LaunchPod - Brandon Harris
===

[00:00:00] 

Jeff: brandon, what's up, man? Good to have you here

Brandon: Hey, thank you for having me. Thank you for dragging me out of bed and early, early in the morning.

Jeff: Yeah, we, we are record-- You know, it's 9:00 in the morning. It's not, not that terrible. I was worried for a second that you were based in LA when we first jumped on, and I was like: "Wow, you are, you are a rock star here, man." But, but you're East Coast with me, so, uh, you only get so much sympathy for being early.

Brandon: Yeah. I have five kids, so the mornings, uh, uh, can be, can be rough at times

Jeff: 100%. I, I, I only got two, but, uh, one of them is a very early riser, so, uh, that's always a fun one. We wanna talk about an interesting topic today around kind of the, the way the world is changing with traffic, , specifically, you know, internet traffic and just the rise of agents and bots and all these things.

And, it's something that's on everyone's mind. And luckily, think you have a background that only a few people have this kind of insight where you, you were at a company called Wiser, which was a big consumer of content. You're at Pantheon now, which, , plays the other side.[00:01:00] 

Um, so let's just jump into it. But before we do that, maybe just can you give us the TLDR, how'd you get here? What was the path to being VP of Product at Pantheon and, what's your background, ?

man?. 

Brandon: , I've been really fortunate to do quite a few different things in life. And so, like, it all comes back from the very beginning. I went to school to be in film, ? And what I learned is that I loved having an idea and then working with a group of people to see it come to fruition.

And so that works in communication space. But, uh, what I found in transition to digital over the course of my career working in, you know, in finance, fintech, pharmaceutical, management consulting, is that the same thing applies, that companies and teams are generally groups of people with an idea of what will be valuable, and you're working with them to make it real.

so that passion of mine translates over. , And in career, I, I just got to Wiser. There's a really interesting question around what does the web look like in seven years, 10 years? And when I thought about that, you know, when I talked with the team about some things that could happen to solve that and, and [00:02:00] change it, it just seemed like a really interesting challenge to me, Um, and I thought there were some things you could do in a very short period of time. I'm really excited about go live and some of the things that are gonna be released in the, in the new vision that we're setting in. But, , that's what brought me to Pantheon.

Jeff: Nice. Yeah, I mean, , I love the kind of background of both product and product design and kind of marrying that together that you, , bring to all this stuff. Let's just jump right into the topic because I do think it's a salient one right now. So I, you know, aside from hosting this and, and working with product folks a lot, I also - run our marketing team, and it is a conversation we've had a lot, but also every single person in product, marketing, engineering, almost any stack within the business has talked about, a huge portion of our traffic is agents, or how much of our traffic is agents, or, or the rise of agents, you know, at the far end of this conversation,

The, the concept of the dead web, uh, has started to come back up of, you know, it's just gonna be agents talking to agents. [00:03:00] Um, like I said, I mean, between Wiser and Pantheon, I think you have a great insight into this. , How should we be thinking about this? What's, what's a good way to kind of frame this from the beginning? . 

Brandon: The web has gone through several iterations since first invented to share research, right? So it, it is just going to the next iteration. If we think about the cognate or the algorithm with - , transportation and traffic, right?

Like, when we really think about that, we were walking, and then we found wheels, and carts, and animals, and then bicycles. We could do some self-powered things, and then trains, then cars. But in each, each time that happens, there's this sense that the new technology for transportation is gonna overtake all of the previous technology, and we'll never use it.

But the reality is, I still have to teach my kids how to ride a bike, right?

And so, like, I use this example because a new mode of, , traffic on the web and a new primary use case doesn't mean that all the other ones become invalid, but it does mean that we have to change how we think about what [00:04:00] transportation looks like. There are now new options, if we think about the information as what was being transported, there are now new options to get that information as an endpoint to where I don't have access all along the way to click and browse through your site.

Now it's, I want a specific piece of information. I'm sending an AI to get that specific information. And in some ways, it's kind of like what a CMS does for us, right? if we have a site, all of our information lives structured inside of a CMS, and we are asking for that CMS to grab the information, to present it in a structure that we want for our website.

