The Agentic Edge: Why Storytelling Beats the Noise

Hey everyone, this is David Mario Smith, live again for another episode of The Crunch. Today's a special episode. I have a fellow creative with me, and I always love talking to creatives because we bring a different perspective. I know some of you are tired of hearing about AI and agentic AI all the time, so we're going to mix it up a little today and talk about storytelling. With no further ado, I'd like to bring on Indy Sen. How are you, man?

Dave, I'm doing well. Thank you so much for having me. It's been a little while, so it's always great to reconnect. Summer's in full swing and I just watched the World Cup. I was traveling with the family on the East Coast, and every four years you're reminded of this: people on the street repping their teams, and a real sense of brotherhood and sisterhood. In Boston I saw Norwegian fans teaching grad students their rowing chants on the street. Who would have thought sports would unite the world?

Indy, thanks so much for joining me. Not to go through a career highlight reel, but you built the story for Box, MuleSoft, and Matterport and took all three public, and you've also been at Salesforce, Google, and Canva. When did you figure out that the story is the product? What was that moment of realization?

For me it started in the early days, when I first became a marketer. My first gig was at Salesforce. No matter what type of company you are or what stage you're at, and especially steeped in technology the way we are, it's very tempting to lead with what your product does: the speeds and feeds, the innovation. That's all great. But it's becoming more acute in this age of AI, where the barriers to entry to build are much lower and there are so many more companies. Speeds and feeds are almost expected. The key thing is leading with the outcome you're selling.

That's been true from 20 years ago to today. We're all coasting on promises of what agentic AI workflows can do for you, and everybody is saying the same thing. So what promise can you make to your customer? At Google we had a motto: if you know the magic of your product and you know your user, that intersection is the story, and it's what helps you stand out in the market. So I always start with: I'm assuming we have great technology, but what is the handshake we can make with the customer? Ideally we take their stories and let them help frame how we talk about things, and we lead with the outcome.

In marketing a lot of that starts with basic positioning. People talk about the minimum viable product; I like to think about minimum viable positioning. What promise are you making to your user or your customer, and what's the red thread of technology enablement and commercial enablement that helps them become a better version of themselves? It will change, but if you can dial that in and start with the outcome, the vision of how your product helps your customer, you have the basic strands of good storytelling.

I'm into music, and in songwriting you sometimes realize a song just isn't landing. It's not the beat or the bars or even the lyrics. I just haven't figured out what I'm saying yet. I haven't figured out what the story is. Is that what happened for you, a moment where you said, "We need to figure out what the story is here"?

Absolutely, and it's okay to strum along. Some of the biggest companies in the world do it. Take Salesforce: Marc Benioff will routinely try out a new narrative. Salesforce is famous for spiking the new big thing at every Dreamforce, a new vision that repackages what they're doing. At the core of it is the story. Technology changes so fast and they want to stay abreast of it, so what's the story they want to tell the market? He'll pilot it first with his exec team and his sales leaders, then take it on the road, and what you see at Dreamforce is the finalized version, whether it lands or not. There's a lot of strumming between the first nugget in his mind and the story that gets delivered.

I see a lot of founders do that too, and I'd encourage it. You're not going to get it right the first time, but think about the core elements. Think about the basic plot and you'll figure out the narrative. Don't stray too far from the plot, and don't be swayed by what other people are saying. That's typically where the magic happens.

I get inspired by other artists, even when I'm writing tech research. But there's a fine line between being inspired and being influenced in a way I don't want, where I need to get away from certain things and find my own story. So how do you find the real story?

Sometimes you have a hunch. I've been in product marketing for almost 20 years, so you spend a lot of time with product and engineering to work out your points of parity and points of differentiation. That's the nitty-gritty of product marketing. But I've always found it helpful to go out and talk to customers, even early. Now there's no real excuse not to: there are so many note-taking apps and so many ways to distill what you hear. Keep your ear to the ground. Nothing beats a story of traction and adoption, of knowing how people use your product. And the more complex your product is, the more the customers distill the value propositions for you. Sometimes I'll compare notes: here's what we're saying, but here's what customers are actually doing with it. Shouldn't we change our positioning?

