
Be the first to curate this episode — add a title and quick summary.
Add title and summaryNo information listed yet. Be the first to add who benefits from this content.
Suggest who benefitsNo detailed summary yet. Suggest a summary to help the community.
Suggest summaryNo questions listed yet. Be the first to add a question for this topic.
Suggest questionLast week on Dashboard, Shawn Busse said he thinks that trying to make your business discoverable on AI bots is “a fool’s errand.” So, this week, I invited Sean Campbell, CEO of Cascade Insights (https://www.cascadeinsights.com/) , a market research firm, to offer an opposing view. In our conversation, Sean talks about what businesses should be thinking about and doing to prepare for the not-too-distant day when most people turn to a generative AI tool like ChatGPT to find products and services.
Transcript from YouTube captions. May contain errors.
[Music] welcome to another 21 hats dashboard I'm here with Sean Campbell who is CEO of cascade insights which is based in Oregon and does market research for B2B businesses welcome to the dashboard Sean hey thanks for having me on and just to throw this in here I love what you guys are doing I've been part of the slack Community love the podcast just you guys do a super job I really appreciate that and you've been a great part of the community you've brought a lot of expertise that we didn't previously have and I really appreciate that um So speaking of which uh Sean last week I had Shan busy of kesis on and and we talked about whether businesses should be working on making themselves discoverable on AI Bots like chat GPT and Sean expressed the opinion that uh that would be a Fool's errand uh his his approach is more you need to build a brand and less think about platforms and tactics you responded to that Sean on uh the 21 hat sounding board our slack Channel uh that you disagreed and I want to hear about that but first tell us about your business Cascade insights what do you guys do yeah so we are a market research firm and I say that we get hired for pain or opportunity uh so we either have clients who've done it to themselves you know they've got poor sales poor marketing poor products or it's opportunity you know they've got a new market new industry new solution and for example on that you know we're doing a ton of gen generative AI research because we only work with the tech sector and we also only work with B2B companies in the tech sector so you know it's not a big stretch to say almost every one of those companies is trying to inject AI in some way to one of their offerings right sure and so there's a lot of research with potential customers prospects you know certain market segments and then in addition to that I'm also an adjunct at a local business school George Fox University and there I teach a B2B sales course and I also teach a generative AI course focus on its impact on various business roles and I'm also going to be teaching kind of a second version of that course apart to next year so it's a space I feel like I know pretty well I mean as much as anybody should say it's moving so blazingly fast that it's definitely hard to keep up on all of it but it is something that I'm kind of swimming in pretty much every hour of the day that's a little scary tell me c could you give us a quick sense of a project that you do for a typical client what what do you do for them yeah so like a real typical thing would be somebody comes to us and says uh let's just focus on marketing for a minute because that's maybe one of the more common things we might address I mean we might also address product deficiencies or sales deficiencies but you know we are trying to get a particular Market segment to engage with us you know and we are just somewhat lost either because we might feel we're not targeting the right Market segment for our offering we might feel we're sending the wrong messages to them we might feel that we're not doing that as effectively as our competitors right so it might start with those kinds of problems you know back to getting hired for pain right and then basically we will go in and say all right if that's the business problem what study could we design that would give you in the end really meaningful actions for you to take and those kind of bookends of Swords end up having in the middle between those two a lot of research interactions and those might be in-depth interviews with potential customers right or even lost customers or it might be a quantitative survey we do we might do focus groups but the methodology is important but to me what's really really important is that we're really locked in on what the business problem is you have and then therefore we can actually design a research study that'll turn around and give you actual things that you should go do to go help with that business problem and that's that's basically what we do every day around here we got about 13 people working for me now across every time zone in the US at least and um and it's a lot of fun you know we get presented with a lot of challenging problems you know we have an informal motto uh because it's not really a formal one when you hear it uh we deliver bad news to good people and the point is that like with research you don't want just everything to be confirmed right like there's there's a point at which you want somebody like us to say but there's this thing you're not acting on there's this thing you don't know right and in a way that usually lands is a bit of bad news because no leadership team wants to think that there's something they don't know or they're not acting on right but at the same time that leads to really good outcomes in the end right so that's where that little like we deliver bad news to good people and the good people part is just you know I I feel like if you're if you're willing to put your business under the microscope a little bit and let somebody else kind of scratch at it and and even more so have somebody else talk to your customers or competitor customers and have them come back really candidly and say you know you're not really ready for prime time here you know that takes a little bit of Courage too but so that's what we do all that it's really interesting I love that line um so let's get to that question at hand uh I think if I recall correctly in your uh comment on our slack Channel you basically said that before long very soon the way people are going to find out uh about other businesses is by asking