea.

And I said, oh, no way. That’s crazy talk. You know, that’s a young man’s game. What are you talking about? Uh, you know, I’m not that old. And I’ll tell you what, once I talked to them, I absolutely fell in love with the idea, fell in love with MVL and what they’re doing there, and got so excited. And, and you know, I think the process took about three weeks.

Elisa: Wow, that’s fast. That’s a quick turn. So you were instantly hooked. So how, how was that process? Tell me more!

Josh: So, in my case, you know, it moved really quickly. They already had a relatively strong idea so we took that idea, and you know, I’d say I put my 25 to 30% spin on the idea. Like, okay, let’s change this. We’re going to modify the product concept this way. Think if we add this capability or that capability, it could make a real difference. And then, MVL really just worked with me. So the way that Madrona Venture Labs works is that they have a set of functional capabilities and functional experts on the team.

So you may have a CTO type and they’ve got, uh, in my case, a guy named Keith Rosema. And Keith is amazing and super smart and talented technically. So, they’ve got sort of the engineering side. Then they may have a product side, they may have a marketing and sales and GTM side, they have a design capability and branding, and then some general management experience and all those things. And so in my case, what happened was we took the different functional capabilities and used them as we needed to refine the product concept to get the pitch right. I probably needed a little less of that and needed more help on the product side. So I’ve got a lot of product perspectives on what the product needed. But I’m not an engineer. I’m not a VP of product. And so what I really needed was the help, like, let’s get the mocks right, let’s get the designs right. Let’s talk about the backend architecture. How are we going to think about data stores? That allowed us to accelerate and move on the build right away rather than maybe being gated until I had a technical co-founder, until we had made some hires, until we had a bunch of other stuff. So we could accelerate that process in some really important ways and allow us to get to market much more quickly.

Elisa: Okay. Switching gears — I want to talk about product-market fit. In your case with Magnify, you had an idea, you spun that out, you did fundraising, you found a co-founder, you started building your initial team, you really went from zero to one. Like you had mock-ups that you were thinking about, and then trying to build kind of three products in one to go to market in a really strong way. And that’s a hard and long process. So, I would love to hear kind of what’s your approach or what has been your philosophy over the past year as you have gone from napkin to bits flowing, product launching software company.

Josh: First of all, it happened over the course of about a year, but it feels like it’s happened really quickly. Like, just to help put context, I think we started talking with MVL in July like is 2021. I came on board in August. We pitched to Madrona as the primary pitch and then to Decibel as the second, you know, as the other investor, in September. We funded October. That was a very quick process. And then you’re, you’re kind of like, oh, great, now I’ve got this check for $6 million. You’re kind of like the proverbial dog that caught the car. It’s like, now what? I’ve got this big check. I got to go build a product. Holy smoke.

So, I think a couple of things that, at least for our journey were specific to us. We had already, MVL specifically, Madrona Venture Labs, had already talked to a bunch of customers and had a thesis on the market. And so there was a bunch of research that had been done that we could immediately plow into it. I’ve led large customer success and, and customer-facing teams for many, many years now. And so I had a point of view on the market as well. So, in some ways, we could immediately accelerate the process of, okay, what do we want to build? What do we want to design? Because I can say, I want this, not that, I want this, not that. And that’s really dangerous for myself to be like, Hey, I know all the answers and all of that because that becomes very quickly a segment of one. But I want to be clear, like, if you are a founder and have a vision for the product, yes, you need to get external points of view, and also, my council is to build what you think you need to build. I don’t think it’s going to be crazy off the mark. If you’re in the space and you know what you’re doing or you know what you want to build and understand and have empathy for the users of your product, then there’s a lot of real estate that that covers. And so I think that was kind of the first moment, or first few months of the process was like, Hey, we know we’re going to need this thing, that thing, that thing. Let’s go work on those and, and design the process.