With AI, now we're saying, "Hey, get this information for us," or, "We wanna understand or learn about this," and it's going out to a bunch of websites in aggregate, which in aggregate become kind of this CMS stack, and it's grabbing all the information and pulling it forward. ,, we have companies that did this before.

So at Wiser, we worked with, , consumer data for pricing, and we scraped and went out to millions of pages a day to understand what the pricing on certain , products were. And so [00:05:00] we used all of those things in aggregate as kind of a CMS for our product that we delivered that information to our customer.

I say all that to say, I think w-what we're gonna see is a compression in the volume of human traffic, and we're gonna see a compression in the volume of traffic that is intended just for browsing, but it's never gonna go away. And in some ways, it makes it more valuable because you know, like we took Intent.

We know that when someone is coming to our site, the intent might be a little bit stronger to actually, , interact with us, to learn about us, to know about us, to transact with us. And so for marketers that are aware instead of being scared, like it's, it can actually be a really great opportunity for us to really zero in and know that we're gonna get less traffic that is not intentional about coming to us

Jeff: Now, especially on my end, like, you know, running marketing here at LogRocket, one thing I kinda worry about here is, I pay with a [00:06:00] traffic basis on a lot of tools we use. our CMS, our, our host, all that kind of stuff. And that's great, and that worked, but like, the agent is not as valuable to me.

Or at least , I think we have trouble explaining up the chain , to, you know, the CFO and the finance folks and, and the people, you know, we need to be answerable to, how the economics work out, is someone gonna fix that or, or, you know, am I looking at this wrong here?

Brandon: No, no, I think you're looking at it right. what we're seeing is that traffic was a proximate metric for the value that we were getting out of the web, right? So how many visits I get, it's a easy way or a way for people to be charged for it. Right now it's a really important metric for pricing, not an important metric for value, , all the time.

And so, like, we'll see that change to where, like, I had this dream where in terms of, of volume, that you can change your volume based upon who's coming to the site and where you are in the month, what is your goal and what you're trying to get out of it, ? And so, like, it [00:07:00] started with this idea of, five years ago everything was how do we block bots, how do we stop them from coming, how do we keep them, , from our site?

And it'll shift to, well, we want some, but not all. We want this traffic, but not all. And so in order for a provider, infrastructure company like Pantheon for example, . if we're gonna block 50,000 visits a day, then that's gonna substantially or could substantially affect our revenue in doing it like that, so there have to be other values that hosting companies provide relative to just getting volume to the site,

right?

Jeff: g- and I get both sides of it, 'cause like I look at it that the agent visit or the scraper, any of that , is far less valuable to me than having someone actually on my site. but it costs the same for, whoever's hosting it, whoever's pushing the data, whoever's loading that stuff to send it.

Like they don't care who they're sending it to. It's the same infrastructure and, and bandwidth costs and stuff like that. Do [00:08:00] you see anyone on the market pricing this right at this point? Or, or, it just being smart about filtering who we actually serve and maybe we can't, you know, if we care that much about cost, we don't serve every bot and every, every little, you know, thing.

We identify traffic better or something.

Brandon: I think it's similar to what we saw in-- I had the opportunity to, serve as interim C-CMO for my company, and so looking at it from that side, when we worked with ads, right, like we started off with the idea of how many people are viewing it in impressions, and over time, the industry has moved to people taking action, right?

Like, I want my attribution because I wanna be based upon the actual clicks. I think a similar thing's gonna happen to traffic. So to, to your question, right now the answer is that we get better at controlling how our traffic is being used, how that meter is being run. But I think where we end up moving towards is that, cost on your side in terms of like, A, we're serving a site, is going to be based upon the value of the traffic that you're getting.

[00:09:00] It's gonna move closer towards the, the action, , part. And that's both properly accounting for traffic that comes from bots, 'cause it can be very valuable, right? Like AI can do some amazing things, uh, valuable to a company that it's interacting with. , But how do we put that in context with the other traffic as well?

And so I think that the way that we charge for it, the visit will become kind of like impressions. They're valuable, but not the same value as a conversion. And we'll start to get better at tracking both and being aware so that the people who are providing the sites also and are working on the infrastructure aren't just paying for the site existing, right?