for a recommendation from someone like chat GPT how close are we to that I think we're already there in some market segments I mean you know in the segment that we're in technology I hear it all the time you know if you've got a buyer inside a Microsoft or an Amazon web services or even like a verticalized uh Focus SAS company you know that's focused on a particular vertical they're fairly Savvy with all these tools right where do they go when they have a question they go go to chat GPT you know now that pretty much all of these tools right there's some exceptions in terms of like how much access they have to the internet or how much real time access they have but for the ones that have real-time access you know a chat GPT things like that it's a very logical leap to say I'm asking this thing to solve problems for me all day why wouldn't I ask it to find a vendor and the problem with that a little bit what I responded to is that AIS don't necessarily Buy on brand now you can quibble a bit and be like well brand is content and messaging and all this stuff yes yes but like sometimes people think of brand as like you know how do I feel about the brand and what does it do for me and you know what kind of emotive aspects do I have and do I like the colors AI doesn't really care about colors you I mean it's not going to care that much doesn't going to really care about the typography that you have on your website right it's going to care a lot about the content and the meaning behind it and how Target it is and it's how focused it is on solving a particular problem and targeting a particular ICP and a set of buyers right and that's that's part of marketing too but yeah it's kind of scary for a lot of people I mean we've spent decades now trying to optimize for Google and basically try to guess at to what Google was going to rank our content for right I mean it's there's a whole industry built around guessing what Google's going to do and now we have a whole whole new player in town that isn't just Google I mean they have Gemini for sure but like now we have a bunch of different front doors that our customers might walk through in an attempt to find us and that's scary it also presents some opportunities you know if maybe you were struggling to rank so there's a lot yeah there's a lot there's a lot there that's going on for sure I don't like to put words in anybody's mouth but I I think Sean busy would kind of agree with you that part of what you just described there as being part of marketing and branding is what he was referring to but but his larger point I think was that just as kind of the the SEO environment kind of got polluted in certain ways it's not what it was when Google first introduced AdWords and you know payperclick uh was a lot easier to use and all that um his point uh that he made uh here on the podcast was that being a generative uh AI uh bot is an expensive thing to do and if they're ever going to make money they're going to have to start charging people and it's going to be a lot like SEO where the big brands are going to take over and his point was that that's going to be a losers game in the long run and that you you need to build a brand uh so that people find you in other ways and you know your your name means something how do you respond to that the problem I have sometimes when I hear somebody say that right is like and I've had this before AI right like you need to build a brand and it's like it really depends on what you're driving at when you say that because if you say Cascade insights you need to build a brand I I tell you we've been a very successful company for like 19 years now and you know what I don't think we have a brand that a bunch of people in Tech know you know we're not the coke of market research and so then it starts to beg a lot of questions about like well what do you mean do you mean a brand in a narrow space that's really wellknown for that what exactly does that mean right because we're also pretty narrow I mean anybody listening to this probably maybe's never even talked to anybody who does what we do as narrow as we do right um and that's fine right so we're already pretty narrow so I think I think that's where I have more of an issue because we're going to find a world where the AI is going to take some pretty darn logical decisions in terms of what it surfaces for you and you won't necessarily have 92 pages to paginate through right not that that was a great experience either I went to a Gartner conference you know a million years ago and it was the thing that stuck in my head about it when a guy was talking about Google and seos he said using a search engine is like staring into a tiny hole in a door and you're looking for a penguin and somebody on the other side holds up a picture of a penguin and says this one and then does it again and then does it again and then does it again and it's like yeah that's that's how horrible it is right because you got to go to Every link you have to look at that page decide if you want that vendor I mean just think of the Allure people are going to have when they go to an AI and they're like just tell me where to go well if they like the answer they get right it's almost like asking for a referral right like it's like the the glory of a great referral right you're like you just save me all that at the end of the day and that's scary because it it changes I think what your marketing mix needs to be and what you need to invest in to make that AI rank you know rank what you want it to rank and and I'm not disag about the paid thing one quick thing on that sure these things cost a blazing amount of money to operate but on the other hand I think it's a little shortsighted to assume they're going to use the business model that preceded this is so unique and so differentiated and so changing to just everything and everybody's noticing that maybe we'll have a different way of monetizing it maybe it won't just be inline ads right so like you know I don't I don't know if it quite ends up there maybe that is where it ends up but maybe not but even from what you just described it sounded scary to me in this sense you the the point you were making is how great would it be if you just ask who should I use and you got an answer hey but if you're only getting one answer it's not going to be very long before that one answer you won't get one answer though like you you know because an AI is interactive you know a lot of times you can just say like give me eight eight