Now, once we got into that, then you very quickly go and get a set of design partners. So, we talked to a number of folks in my network and other networks and saying, Hey, I’m going to test the positioning, the concept with you. And so there was a sort of quasi pitch deck that we built, that we tested with them, and a bunch of questions and some really helpful scripts that actually Decibel had given us around ways to ask questions and do focus groups and things like that. There were some really good things about product that we learned during that process and learned by getting input and counsel from some subject matter experts into the design. Then you start putting mocks in front of your customers. And, you know, and mock we’re talking like stuff in Figma that we’ve, we designed and, you know, it’s actual clickable demos, but it’s, there’s no code. It’s just all, all slideware. I think. Here’s one of the things that, you know, sometimes you’ll hear, was that “nos” were much more helpful than “yeses.” People will often say things like, yeah, yeah, yeah. Or, I like that. Because people generally don’t want to offend other people.

Elisa: Said another way, if someone’s saying “yes” maybe they’re just not engaging. Maybe they’re not into it, so, if you get someone saying “no,” it means some part of them cares enough to give you that feedback, and some part of them thinks the problem is annoying enough to talk about.

Josh: That’s right. And so what would happen is sometimes we test product concept. And I’d be like, oh, this is like the best idea ever. And I, I remember this one call with this one woman who is a chief customer officer at a, a fairly large company, and she’s like, I don’t get it. This doesn’t make sense. I’ve already got the solution — it is a solved problem. I’m like, here are the differences. Here’s why this isn’t a solved problem for you. And she’s like, yeah, I’m not compelled. And I remember getting so pissed at the end of that call. I was externalizing my failure to persuade and convince and you know, you ask all the why, how, what questions in that conversation, trying and unpack, well, what would make it compelling? How do you solve this problem today? So, we got a ton of data, but I remember the whole time just sort of seething inside being like, how dare you not believe in my vision? And like…

Elisa: That’s the fire you need because you’re thinking back to, I mean, you’re talking about building these mockups in Figma, which is basically a, a pretty picture for prospective customer to look at and react to, but they can’t really click through get genuine insights. But it gives them enough context for the conversation that they can give you feedback. And so I can imagine that kind of a conversation with a number of prospective customers, kind of put some fire in your belly. Like, how do I translate my belief and conviction, and what my product can solve onto a page and do it in the flow that resonates with my prospective customers?

Josh: That’s right. So after we did, you know, a bunch of these mocks, we did proof of concepts. And this is, you know, this is not very revolutionary, but we did a lot of stuff manually — some in code, some with a little bit of mechanical turk, like human beings sort of doing the work to show that yes, indeed this would work and that we can solve these problems. In one case we worked with a company and they took a year to build basically an equivalent product with us and generate some insights. They had sort of built something homegrown and we have a product that can replace their homegrown solution. And we were able to do something that took almost a year for them to do in the span of a couple of weeks. And that is awesome. Like when you can do that sort of proof of concept and get somebody excited. And in particular, I’d say a great indicator of product-market fit is when you start running into multiple customers that have built something similar as a homegrown solution. Because that means it’s important enough for them to invest dollars. It’s important enough for them to invest time. They think it’s a real problem. And so even if you can’t displace a homegrown solution immediately because maybe you just are not far enough along and your development process or whatever it is, that’s still great validation of product-market fit.

And so anyway, . We did a bunch of proof of concepts, got sort of good validation, good fit. And then ultimately we get to a place where we are today, which is releasing the product to getting to minimum viable product. And what we like to talk about a lot is MDP, which is minimum delightful product.

Elisa: You brought up an interesting point, and I’m curious if this has been a part of your discovery with your prospective customers in selling and figuring out if there is a space in the market and budget in the market for me to build this incredible solution that a homegrown product couldn’t fix, for example. And what we see with a number of our other companies we work with is that decision some companies have between buy versus build and how do you make that decision? Engineering resources are scarce. Time is always a really critical factor. And so what it sounds like with that discussion you had is, Hey, what you guys, took you a year to do with part-time resources, that we could do in a number of months.

Josh: Weeks days, you can’t sell me short.

Elisa: Okay. Weeks to days. Plus, even more importantly, what you plan to do and what’s on your roadmap for the year ahead in creating value and helping them drive effectively top-line revenue to the business is a very strong value proposition.