, And that's kind of where we've gotten to because of what was available in the technology

Jeff: as I think about like th- and this is something we struggle a lot with or, or think a lot, I think a lot about here is, you know, at one point we had a very popular blog, and when we had visits on that and people landed on it, you know, it was branded, , people saw, , messaging about us, and we, we added value to their day-to-day workflow and to their day-to-day, life.

[00:10:00] But in return, we got, you know, brand equity and trust and stuff like that. And sure- certainly, you know, if we're answering questions, uh, via an agent and it's citing us, that's great. But to us, the cost is the same, right? The agent came, loaded the page, download stuff, da, da, da, da. It cites us. My wager is, , very few people are actually clicking on the citation links once the AI gives them the answer.

they're getting the info, you know, so they're getting the benefit, they're not seeing the stuff that wrapped the answer, which, I understand as a user and consumer, It's very nice. I, I'm a big fan. I think it's the right way to go. But I do think this is one of those things that's gonna be an interesting thing around content creation, what's the incentive and, and all that when, when we've kind of removed some of that.

And yeah, I know they're doing ads in, in the platforms and stuff now. Yeah. It's gonna be an, it's gonna be an interesting one. ,

How do you look at like grading, I guess? Or, or, you know, what do you, what do you think that turns into? 'Cause I, I, I have an answer here about as a marketer, , but my thoughts turn pretty dark and, , black hat if I'm not careful really, really quickly.

Brandon: So that's actually where I was gonna go with grading, 'cause I [00:11:00] think where we go is answers on both sides with grading so that , a company or a website can determine what value a particular type of interaction or traffic is for their site, right?

Like, whether that's happening, on their side of like the CDN and saying, "Hey, allow this much traffic," or setting a ratio, or whether it's, , just grading for advertisers or, or, shipping back the cost. I've had people reach out to me about setting up networks. - How might we allow our customers to charge for access, and what might that look like?

Some of the large publishers are doing that, but what might that look like for websites as a whole? And I've had some folks with startups reach out to me about that. But I think it also ends up applying on the other side, where the grading of traffic can be done on both sides Ultimately, the market will even itself. No one's going to want to pay the same to get worse quality, and so the market's gonna have to figure out how to solve that. Right now, there's the commodification of hosting as a result, but I think that what you're saying, the [00:12:00] grading the traffic and, and understanding the different values of the pieces makes it possible for everyone to win, 

Jeff: yeah, that, that, that's where I, I kinda get curious 'cause like, like I said, I wouldn't necessarily do this 'cause I, you know, ethics and, and morals and stuff like that. But I do start to think about, the kind of handoff there of we're gonna grade traffic and we're only gonna serve certain ones, which, which the inverse part of that is we're not going to serve certain sources or certain things, , maybe sometimes, maybe all the time.

, Or like you said, maybe it turns into a charge model. That works if both sides are being completely truthful with each other. But it gives you, you know, motivation on, on at least one of those sides to be, let's say, less than truthful. I could see a whole class of businesses who are not even black hat businesses, but just don't wanna spend a ton of money for the content access to start to, you know, it becomes an arm race of like, can we misidentify who we are?

How do we identify? How do we, you know, beat that? And, and just back and forth, back and forth, and, and it just becomes a new tool, , you have to have as a publisher in your stack maybe to, [00:13:00] to better identify and stay ahead of that kind of stuff. , Do we see that yet? I mean, that seems like something with like bots and scrapers is kind of already, it's not a new thing.

It's just a different, different flavor of an existing battle almost.

Brandon: I think W-we see that now. We don't see tools that are being, I think, built well for everyday use for that, that can make the real time changes. So within the CDNs, when we set them up, , it's one of the beautiful things about what we're doing with our AGCDN and some of the offerings and how we're evolving them at Pantheon, is towards being able to make some of those, , filtering and changes.

But I think the next wave of what you're talking about is like, how do we make that a product that I can have more real time control over and determine what's valuable for me? , Eventually getting it to where it follows campaign times, whether it's Black Friday, whether it's, you know, Amazon during the summer, whether we're trying to get a particular push.

And I think that we're still probably a year away from having some really mature options for that.