great doctors that deal with Advanced knee surgeries right and it'll give you eight but it doesn't necessarily give you their website immediately right maybe the Link's in there right and then you say to it tell me the background of these doctors and tell me what their experience is with my particular type of uh uh ski based knee injury right all of that changes kind of this infrastructure that we built up that assumes people are doing that on our website and that's I think really the big thing that's differentiated right like you what you see in Google search is a really narrow subset of your website kind of by Design because it's going one penguin two penguins three penguin right and it's like and it's this meta a link but with an AI it's interactive which which is obvious right but at the same time that alone very radically changes the buyer's Journey right they may spend 80% of the funnel just hanging out in an AI chatbot window that's a big change all right that makes sense I buy that um tell me this if a business is thinking about what it needs to do to prepare for this as it comes or if in fact it's already here what should businesses be doing uh to make themselves discoverable through this channel the very first thing is figure out if you're discoverable like in other words it's only 20 bucks a month you know what I mean for chat PT Plus right you know what I mean like so sign up for that sign up for Gemini sign up for Claude you know because the first thing is is it a problem for you right like maybe you're one of the fortunate ones right which should be some people at least right because what what are these things built on they siphon the entire internet now they're going to interpret the internet differently than Google's SEO algorithm so you you will guarant it's guaranteed that it'll look different and you may not rank but first test and then the second thing I would do is you know think like an AI and maybe even spend a little bit of a time understanding how these llms are constructed I'm not talking about like eons of time I'm just like go watch some TED talks to start right go I mean you know there's more you can get into but understand how one of these AI platforms think it's not super hard there's all kinds of content about this like why does it serve me up the things it does why does it kind of think through it that way and understanding kind of fundamentally how these things are built might give you also a little bit of a clue of the types of content you need to build and the last thing I'd say is things that are meaningful things that have depth things that have context you know these also positively impact SEO and they also positively impact an AI but to go back to what I said earlier does your typography matter like it used to if somebody spends 80% of their time in a chat window no because they won't see your beautiful website not until the very end and by then maybe they haven't picked you they picked two other guys and they're looking at their websites so that's that's where I think it's really interesting you know what I mean like a lot of the things we've invested in to show somebody our website don't really matter to the AI nearly as much right and if the AIS were somebody spending their time in that very plain vanilla chat window interface that's got to change some things well well let me give you an example um the example is me and 21 hats uh I did sign up I have uh I have a paid version of chat GPT and I asked it uh could you recommend a peer group podcast uh uh for me I got a podcast that features business owners who have a peer group conversation about what's actually going on uh on a day-to-day basis in their business uh not a a podcast with uh someone looking back on on you know Distant Memories of their struggle and then their future great success and um it gave me a list of five podcasts that in my opinion are fine podcasts but didn't really fit that description um which I think is a fair description of what I tried to do here uh at 21 hats um and then I tried it again a few weeks later and it gave me a uh gave me two responses it said we're going to give you two choices here the the first response is uh another five podcasts like the first time which were um you know fine but but not really responsive to what I was asking and then uh a second response uh which started with it saying you know this is almost a direct quote you know that sounds like your podcast the 21 hats podcast and it listed that and it described it as a peer group conversation and then it listed a few others which I think were closer to the mark but but not right on target um and it it asked me which of those two responses I preferred I said I preferred the second one figure right exctly yeah right uh and then I asked it um would you recommend the 21 pod uh 21 hats podcast to someone who was looking for a business owner peer group uh podcast and it said yes it would and I asked it well what what could I do to make it more likely that you would make that recommendation and it actually gave me some you know really generic uh suggestions um and I I know you got to follow up and I probably should have but I didn't have time at that moment and I haven't gone back to it so I guess I'm asking you the same question what would I need to do to make sure that it actually makes that recommendation to other people not just the host of the podcast well one thing is we're dealing with a blackbox to some extent like we have always dealt with with Google right in the sense that there's there's always going to be a certain amount of inability to connect a with B because we can't really peer into the platform but I'll tell you another thing that makes it harder sometimes with llms and you know I'll just direct folks to this because this would like nerd out for like way too long but if you if you look into how like vector databases work and how the plumbing of these things are actually constructed right and I I could talk for longer about this but I but I won't because it just everybody's eyes glaz over right who's listening but the point is is that um there is a challenge even for people who construct these things sometimes in determining exactly how it got to the path of XYZ answer right um even if ultimately that answer is a logical one right and a useful one and you know if anybody's curious about that just go start searching about Vector databases and how they're used in building out llms and tokenization and a whole bunch of other things