The other piece: meet the customer where they are, but also meet the market where it is. I'm a huge pop culture nerd (music, movies, comic books), and you can often take your cues from where culture is going. I just saw the trailer for Avengers: Doomsday, and I remember when the first Avengers movie came out about 15 years ago and everybody was going, "Oh my God, superheroes coming together." I was at Box then, and my team was in charge of our APIs, making Box better known as a platform rather than just a killer cloud content management product. Our long-term survival and growth would come from being a platform company.

We were jazzed about our Content API, and we realized that if enterprise software was going API-first, we could rally with the other APIs out there. So we reached out to Twilio and Stripe, and before we knew it we had a meme: instead of "Avengers Assemble," it was "Innovators Assemble." Don't build applications on monolithic stacks. Take best-of-breed APIs for content, SMS and communications, payments, and so on. We did a bunch of co-marketing together. That's the multiplicative effect of storytelling backed by a strong meme. So listen to what your customers say about your product (some of the sharpest distillation of what you do well comes from there), and also embrace the megatrends. It's often easier to have a cultural zeitgeist moment in the background to piggyback on and get people excited.

You used the word "hunch." That internal feeling, the instinct, the thing you're drawn to, carries over into the story too. You know when you've hit it. It strikes something in you.

There's art and science to marketing. I've learned to trust my instincts. I'll pitch certain ideas and sometimes they'll get shot down, and sometimes they'll be accepted. As I've grown as a leader, the key thing is to give people that agency and that instinct. Right now I see leaders left and right saying, "Just use these agents to carve out the work so you don't even have to think anymore." I think the onus is on us to think even more sharply, and use AI for the grunt work or as a sparring partner. Data counts for a lot in our field. You can't wish your way into product-market fit; that has to come from real customer traction. But how you tell the story matters too. Can you tell it in a way that isn't dry, that's exciting, at the very least to you and your teams? That enthusiasm is contagious. Every time I've seen a company in unison about how excited they are, you feel it in your bones when you go to their events. There's no shortage of that at Canva Create. And if you can get your user base just as excited as you are, that's a winning combination.

There's the old adage that people will forget what you said but remember how you made them feel. One of my favorite rappers is Nas. He has a song, "I Gave You Power," where he tells the whole story from the perspective of a gun, not about a gun. You feel the entire weight of it without him ever telling you what to think. The choice of perspective does all the work. With a good narrative you change the vantage point, and people feel what you're saying without you having to tell them.

What distinguishes Nas from so many great rappers of that era is that visceral level of storytelling. You're engulfed in it. Your mind registers what he's saying before you fully get it, and every time you hear it again it reinforces what he wants you to feel. So good storytelling is also about repetition. You might have a banger blog post or a sizzle reel, but if you don't follow up and reinforce it, people will forget, especially these days.

And on gut instincts: one I couldn't act on was back in the Box era. We were trying to distinguish our conferences with different speakers, and I kid you not, we were all set to have Nas as a guest speaker at our developer event. Sean, a product manager, and I were so excited because we were fans. Nas was also one of the first artists to invest in the tech scene, about 15 years ago. I think it would have cost us around $40K for an hour and a half of his time, and he was in town. But it got shot down, and I don't think some of the leaders, who shall not be named, knew who he was. There are only so many battles you can fight, but it's still one of my biggest regrets. He probably charges at least three times that now.

As a songwriter and producer I deal in flow. Even when I'm writing research, it has to flow a certain way, with beat and rhythm. You start from where you are innately, and that becomes your instinct and how you move.