And I don't think it's [00:14:00] gonna come from the infrastructure side, to be quite honest. I think it's gonna come from, from independent companies, either on the marketing side or new startups that are, that are building something that tackles that well

Jeff: I wanna go back to one thing real quick , 'cause you, you said it and we moved on and I think it's, it's a really interesting thought and deserves a little bit more kind of expanding maybe, but the web as the CMS.

cause, you know, up until very recently, - it was easy, right? You had a database, it had page structures, it had HTML in it. It was very deterministic. you, you know, at Wiser, you, you brought this up, you guys were scraping millions of pages to pull in pricing data and, and show that to your customers, and that was part of the value you brought was it was a price aggregator.

I'm probably dumbing it down a little bit, , but one thing I found, , and I think this kind of parallels it, one thing I found really interesting about kind of, AI agents and, what this behavior brings is sure, it's great that you can have it go off and do things and blah, da, da, da, da, and it will, pull, flight times for you, whatever you want your little agent to do.

The other thing that's been interesting I've found is being able to kind of build, like I have this product, this product, and this SaaS product But because of how they all kind of hydrate data internally, [00:15:00] now I can use this piece of this product, this piece of this product, and this piece of product three and turn it into its own product almost that is only, you know, N of one.

It's just mine. and, and that's an interesting point, I think, this concept of the web as the CMS is where does that go? Like, could we start building websites that are dynamically pulling, , information I know you're probably going to want in from other sources and showing that? Or, you know, how far does this go of, of the web as the CMS versus my little database as the CMS?

Brandon: , I think it absolutely goes there. , 15 or so years, I'm not gonna date myself, 15 or so years ago, I was at Gannett, which is, I think now Tegna, um, used to be a large media conglomerate. And one of the things that we were working on then, there was a project that built the websites as someone called on them based upon the content that was available.

Tegna, Gannett has hundreds of news sites. and what you're describing is, it's sounds like an evolution of that, where it's saying, hey, if the web is the CMS, if all this [00:16:00] data is available very quickly in real time, I can assemble it into whatever I need it to be to either build a new product or a new page on demand.

Whether that demand is happening from the user in r- real time, or whether the demand is happening from me, to your point of like I'm cobbling together a product based upon information I can pull from everyone else. like that is absolutely going to happen, and I think it's going to go two ways. It's similar to what we see in social media, right? Like there's a lot of AI slop, , but there's some really good things that happen, and some really interesting content that's also produced. I think that you'll see some things put together that are very interesting and useful, and you'll also see a good bit of slop.

And so it gets to the, much like a CMS, the value isn't gonna be just having your information out there and available on your site. It's gonna be about how do I imbue trust into how it's written, into what's there, into curating, thinking both about someone's gonna come and look for the information, or something [00:17:00] is gonna come and look for the information.

How does it understand what is valuable and what is real, and how much should be involved in the context that it's gonna pull? There's gonna be limitations on how much information it may pull at a time, how many articles would be linked. And so other things in how I structure my site and structure the content, and write the content and when, that would imbue that so that when my stuff is pulled and caught up, I am recognized for it, right?

And it imbues, like it's gonna be, , more seen, , by a bot. so like I think that if I could, structure it a little bit more, there are multiple thoughts that the web becoming the new CMS requires. Requires us to think about- How trustworthy our content is, and how do we communicate that trust?

How do we insert ourselves into the content with fingerprints in the same way that we're doing now, other than looking for AI's fingerprints? How do you insert your brand's fingerprints into your content? You know, broadcast background, so, like, there are advertisements and there are [00:18:00] in-line ads. There are product placements.

How do you provide brand placement within the content that you create so that when it is pulled, it is still representative of your brand or harkens back to your brand? It also requires us to think about what are ways in which, to your notion of black hat, we can sequence the content that we're producing, whether that's by volume or by order, so that that communicates something that we want and draws in the bot or the AI that we're looking for.

And w-we've done this as an industry, which is fantastic in a way. Google forced everyone to do this, right? Like, we had that huge adjustment to how Google was looking. I think it just takes another level up, the complexity and the need to understand it and operate on it well. The beautiful thing, in some ways, that's happening with the rise of AI is it's actually giving more importance to the marketer.

