How to Stop Guessing and Start Testing with Casey Hill, DoWhatWorks

Speaker A: SEO is fundamentally changing. And as a byproduct, there's been this kind of narrative out there that everyone has been saying, like, traffic is dead, traffic is not important. And from my own experience, I actually see a very strong correlation between all the core parts of the funnel. When I have months where my traffic is dramatically higher, my leads are also higher, my demos are also higher, and my sales are also higher.

Speaker B: I'm Jonathan Gandalf, and welcome to the Content Cocktail Hour. Powered by Audience Plus. Our mission is to shake out the deepest secrets of B2B marketing professionals and stir up fresh conversation, all to help you connect your brand with the right audience at the right time. Let's raise a glass. Welcome back to the Content Cocktail Hour. Thanks for joining another episode. This is Jonathan here, co founder and CEO of Audience Plus. I'm also excited to welcome back Casey Hill, a recurring guest. Now, the Content Cocktail Hour. Last time we talked to Casey, he was leading growth at, ah, activecampaign. Now he's the CMO of Do what Works, which I, uh. It's fascinating business I'm excited to talk about. You're going to walk away from this episode with some, like, real life lessons that are very tangible, tactical, that you'll be able to apply. Not only you. I'm excited to apply some of those lessons myself, so why don't I stop talking there? Kasey, tell us a little bit about your role, your background, and what you're up to at do what Works.

Speaker A: Thanks so much for having me, Jonathan. Um, so, yeah, I'm the CMO over at do what Works. We're a technology that essentially can detect any AB test run by any major brand on the web. When I first heard that, by the way, when I was in interviews, I was like, this seems really sketchy. Like, are we hacking people? No, this is all based on public information. We have a patent on a web crawler and an algorithm. When people run split tests, basically those are detectable variants. And we just are taking this in mass and looking at what brands decide to keep. So you run A versus B. What version do you stay with? And then based on that, we look across hundreds and hundreds of different brands in the same industry in the same exact application, and we make that easily searchable. So if someone's like, hey, I'm designing my homepage. Should I have one CTA or two CTAs right there on that hero section? We can say we looked at 493 different people in your space who tested it, and this is what the results say. So it's really great for helping guide decision making and hopefully make AB testing more efficient. Uh, Optimizely put out a study that said only 11% of split tests beat their controls. So what this means, essentially, is people are doing a ton of testing and spinning their gears because they're ineffective. So our thing is not to replace testing. We want you to continue to test. But hopefully some of the ideas that we talk about today might spark a more educated test across some of these different areas. That's the idea here.

Speaker B: That's a really interesting perspective of not only, like, are we going to help aggregate data from tests, we're going to help you run better tests. And it becomes this, like, very virtuous cycle of helping you get data from

Speaker A: test to run better tests and get

Speaker B: better data and like, continue to iterate on that. I think that's, uh, really fascinating.

Speaker A: 100%.

Speaker B: Yeah.

Speaker A: I mean, we saw this. Not to go, you know, down the rabbit holes, but, like, in the ad space, I think it was back, like, 2016, 2017. People who are running paid ads might have remembered that suddenly people started to lean heavily into these algorithms because when you manually targeted, it was actually less effective than using big data to inform decisions. And obviously there's some differences between that and what we're doing, but the idea is to use data to make better decisions. And I think this is something that we're bringing into the AB in the split testing space.

Speaker B: Fascinating. We'll dig more into all of that, but first, this is the content cocktail hour. Kasey, I have to ask you, what are you enjoying drinking nowadays? Yeah.

Speaker A: So this is going to be a boring answer, but it's coffee. I'm a coffee and water guy. I wish I had a more marvelous answer, but, uh, that is the. That is the reality and the truth.

Speaker B: I can appreciate that. I just had my diet soda as well for the afternoon, so, well, caffeinated. Also mixing in some water. My desk is always an array of different beverages on it, so I appreciate that.

Speaker A: Love it.

Speaker B: Let's, uh, let's dig into one of the tests you ran. Well, and actually, before I do that, I just have to say I love the name do what works. Uh, a little bit of a personal anecdote here. I remember I, uh, was leading marketing at a healthcare analytics startup prior to starting the Juice, and we were interviewing CMO candidates, actually, and I was fortunate to sit in on those. This person was going to be my. My boss. I wasn't ready for that seat. That's a separate conversation for a Separate podcast episode or maybe therapist, but neither here nor there. Uh, I remember, like, you know, we'd always ask them, you know, some generic question what their plan was when they came in. I remember there's one candidate, she just said she viewed marketing's job as doing more of what's working and less of what's not working. And I remember, like, it sounds kind of silly to say out loud, but it does feel like that's such a obvious but difficult thing for marketers to do. I always feel like we've been given a stack of chips, and it's like we're running all of these experiments, and it's like, as soon as you find one that's working, like, shift your chips to that experiment. So I just love the simplicity of what you all are doing. And I think there's something really powerful and unique to the marketing role of, like, experimentation that sometimes gets overlooked 100%.

Speaker A: And I think that one of the things that is really important to note, too, is that there's this place for creativity. Creativity is super important. You want to be unique, you want to differentiate, you want to stand out, and that comes across in your brand in so many different ways. And sometimes I think one of the misconceptions is I'll talk to people and they're like, well, I don't want to see what everyone else is doing. Like, I want to be unique. And what I tell people is there's absolutely spaces where you should be unique. But something like the clarity of your button text and the fact that saying, like, start a trial versus get started, start a trial is more clear, for example, like, those are the kind of logistical things that exist all through your customer journey that you really just want to bring clarity. And if you can, shortstop, if you can increase by 0.8%, your button at this stage and 1.5% at this stage, and you can take the exact same information, but you can use a grid instead of a carousel. Like, all these small logistical things, those really matter. So it's not about. I want to emphasize that I'm not saying this like, you should replace creativity. You should blanket copy your competitors. That's not the idea here at all. Instead, it's to look at those core logistical pieces along the journey and understand, oh, wow. I was thinking about putting maybe a video above my pricing page tiers, but I looked at 100 different brands that did that, and it always loses. Maybe the best spot is to put it below those. Like, you know, you just kind of are, uh, trying to do what Works based on that information.