As we grow into our roles we learn to trust how we operate. I'm well aware of my own circadian rhythms: sometimes I'm in peak mode, and sometimes I know I need to sleep but can't get there. There are natural rhythms to how you operate as an individual, and as you move from organization to organization, other teams have their own rhythms too. Technology will only make it easier to get more done, and everything pushes us to hustle harder. But it's really important to obey your own rhythm and know how your teams work. At every company I've been at there's crunch time, and then after crunch time there's more crunch time. This whole 996 phenomenon: it can't always be crunch time. Even our devices have to plug in every now and then. So one of my tenets is to know your rhythm, and know when to plug in and when to plug out.

I'm a runner. I don't think I'm as fast as I used to be, but I'm probably smarter than I've ever been about listening to my rhythm and sticking the landing. I've run more marathons in my 40s than I did in my 30s. I still enjoy it, but it's harder every time, and it makes no sense to expect more hustle and grind without recharging. I see so many people burn out so easily now without recognizing it, because they think technology hasn't caught up to their ambition. Don't forget that ambition is built on patience, on being deliberate, and on focus. If you don't give yourself rest, all of it is for naught.

So the marathons taught you rhythm and pace. When you're doing a launch, do you get nervous, or do you use what you learned from marathoning to center yourself back into a rhythm?

Whenever I've orchestrated launches, or helped my teams or the folks I advise orchestrate them, everybody looks for the big-bang moment. Was it well received? Did we break Techmeme? Did we get the media coverage we wanted? All of that is important, but so is what goes into the work before the launch and what happens after it. One of the greatest leaders I've worked with, period, was Carol Carpenter at Google. When she was coaching us as product marketing leaders, she said: we all want that big pop launch moment, but don't forget to stick the landing. So for me it's always launch and land. And if you follow the rocketry imagery, a lot of work goes in before the launch.

I prepare the way I prepare for races. There's the work before, how you show up on the day, and the recovery. Some days you hit your goal and feel great, and recovery still matters. Other days you don't run the race you wanted. I trained hard for the New York Marathon three years ago, thought I was ready, and then cramped up. There was nothing I could do about it, and I had to be kind to myself in recovery.

So with launches there's a pre-launch checklist, a launch checklist, and a post-launch checklist. A lot of people focus on the pre-launch, and as soon as the launch is done they want to sign off for two weeks. That's tempting, because your teams are exhausted. But the landing moment really matters: the day of, two weeks later, 90 days later. Do you have the discipline to look back at the metrics and report out? Otherwise it's a flash in the pan. Being on Techmeme feels great, but what metric are you going to show the business, whether that's sign-ups or something else? Have the discipline to give a 24-hour report, a 90-day report, whatever the intervals are. Repetition, even internally, is your friend.

That's so similar to rolling out a record: the teaser, the build, what you hold back, who hears it first, and then whether you have the follow-through to keep it alive. You don't just have the baby and leave it at the hospital. You have to raise it.

Some of the companies we all admire from afar are absolute launch machines, but what does that launch velocity cost internally? I think Anthropic did something like a launch a day for 52 days. In 52 days they shipped more than the average company does in 52 weeks. It's great, but it's also dizzying, internally and externally. Very few companies can get away with it. They're a business, but they're also a research organization, so it makes sense for them to ship quickly. They're pumping out innovations and insights for customers and for the academic side. A lot of people say, "We've got to launch every day like Anthropic." I don't think you do. They're the exception to the rule. What you do need is to stick the landing when you launch something, and to be super clear about what the launch is supposed to accomplish and who it's for.

Another nugget from Carol Carpenter (Carol, if you're watching): you're doing a good job in marketing if you're reducing the cost of customer comprehension. Your product may launch at high velocity, but customers and prospects can only take in so much at one time. Even if you have tons of innovations, package them in a way people can grok. That could be standard release marketing, reinforcing things, release notes. But especially in enterprise software, don't expect people to act on every launch. People like innovation, but they don't like change, and change has to be rolled out with a lot of consideration. Doing things fast doesn't necessarily earn you trust.