It's giving more importance to the tools for marketing and the impact, right? Like, that, that's happening, and I can say that from the infrastructure side and from the side that's doing the bot [00:19:00] side, , and having been on the market. Like, it becomes so much more important, and I think marketers and t-companies and teams that understand the multiple layers of this are going to do extremely well, because there are gonna be lots of companies that struggle to adapt.

There are gonna be lots of companies who fall into the trap that we saw when we went digital of just taking my paper product and putting it on a website. There are gonna be lots of companies that think that what I did in 2000 for my website and for my content is gonna work in 2030 and 2026, and it's not.

And so for the people who understand and can adapt well, they're going to be far more impactful. They're gonna see returns, 

Jeff: Yeah. 

Brandon: encouraging

Jeff: think it's interesting you bring up kinda trust and how do people move forward and, and, you know, the concept of you could have a dynamic site that builds on demand, pulling kind of whatever it needs from across the web

both an interesting and a terrifying one. Um, 'cause basically, I mean, [00:20:00] uh, you are, you're at some level, you know, playing information DJ, , where you're just kind of like remixing, , sources but there's a million iterations where that could go badly and wrong that is not nearly as good, and I think that's what we run into, you know, here is, is yes, you have the opportunity to pull whatever you want in dynamically. If you can build it fast enough, great.

, But you also have, you know, and, and we as consumers have the opportunity to have a lot more crap foisted upon us.

, But I think that, that goes into something really important is Trust and how does it play into kind of creating content quickly and, and customized for everyone? 'Cause that can be really good. If you can use that tactic to serve me better information and really, really help me with my needs, because in the end, we don't make software 'cause it's fun.

We make software 'cause for a long time that has been the best way to solve a problem. At the same time, if you can use this tactic to answer my questions better or to, to make me better off, I don't care how you do it. Use AI all you want if it's trustworthy and good and seems high quality.

, But you guys over at Pantheon actually [00:21:00] published research on this whole kind of concept of, trust and how do people internally kind of like ingest and and almost like mentally fingerprint a site and, and determine like, do I trust it?

Do I not? , I mean, I'll let you describe it because I, I only know surface level, but I thought it was really cool point and kind of runs off of this of, of what does this turn into from a trust standpoint? Like how do we kinda look at stuff and know what to trust?

Brandon: Humans are phenomenal at recognizing patterns. It's, it's our way of survival. And so our ability to start to recognize patterns in content that we're seeing in things on the web is how we identify slop really quickly. Like, you see the M dashes now, and you're like, "Wait, is that AI, or is that, you know, the two statements, it's not this, it's this?"

we've started to recognize patterns in how it writes. , And to therein, the, LLM companies are starting to insert patterns so that they're fingerprints of AI that they can look back at and see, how it's being used. And so I think that [00:22:00] what we have to realize in terms of trust, ways that we can impact that, is that we could be more intentional about actually writing final copies of things.

I think we could be more intentional about how we're structuring things. I'll give you an example, like at the very core framework, is that if your site takes a long time to load, it is gonna affect how much people trust your site. in the same way that if they come to your site and, you know, there's a lots and lots of movement, and it looks like GeoCities, like, they're gonna have concerns, if you're not the government, which everyone expects to be terrible for their sites unfortunately at this time, or paying bills and such, like, the expectation is I'm going to get a consumer-level experience, and if you're not providing a consumer-level experience on your website, there is going to be question about how much I can trust you.

If patterns take too long, or if they deviate, , from a common pattern, and there's not a good reason why, people will wonder. , [00:23:00] Checkout is a great example. So I've had the opportunity to, you know, in my time at Meta, work on pulling together the checkouts from multiple different organizations into kinda, , a Meta Fintech into what it looks like.

And one of the things that we saw when looking out, was that when you have a deviation from an existing pattern with checkout, people pause, and that pause will cost you money, right? And so, like, one thing that companies can think about in terms of building this trust is what are the existing patterns that I need to follow in terms of my content quality, the content type, , things that speak to my brand, and, , UI patterns that I need to follow in order to maintain that trust thread, and what are places where I can deviate because it is going to set me apart as having a very different experience?