Josh: And I think the way that you get there is you have to demonstrate sort of competency, and I think you have to create transparency and trust. It’s a two-part process because first, the product has to work or you have to be able to show that you can actually do something. And, and all three of our initial customers that we did, uh, some of our initial proof of concepts with, in every single instance, we had to demonstrate competency and trust and the product work, that our ML insights were significant and superior, all of that sort of stuff. And if we didn’t do that, I’m not sure we’d be in the place we are today with the product. Because that pushes you to excellence, it pushes you to build something that’s very compelling. From our point of view, it’s not enough just to meet a minimum bar. It’s like, no, we have to demonstrate real and significant value. Then I think at the same time, the whole time I’m in conversation with the executives that are sponsoring projects saying, Hey, I’m going to be really transparent about where we are, here’s what we’d like to do. We’d like the opportunity to earn your business. We’d like the opportunity to partner with you. You know, is that a ridiculous question? Like, For you, a big company to work with, a small company, you know, and then they say, no, it’s not ridiculous. That’s great.

And I say, okay, good now what would you need to see from us to feel comfortable to do that? That then also shapes your product roadmap. That shapes your deployment model. It shapes a bunch of different things around what they need to see from you. And I think that that becomes really important when you’re moving from kind of zero to one. When you’re moving from some proof of concepts to actual revenue. What are those components that you need to have? And it could, by the way, it could not just be product features. It could be, well, am I going to get a customer success manager to help me manage or, what are your service level agreements? Like, there are a bunch of things that sit in there that you have to understand, and if you’re open early days in that process and you’re clear and transparent about what your priorities are and what their priorities are, you can usually get to a place where that translates to revenue in a relatively short period of time.

Elisa: Okay, one last topic.

Josh: We still haven’t talked about the product, by the way.

Elisa: Okay. Let’s get into the product. Uh, well this is almost a segue into that, and it’s thinking about how do you build intelligence. You have effectively multiple different products that you’re building right now inside of Magnify. And one of the beautiful and magical things is how do you build artificial intelligence, machine learning, deep learning into the data that you’re collecting and pulling, to create these incredible insights for your customers. So let’s hear more about what those multiple products are, and then I would love to hear how are you thinking about building an intelligent application.

Josh: So, Magnify automates and personalizes the customer life cycle for every single user. So, the way to think of it is it’s taking all of the data in your systems, across all the disparate disconnected systems, taking that in, generating business insights on them, and then automatically engaging every single user to drive product adoption and satisfaction. And if you can do that, you’re going to grow your revenue, you’re going to improve your retention rates, you’re going to get your upsell, and your customers are going to feel better for it, which is, which is arguably the most important thing. And we do that through three components. The first component is actually, that we have to ingest all this data. So, there’s user behavior from product telemetry, and from other systems. There’s sales revenue data, there’s customer success data, there’s marketing data, there’s all this stuff. So, you have to pull all these different disconnected systems into a single repository.

Now, some folks have a CDP, customer data platform or sort of other data warehouses that grab all that information, which is great, and that makes our life a lot easier. Most do not. And so in our experience, we take all that data and we connect and stitch it all together and we create what we call the lifecycle graph. And that is where every single user is in their adoption journey. And how that connects to other signal like revenue or retention or churn or growth. Okay. Then what you do is you do the second piece of our product, which is we then apply user insights on top of that. So, we can then say, ah, we now have looked at every single user, we can understand where they are in their adoption journey and what is the next best step for them. So, user 258 is in this place in their adoption journey, and here’s what they need to do next. And that’s different than user 259 and different than user 260, and so on and so forth. And what that allows you to do is actually create a real level of personalization for every single user. That’s fundamentally important.

What matters there is that we can also then say, how does that correlate or connect with things like churn, revenue growth, trial conversion. So, we can then say, great, here are the risk indicators. Here’s revenue prediction. Here’s how we think about the world both at the account level, and at the user level, and this set of insights, we can literally predict revenue quarters out now for our customers. Then, okay, cool. Great that you’ve got these insights, but frankly, insight without execution is kind of worthless. And so what we ultimately do is we then connect and drive orchestration/ execution. Most companies have lots of different digital touchpoint systems. So maybe they have an email system, maybe they’ve got some sort of end product platform — there’s this laundry list, this Cambrian explosion of tools: Slack, text, chat all of that stuff that sits out there. And each of those are disconnected. And so what we do is we then can connect via API to every single one of those different systems. And then drive the next best touchpoint. So, let’s say user 259, they need to use this feature. The next time they log into the product, that feature pops up. Did they use it? No. Oh, maybe we send them a text reminder. Maybe we send them a Slack reminder. Maybe we open a support ticket because they could only get a third of the way through and they seem stuck somewhere. And we execute and drive each of those actions in those different existing digital touchpoint systems, which allow our customers, the users of Magnify, to ultimately get more value out of their go-to-market investments.