That goes back to storytelling: be deliberate, but be incremental. It's fine to keep a steady clip of talking points, but to me basic storytelling is scaffolding. You can't pull people in different directions. Internally your teams may say, "Here's where we want to be in a year," but you have to hold your customers' and prospects' hands as you tell that story in the market. That's a core part of responsible storytelling. The number one person you don't want to confuse is the customer. The second is the analyst.

Don't worry about us. [laughter]

As a tried-and-true product marketer, I think those stakeholders sometimes get disregarded. People say, "We just want to make the headlines." You'll make the headlines, but maybe not the way you intended if you're not disciplined in your storytelling.

You're making something people are supposed to feel, not just receive, so you have to keep in mind who you're writing for. When you're building the narrative, who's the person in your head?

It goes back to your ICP, a core tenet of product marketing. You have to have a strong view of your ideal customer profile, and then be really specific about the personas. The products I've worked on usually have a business decision maker, an influencer, and a practitioner. With developer products, and with productivity and collaboration tools, you want the actual user interested, but don't forget whoever influences the purchase and the budget holder who ultimately decides. Whenever my teams work on briefs, we think through exactly who we're going after, because the default setting for so many companies is speeds and feeds: how fast, how efficient, how performant. That's all great, but who are you trying to convince?

Then I work in reverse, like that Nas song. Picture the outcome, then work back to the messaging elements, the reasons to believe, the value props, and down to the core product positioning, so that by launch you have the product and messaging primitives for each channel where you'll reach the people you want to convince. You have to picture that end state before you put pen to paper.

You're not going to one-shot it. One piece of messaging won't work for everyone. At WeWork we already had a strong brand and product perception, but there was a big difference between enterprise companies using WeWork and solopreneurs at a hot desk, and then between markets: the vocabulary, the appetite, the mores. My responsibility was to create the core positioning, knowing it would get remixed and that the value props and reasons to believe would differ from market to market. Our tenet was "core creates, field executes." Whatever stage your company is at, you may feel in your bones what your product does well, but you'll have to land it with the decision maker, the buyer, and the user in at least three different ways, often across different channels. So many people interested in AI follow OpenAI and Anthropic and are ready to consume that information, but that's a position of privilege. Most companies don't have that captive ear. So how do you make sure your message lands on the right ears? You have to think about who you want to listen.

My dad is a jazz musician. I grew up going to his rehearsals in New York City, in places a seven- or eight-year-old probably shouldn't have been. He always told me, when you're writing or performing a song, even alone in the vocal booth, picture someone in front of you. Never sing or rap into the ether. Visualize a real person and tell that person the story. When you picture that persona, what you're creating feels more real to you, because now you're conveying a message to a real person.

I had the huge honor of being on stage at Canva Create, tapped to deliver a message on behalf of my team, in a theater with 3,500 people staring at you and who knows how many streaming live. One thing I remember from public speaking classes is exactly your father's point: look people in the eye. Even if it's dark, look at one section of the audience while you deliver one part of your message, then look at another. There's a teleprompter, so it's tempting to speak into the ether. But my confidence came from telling myself: I don't know who's in the audience, but I'm looking some of these folks right in the eye and telling them this story. Otherwise it feels so impersonal.

One concert I wish I'd gone to was Usher's residency in Vegas. In the clips he skates up and sings a whole song to one person. Even if you're not that person, you remember the visceral connection he makes with the audience. It's still an art, and it's emotional. Is it instinctive? I'm sure, but it's probably planned too: "Go up to section C today." There's forethought, and then there's how you show up. At keynotes you can always tell which executives are focused on the notes. The delivery is flat, even if the content is interesting. Connection doesn't come naturally, but you can practice it. Look people in the eye, know who you're speaking to, and speak to them. The enthusiasm and authenticity will be felt by whoever is watching. Authenticity is a key part of storytelling.

AI has made polish free. You can generate a slick launch that looks great, but the authenticity and the personal storytelling are something else. Does AI make what you do harder, or more valuable?