that you'll bump into pretty quickly um and it'll kind of open your eyes to a little bit of that and it's at the heart of when people talk about llms hallucinating it's one of the reasons that's possible okay even though the amount of that has greatly degraded uh not degraded de gradely lessened um since you know let's just say a year ago or whatever um but I think the biggest thing to focus on is think about meaning and depth and value right which if you write to those things if you create content that hits those particular call it little you know triangle of values that's going to be something that I would imagine in most cases a lot of the generative AI platforms that are out there now are going to see as something that they want to surface right um because these things are driven by content at the end of the day and so you know that's that's fundamentally what's happened we have something that is super powerful has impressive Computing abilities and it siphoned the entire internet um you know before it did that and so ultimately that's where you want to go but some of those softer squishier things that we use to impress humans maybe not as much but it sounds a little bit similar to building a brand and practicing good SEO habits I mean I think what I heard from you is if I have positive experiences with my customers and they say good things about me on the internet and if I get written about in the Wall Street Journal or the New York Times or somewhere uh that has some reach and credibility uh that all those things will make it more likely that I show up um either in an you know a Google search or in an llm is that well to to a degree like I said I think I think the distinction I'm trying to make is that you know when I hear build a brand you know maybe this comes out of the day job you know that I've got like if we do a brand study like if we analyze a company's brand um we may be looking at things like awareness and usage and consideration and advocacy but the team that is focused let's say in a company on building the brand may take into all those things into account they may say like you know brand development includes everything that's ever said about us that's our brand but there's also a slightly more narrower version of that definition that you equally see which is like our brand is you know kind of basically our core brand promise and it's our core you know value props and it's the colors and the typography and how our websites laid out you know the kind of things like a web development firm might share as being critically important now they wouldn't neily discount the value of good content and good SEO but I think it's just you know you're selling to a machine even more so than you were with Google in a way right um and I think that's that's got to change a little bit maybe at least where you put your emphasis when you're thinking about like where's my marketing dollars going to go and what do I want to create and who am I creating it for can you tell us a little bit about how you've approached this for Cascade insights what what steps have you taken uh to try to improve your discoverability well I think we got a little lucky because that's usually where we have spent our time over the years now on the other hand some of that is a fory function of being a smaller company that sells to large companies right you know like I can't possibly spend enough to make you know the brand of a dozen or so person companies stand out to everybody in a large multinational I mean I just there's just not enough Revenue in the building I don't I don't care what we were doing you know what I mean sure and so what I've had to do is basically write stuff over the years that's incredibly on point that's really focused at a narrow audience that provides a lot of value and frankly to somebody who's not on our lane they wouldn't want to read it you know I mean like that that's to me a big piece of it right you know it's almost like you you have to write stuff that is incredibly attractive to one audience and it's almost really unattractive to somebody else right it's not generalist writing and that the minute a lot of these platforms came out you know the people on my marketing team we started looking at them because it's you know we're doing research in the space all the time we're like huh we already rank pretty well that's interesting you know what I mean and what's really weird is we rank better than some of the bigger Brands the ones that I know that are out there they don't show up and um you know I again you know it it's a difficult thing to peer behind the veil of these things partly by Design partly just kind of the technologies that are actually used to construct these things and partly it just you know there's going to be IP issues that we're never going to get around in terms of what these companies are going to say but um but for us it's worked out pretty well I mean in a way it's actually great we already had really good SEO ranks for the audiences and terms that we wanted but we actually rank even better in the Gen tools so I to me that's a win at the end of the day and it also suggests that what you were doing for SEO reasons translated correct correct again I think I think as long as you were you were you were focused in narrow right you know I think of like taking nothing against you know guys that like blow Moss off your roof right but we hired somebody to do that the Pacific Northwest just grows Moss left and right says couple couple obvious characteristics as to why it might right and so like you know everybody around here is blowing Moss off their roof it feels like every six months right otherwise you just have a perfectly green roof it looked pretty for till your roof fell apart but like um that guy who's just broadcasting we do moss removal I don't know how that's going to play out you know um there's a lot of things that are unknown you know how good is are these gen platform is going to be at localized search you know that's been one of Google's kind of shticks for a while right you go in and you're like find me the top you know Moss cleaning companies in my area right and this isn't me being dismissive about that like like I mean you know what I mean I'm not being dismissive about that as a line of work at all far from it I grew up from in Chicago in like a really bluecollar neighborhood like I tell people like if they watch the movie Rudy that was basically like everybody I knew growing up is like like could have been