Elisa: And I love this because it’s personalization and almost magic at scale. Those three product pillars that you’re building inside of Magnify to deliver that, that is the hardest technical challenge and the most beautiful and magical solution when you can deliver it. So that’s the vision for Magnify today.

Josh: That’s right. And so you’d asked the question like, Hey, how does AI connect to this? And so here’s what I think some companies do that is the anti-pattern, which is you can’t just sprinkle some sort of user, magic pixie dust on something and it will all of a sudden fix these problems. The way that we’ve thought about AI and ML is that it has to be incorporated from the very beginning. We had applied scientists on our team from day one, and they were influential in understanding what are the product requirements, and how you think about that. For instance, we pull in all this data to create the customer lifecycle graph, but you’re pulling data from all these different systems and each of those data fields has different contexts. There’s different metadata around that that the AI systems need to understand in order to be able to correlate and compare signals across different systems. So if I just said, well, I’ve just created this AI insight engine, but our applied science team isn’t talking to our product and engineering teams around their requirements upstream for data ingest, and the work that we’re going to have to do and to our customer success teams on what is needed for customer onboarding and onboarding wizards and things like that, all of that needs to be incorporated into the product. So, when you think about AI – AI is delivering value because it’s giving these insights, and it can optimize the journeys, the next best steps for every user. There’s all the stuff that it can do, but In order for it to do it well, it has to be woven into the very fabric of your product from day one.

Elisa: We should talk about this in a part two because this is exciting and I know that the journey for Magnify is just beginning. But I have a couple of lightning round questions I do want to end on.

Josh: Bring it on.

Elisa: Okay. So the first one, what do you think will be the greatest source of technological disruption in the next five years?

Josh: Hmm, I’m really torn. I’d say the short answer is the generative AI I think will be a significant source. If you asked me 15 to 20, I think it’s going to be something in biotech.

Elisa: Second question, what is an important lesson you’ve learned becoming a first-time CEO?

Josh: I think how important it is to constantly reinforce the vision and the messages you’ve already communicated over and over and over.

Elisa: It’s so funny hearing you say that because every single CEO I talk with says the same thing. It’s like half my job is just reinforcing the vision, repeating the vision, making sure that what we are doing and our go forward is ingrained with all of our people. Alignment is huge, especially at these early stages when you’re building your first product. Okay. Finally, what’s one thing you’re watching or reading right now?

Josh: Ooh. I just finished binging “Welcome to Wrexham” with Ryan Reynolds and Rob McElhenney, which is a story of them buying a tiny little English football slash soccer team in the middle of nowhere North Wales, and them kind of going this turnaround journey. And neither of them are business types. That’s very clear. And it is gripping television in a fun, lighthearted way. I cannot, it’s like, it’s like a real, it is not just like it is a real-life Ted Lasso. If Ted Lasso were the owner of the football team. It’s a great show.

Elisa: Well, Josh, thank you for spending so much time with us so thoroughly enjoyed the conversation, love working together, and it’s exciting to get more of the Magnified story out into the world.

Josh: Yep. Fantastic. It was great to see you, Elisa, and I hope you have a wonderful next few days.

Coral: Thank you for joining us for this week’s episode of Founded & Funded. If you want to learn more about Magnify visit That’s Thanks again for listening, and tune in in a couple of weeks for another episode of Founded & Funded.

Related Insights

    A-Alpha Bio’s David Younger on machine learning in biotech, building cross-functional teams
    Fivetran CEO George Fraser on Data Integration and Connectors
    Stack Co-Founder Will Rush on Teen Crypto Investing

Related Insights

    A-Alpha Bio’s David Younger on machine learning in biotech, building cross-functional teams
    Fivetran CEO George Fraser on Data Integration and Connectors
    Stack Co-Founder Will Rush on Teen Crypto Investing