Great question. Six months ago I'd have answered differently. There's no doubt AI is making some things easier, particularly where the outcome is deterministic. I've built websites before, but I definitely vibe coded my own consulting website. I knew what I wanted and I'm super happy with the result. It would be silly not to use AI for that kind of thing. But much of my work as a marketer involves probabilistic outcomes. Preparation takes you only so far, and there's always an element of chance in whether your messaging lands. That's where the human in the loop matters more than ever.

Six months ago I'd have said AI can one-shot the deterministic stuff and get me 90% of the way on the probabilistic stuff, like an essay or an artifact I'm developing, and that it's great not to face a blank page. Now I think that for probabilistic work AI is a really good sparring partner, but it won't get you the last mile. And even at the first mile, it's more important than ever to give it the right directives so you get something worth reacting to. Let AI handle the prototyping and throw the darts at the board, but you have to judge whether you're on target.

I just read a David Brooks piece in The Atlantic on how AI should be used. I'm paraphrasing, but his point was that people bring different levels of cognition to the table. For the deterministic stuff, nobody should worry about having AI do it for you. But if your work depends on your judgment, taste, and acumen, it's a double-edged sword. Rely on it too much and you start creating slop, and the slop makes your thinking sloppy. If you stay sharp on what you want, AI is a great sparring partner. So my answer is both. I can't imagine my job without AI now, but it's one tool in my arsenal for doing work at a higher level of quality, and that quality comes from people. Everyone has their own bar and their own way of saying things. I use it as a proofreader and an editor, but we're all works in motion, and how I want to say something changes from week to week. It's not a silver bullet. Wield it, but don't use it for everything, only where it adds value to your contributions.

The gem there is judgment and taste. How do you account for taste? I can vibe code a prototype with all the brand colors and palettes, and it's still ugly. My brother is a UX and graphic designer, and we talk about this all the time. Judgment and taste are the human in the loop, the emotional part you can't replace.

Agreed. AI will get better, and so will you. People like to say the AI we're using now is the worst it will ever be. The flip side is that with these tools we can do things we used to aspire to but didn't have the mental or financial capacity for. Years ago I'd have paid a designer for my website. Designers listening may think it's crap [laughter], but I hope they feel empowered by these tools, not alienated. That's the promise of any good technology: it empowers you from behind the scenes. There's the Arthur C. Clarke line that any sufficiently advanced technology is indistinguishable from magic. Great technology happens in the background. We're the ones inventing and building all of this, so we should never take a back seat to the technology. That goes back to storytelling. You may have a killer product, but don't put it in the foreground. Put the outcomes, the story, and why you're different in the foreground, and let the technology be the proof point.

I cover this space: agentic AI, CX, and the convergence of CX and EX. Salesforce, Adobe, NICE, Genesys, Microsoft, Zoom, Google. Everyone is telling the identical story: agentic, autonomous, context-aware, intelligent orchestration. You could swap the logos and not notice. What are you telling founders and startups right now?

If you're a startup without "AI-powered" somewhere in your deck, investors will thank you for your time and move on, because they want a thesis that uses it. Not everything will be AI-powered. Capital-intensive industries like oil and gas still depend on mechanical engineering and IP innovation. But what I tell founders is to think about the point of view you're bringing to the market. Have conviction about how you're addressing a problem, why you're the best team to solve it, or how you're thinking about it in a novel way. That's what investors and customers want to hear.

Two things. First, go back to minimum viable positioning: here's where our product is right now, and our approach is different because of X, Y, and Z. Of course those will include AI, APIs, whatever you have. Second, show early traction in what people are building and how they're using your product. Even a few early design partners set you apart from someone leading with technology they're still incubating. Design partners and early traction stories, even if you end up pivoting, build confidence in whoever is looking at your pitch, whether that's a prospect, a customer, or an investor.