an extra in that movie basically and so like 100% my dad even wanted me to go to Notre Dame and I didn't go and I didn't well I didn't have the grades but there was that too life regardless but he would have wanted it like just like the guy in the movie and so um you know I think my point is is like that I I worry about that side of it right you know like because because fundamentally to go back to something I said earlier right everybody probably's heard the term the buyer's Journey right and the idea that that has changed and that a lot of that is digital now and a lot of that is on the internet and we've known that for a long time now and how that penetrated different Industries and jobs um uh jobs to be done so to speak you know the people were hiring uh folks for obviously was um you know varied like certain industries that happen quick and certain industries that happen slow but it's definitely true of almost every industry now right that you know you expect customers are going to go through the internet for a while before they get to you now it's a whole different set of doors right I mean for a long time like all and think about it all we ever cared about was Google I mean who really uses Bing at the end of the day I mean I've wondered about that right there's a door and and yeah maybe it was hard and it was complicated but even that alone is kind of frightening right because there's no guarantee that chat gbt and open AI win this battle who maybe it's anthropic with Claude maybe it is one of the big guys like Microsoft with co-pilot or Google with Gemini or maybe it's somebody we haven't even heard of yet it's still early have you seen differences in their approach I mean is it are well sure sure sure well because because CL doesn't have direct access to the internet currently um Claude is like the way we used to talk about chbt before it had access to search you know where like its data set was capped at a certain window right right and then um and Gemini I used to think was pretty horrible until like like a lot of people did until just somewhat recently and it's really strong I mean if no one to looked at their tool called Deep research like say you're going to go make a complex purchase decision of something kind of you know fairly meaningful meaty B2B millions of dollars or even just you know tens of thousands of dollars you can go into deep research and say I would like a report on everything in this space and it will just churn away and go grab like and it'll show you the links it'll grab 80 90 links and pull it all together you know what I mean and give you a nice report I showed my students that in my gen class I teach at the University and I I had to quickly say just to be clear in your other classes AIU needs to be approved by the professor you know what I mean like before you just I could totally see the eyeballs go wait a minute I don't have to write reports anymore no no no no no that's not what I said um but but anyway the point is is like yeah there are differences and that's why I think you have to explore perplexity is another one a super powerful tool right they are clearly moving by what you can kind of see publicly toward more of an ad model right um that's a really powerful so you know but that goes back to now everybody's got to go through nine different doors you know at that point so has your approach that you said uh you know you saw you were ranking right from the beginning has that been true on all of these uh different tools no no not necessarily um you know so different different tactics work better on different well well and part of it is um and this might resonate with listeners too or or at a minimum I think it resonate with with you I know just from watching you in the show and the Community right um this will tie in in one second I promise it only take 10 seconds to here like I I've said for years like in as an owner of a moderately small company if you're not used to something sitting in the corner rattling making noise that you can't quite get to yet you're going to lose your mind because there you can't possibly fix everything you want to fix as a business owner every day right so you have to figure out which when you're going to tackle a and be and like as you might have seen on your show some guys are not good at making that decision and then life sucks for them so to me where I'm at now is I know we don't rank as well as I'd like in some of them so then what as a 12 person company do we do right like we can only create so much content at at X volume right and so what we've done is for our audience and our ICP and who we target we've tilted our content calendar a bit and changed some of the topics that we're going to discuss for what we believe will lead to some better ranking in those engines right but the problem is that takes time right I'm just really thankful that what we've been pushing out for content for the last 19 years or you know not that it really cares about something we really wrote 19 years ago but let's say the last five years right is something that in the end when they turn these things all on and they gave us access they were like ah yeah you guys are all right so I'm just thankful for that got it Sean I expected to get you out of here a lot quicker I apologize I just couldn't stop asking questions no no no I'm glad to pay it forward man you guys had been great and I seriously you know it to anybody who's only listened to an episode or two go back listen to The Back catalog these guys are great they're super Sean thank you very much I appreciate that Sean Campbell is CEO of cascade insights which is based in Oregon and does market research for B2B businesses not for Moss blowers uh but we're going to have you back and maybe at some point you'll have an answer for those Moss blowers um I'm sure that would resonate with my audience awesome thanks man thank you Sean take here everybody have a great week [Music]
About 21 Hats
21 Hats is an online community for business owners. Entrepreneurs have to wear a lot of hats to build a business—but some hats fit better than others, right? When you’re not sure where to turn, the 21 Hats community is here to help. The 21 Hats Morning Report scours the web every morning for the most important stories for business owners (https://21hats.substack.com/p/coming-soon). The 21 Hats Podcast has been tracking six businesses throughout the crisis in weekly conversations (https://21hats.com/).
People who have contributed edits to this page.