If I come and answer the door and I'm holding a knife, right, like, you're gonna be concerned, right?

Jeff: Yeah. Yes, I am 

Brandon: because no one expects you to come to the door answering a knife. And so it's, it's a similar thing. However, if you walk into my house, I'm in my kitchen [00:24:00] holding a knife, "Oh, that's cool. We're about 

to eat."

Right? Like, I figure you're a chef the same action done at different times can have completely different meanings, and I think that same thing applies with the patterns that we use on our websites and our content, where us being mindful of, the experience itself and what the experience is communicating it's a huge importance because humans notice.

We are still, at this point, better at noticing those type of repeating patterns, , than AI is. And so, like we're gonna notice when those things are off. People will know when everything you write comes from AI. , And that's not just in visual content and written content. They will know if you are just following the patterns that everyone else is using.

A WordPress site looks like a WordPress site. People who don't understand why will realize why a lot of times, and that's not always the case, to be very clear, but 80% look the same. And so like if you're not intentional about where to keep to the pattern and where to deviate from the [00:25:00] pattern, you can get caught up in whatever that pattern represents.

Jeff: Cause this is something Pantheon actually put out research on, right? Where, , , there were markers that people looked at a- and discerned if it was slow, if it was glitchy. It was assumed to be AI, which can be problematic because one, if you're trying to do something kinda gray hat and you are trying to, launch something, just spam the world with a million slop sites. But also if you're just trying to do something interesting, and you don't quite execute it well, all those things people are gonna recognize.

but at the same time, there are places where, you can make something with AI really, really good. 'Cause I don't think the problem is necessarily AI, it's bad AI. It's using the tool badly. ,

Brandon: It's, I think it, it's, it's 

Jeff: yeah. 

Brandon: intentional, right?

Jeff: Yeah. Yes. It's not being the hu- like, it's a great tool. It's not everything. but also like you, I, I, I love the point you bring up about there's areas you can deviate from and areas you, shouldn't.

If I land on a website and I'm trying to learn about, you know, what your product does, this is well researched at this point, like there's pretty good constructs. People expect to be able to go look at certain kind of sections of your site and [00:26:00] understand what your tooling does.

If you completely deviate from that, that's probably bad unless you found a really insightful way to do it better, in which case that alpha is really good.

And now the question is, how do you decide what's good and bad? And we need an answer, by the way. Like I, Brandon, we need the answer in, in 30 seconds please, okay?

Brandon: I, I think this is where I, I, I may, I may sound like a homer here, uh, but this is where session recording and watching people use your stuff is insanely important. Because if, if you don't see how people respond, you won't know, and you may not realize. If you don't ask them, they may not tell you that they see your pattern, and they think your pattern's weird, right?

Like, you may think you're fantastic with the polka dots and the stripes, and they may say, "You're not really pulling this off. , You need to try, a different approach." And so I, I think that how you know is you look at-- I mean, you can look at your competitors, still involve some research. But talking to the people that are using it, seeing how they're using it, being intimately aware of where people are [00:27:00] falling out in your funnel and why, looking at the analytics of, is there an error that's causing this?

Like, understanding that part and actually looking at your journeys is the best way to know why, , and the outputs of those and comparatively.

If you figure out for your organization what is actually important to the people who are using it, what are they actually benefiting from? What is it the part that you're deviating, how is that affecting what they're doing? Deviations you know about, and deviations that you didn't plan for, if you will.

And then figuring out, do I need to support those? Do I need to protect against those? , One of the things that I, I like to talk about is like understand your superpower and your kryptonite. Superpower, amplify it, like get your super suit, get the utility belt. For your kryptonite, protect against it, because it's not just things that you're weak at, it could be things that drain you and drain your value.

But you wanna understand those things, and you're not gonna do it unless you go looking, if you think you're gonna do what you did in 2020 and survive in [00:28:00] 2030, you're gonna fall off as an organization. And if you don't look for the places that deviation is going to occur, in machining they call it drift.

If you're not looking for the drift, you won't see it till it's messed up your tool.