Stepping back, the hyperscalers are saying many of the same things because they're building the same things. Meta invested like crazy in data centers and now seems to be positioning some of that as capacity it can sell, stepping into AWS's business overnight. They'll probably succeed in some ways, but it feels like putting the cart before the horse. Apple, on the other hand, has done something interesting in this AI era. I still tune into all these developer conferences. Ever since Steve Jobs launched the iPod (that's my origin story as a marketer: I didn't know who put that story together, but I wanted to do it), I've been a nerd for this stuff.

At WWDC, the most interesting thing to me was a diagram that literally felt like a PowerPoint slide: a circle showing their devices, their services, and the areas they're investing in. That's the core of what a great marketer is supposed to do. It showed how they, as a vendor, understand the world as fast as it's changing. Their thesis was: we don't know where all this is going, but here's the part we play and here's what we're interested in. What interested me most was the negative space around that diagram, because that's where the islands of innovation are: the next great app, the next Uber, the next Airbnb. Apple is laying the groundwork and showing developers the new picks and shovels, and that gives them license to think about the negative space. AI won't do that for you. The thousands, millions, even billions of entrepreneurs out there will say, "Now I know what I have to work with, and I want to build something that does this."

So my advice to founders, and even to the hyperscalers, is not to forget that an opinion and conviction about the market is the guidance people are looking for. Nobody has figured out where all this is going. You instill the most confidence by saying: based on where we believe this is going, here's what we can offer, and if you come on board we'll make sure you stay on the cutting edge. I saw that in the early days of Salesforce, at Box, and later at Canva. With the companies I advise and talk to, I know immediately when a founder is convinced of what they're doing rather than running things by the numbers. Hold strong opinions, loosely held. This could change overnight: Anthropic could release a new model or module that Sherlocks your product. But for now, here's where we are. Leading with that confidence and giving your customer, investor, or prospect the lay of the land as you see it is where you'll shine.

Have you been in rooms where the story is better than the product?

Yes. You need the vision, the conviction, and the ability to execute, including the bona fides of your product. That's true for hyperscalers too. Once upon a time WeWork was a hyperscaler in its space, and we were very ambitious. But we were also putting things out there that had nothing to do with our core business. People would ask, "So tell us about this wave pool thing you're doing." I was in marketing and had no idea what that was. It was in my FAQs for press pre-briefings: "Maybe you should talk to so-and-so about that." Focus matters. Don't give in to FOMO, or to the fear of being misunderstood or of what people expect from you. Come up with something you're truly convinced of, that you have the evidence and credibility to prove, and go for it. Your product may pivot and your messaging will pivot, and that's one of the affordances of this age: it's easier than ever to change things.

It's the conviction of the story, knowing who you're telling it to, and letting them feel it as they receive it, across personas, users, analysts, and partners. Now it's all about the ecosystem, and how the message is received across it.

Maybe the last thing I'll share is one of my beliefs: a good platform is a good story, and a good story is also a good platform. That matters more than ever because almost every company is about to be in the ecosystem business. Every SaaS company under the sun now has to think about what its agentic interfaces look like. They've honed their go-to-market on selling per-user, per-month logins, and now they'll be selling packets of API calls and tool calls to agents. Almost overnight every company is becoming a platform company, whether it knows it or not. A couple of months ago people were saying SaaS is dead. I think the companies that don't make that platform pivot, or at least open the aperture to include a platform component in their go-to-market, are the ones that will die. For the ones that embrace it, it will be a lever of growth. That's why these are such exciting times. A good story is a great precursor to platform success, and vice versa.

We usually cap the show at 30 minutes, and I lost track of time because I was so enthralled. That's a testament to you, my friend. You've been so insightful and engaging, and this will be on demand too. Indy, thanks for coming on. We've got to do it again.

Like Captain America says, we can do this all day. Dave, thank you so much for having me. I want to be respectful of everyone's time, but I hope this resonated, and I'd love to do it again. Always a pleasure, my friend.