Jeff: I'm glad you brought it up 'cause that way I didn't have to ask it and, and look like I was leading the witness here. But, this is your point, but it's the idea of there's a lot of things you're gonna know, right?

If someone tells you they hate what you did with the site, you're gonna know that and you're gonna feel that. But that's not usual., If people don't trust you, they're not gonna go away saying, "Oh, I don't trust you.

You seem sketchy." They're just going to kind of slowly accrete away and, and go away. It's the, it's the failure that never generates a ticket or never generates a response. -- There's no sign,

it's just people slowly melt away. And I'm glad you brought up, a huge benefit is watching people and understanding those patterns and , how do you understand that scale, , which is exactly what we were founded to do, was how do you get those insights without having to watch 1,000 sessions?

How do you automatically drive those? [00:29:00] But it's important because the world is changing if you are truly innovating and following what we need to do, you're gonna make mistakes, and the best way to find that is to watch for, like, where are people giving signs that they don't trust what you did, that you, that you messed up, and find those as fast as you can, revert or change or try a new hypothesis.

But speed is not the goal of product management. , It's using good insights and good understanding. But speed can help once you have that understanding.

Brandon: , I relate a lot of things back to parenting, but you expect your kids to grow up and to change, and so you look for those changes. You look for changes in their behavior, in how they talk to you, in the things they're doing, the friends that they're around. And because you're looking, , as a great parent, which I assume everyone's trying to be a great parent, , you go looking for those things and you try to understand them and respond to them, and I think it's a similar thing.

You need to look for it. You need to understand you're not, it's not all gonna be evident. To your point, like, uh, Creeding, my kids may not come to me and say that they've had a change in how they trust, , boys or girls or [00:30:00] teachers or us, or we did something.

We may see it, just little shifts in their behavior in how much they wanna play, how much they wanna talk, right? And so, like, it's a similar thing with your customers in that you need to be looking for shifts in behavior. Even if you've got, you know, the same process that's been working for a long period of time, you need to look at are people changing that process because maybe the standard has changed?

So there could be shifts in behavior that aren't just a response to me failing or me changing. Your response to the changing expectations in the market, and this goes back to what we were saying with the AI, like, there's a compression.

There is a, a new opportunity for transportation. There is now a new way to get information, and so the compression of the old way is happening. It's still important, but you need to go look at your site and look at your product and understand how that is changing the traffic that is currently on your site.

Is it becoming different? Is the mix of searchers versus browsers now different when you have AI?

search, [00:31:00] you know, bar becomes maybe less important, and maybe your site needs to be set up for browsing, and providing that latency. And so, like, these are the things we have to look for

Jeff: Yeah. And, and, you know, to your point, as much as, sure, no one wants to pay, , to deliver the experience of your site to a lesser page view or a lesser visit, at the same time,

we saw the time when people started to wanna talk about, "Maybe I block Google from indexing my site. Maybe I don't, I don't want them crawling my site." That turned out to be a bad answer for almost everyone who did it. would wager you, you want those agents on your site. You want them serving your information up, because at the base of it, as a business, we need to be discovered.

However that happens, I need it to happen. So it's on me to figure out what that new world looks like and how I deliver that information to people best, and how I use the channels that people can use. I can't choose what channel my customers use. I can only choose how I show up there, hopefully.

So,

um, Brandon, , this has been a blast, man. I appreciate you coming on. I think this is really insightful. Really good info about how do we think about the changing behavior of, [00:32:00] of traffic, , and how people see us on the web, and some of the most foundational pieces of building a business and think about building products even before, you know, you look at what is the feature, what's the thing.

It's just how do we, we be found, , how do we show to our users. , it's been rad having you on. Thank you for coming.

If people want to reach out and kind of dig deeper on this conversation, , is LinkedIn a good spot, or is there a better place?

Brandon: , LinkedIn is great. , As I tell my teams, my DMs are open, so anyone can reach out to me. I love helping people, , over career, big companies, small companies. , I believe in access to information for people who want to, to use it, so I'm happy to help anyone I can

Jeff: Love it. Well, thanks for coming on, Brandon. have a good rest of the week, and yeah, look forward to staying in touch, man. Thank you so much

Brandon: Thank you, Jeff