Speaker B: Yeah, it's a classic cliche that constraints can breed creativity. And that's kind of what you're providing is like, here's the data constraints of what does and doesn't work. Now you get to go be creative within that. So let's dig into some of these examples. There's one, uh, specifically I wanted to start with, uh, I am guilty as charged here. So again, this is one where maybe I'm going to walk away with some to dos and action items myself. But you found that social proof. I know there are different versions of social proof that we'll probably talk through, but social proof actually doesn't always work for brands. Can you tell us more about this experiment that you're, uh, kind of gathering the data on?

Speaker A: So one of the most surprising things when I first got into this huge database of tests and was looking, you know, we got tens and tens and thousands of tests across all these different industries, was that customer logos were consistently losing, like at a very high percentage. Like 8 out of 10 times. When brands were testing this head to head, they were losing. And that was kind of stunning to me because almost every website, especially in B2B SaaS, which is one of the, the core areas that I operate, uses logos. And it's typically always the same thing. It's right there on the top. It's right after that hero section. You got your hero. Then right below you got that banner. It'll often say trusted by a hundred thousand plus brands or trusted by ten thousand plus brands. You got a scrolling logo bar. Right.

Speaker B: Are you reading off of our website right now?

Speaker A: I am not. But, uh, having looked at a ton of websites, it is, it is a good bet. And so that was super curious. I was like, wow, I wonder why that is losing. And there was some really good takeaways when I dug into the data. And I can provide some specific alternatives for folks that I think will work better. So the first kind of question is though, but like, why is it not working? I think one of the things is it's a static asset. There's no ability for you to engage with an average logo bar. So you're just stuck there kind of like trying to find if there's a logo that jumps out to you. And if, let's say the first three that scroll by, sometimes it's static, but many times it's kind of scrolling by. If the first handful you don't associate with, then it's kind of almost a negative signal. Right. And a lot of times companies try to do their biggest logos. But if you serve big companies and small companies, I'm a small company. I come in, I see like Stripe, Wells Fargo and Disney in the company's head. They're like, wow, they'll be really impressed. In my head, I'm like, okay, that doesn't seem like for me, right? There's this association problem. There's also this challenge up. There's no ability to validate it because you can't click in because they're not attached to anything. You don't know, does this person work with State Farm, the big corporation multinational, or one tiny branch in North Dakota? Do they currently work with them or were they a customer eight years ago? And they churned like. There's so many question marks that come. So my guidance for folks, if you really want to use customer logos is a couple fold. First, I want to provide two good reference examples. There's a company called Clay and there's a company called Hex. What both of those companies do is they attach case studies and they have little like pluses next to them. So you actually can see like a short preview and then you can click in and get the full case study. And for both Clan Hex, they don't have every single logo with a case study. But the fact that at least a handful of them you can interact with, it gives this sense of trust that you're like, oh, I can dive deeper, I can interact with. That's a really cool dynamic. There's another company called Seven Shifts that has a, uh, toggle by type. So they serve like restaurants and they have one for like pizzerias, one for coffee shops. They're speaking to that problem and you can toggle between based on your business type. So that speaks to the like, I want this to be relevant to me. If I'm coming in and I'm a coffee shop, Coffee shop might operate very different from pizzeria. So it's really nice to be able to just click in and immediately see a business just like mine. So one of those two models I think is more effective. The last thing I'll say, and this is, uh, just a quick personal preference on logos, is everyone puts them at the very top. Now the way I think about the journey on a page is you need to sell your value prop and then you use the trust to bring them over the line. Starting with the trust before you've even sold them, I think is going to lead to diminishing returns. I think the reason everyone puts it up there is because that's what everyone else has done historically. And it's just a norm. They're like, oh, I look at every major incumbent, they stick it at the top. I'll stick it at the top. But I would argue to you that it's more important to sell. What do you do? How do you do it? Why is this for me? Sell those three core value points and then use the trust to carry people through. I think you lose something when you just tack it right up at the top.

Speaker B: Yeah. This is fascinating and I think we're about to have our first ever screen share in the history of the content cocktail hour here. I'm going to try this for those uh, watching here. Here's a. I pulled up that Clay website and yeah, I love this experience of like the actual. You can click into the case study from the logo. That just seems like one of those like, oh yeah, duh. Opportunities uh, for brands here. The little preview of the case study here. I think that's really uh, good. So I appreciate that real life example. It's interesting what you mentioned about the norm. I joked like that's, that's the structure of our website. It's the structure of almost every B2B SaaS's website and it becomes so ubiquitous that like we stop even thinking about testing or like we forget that it even can be tested. Are there other examples like that where like we've maybe like it's become so ubiquitous that we're not testing it and then we should be testing it? From what you've seen it do what works.

Speaker A: Yeah, I mean there's a ton of different categories so it, it often happens in cycles. So when I was starting out in marketing as an example, the one CTA thing was always like, choose your highest value path, take people down your highest value path. Don't have a lot of distraction, make sure you have focus and take people through. Then I started to look at a bunch of tests and when people tested 1 versus 2 CTA at the top in the hero section, 2 CTA was consistently winning. I think the reason is if you have diverse traffic coming to your website, they're going to have different levels of intent. Right. In the case of a software company, some people want to hop in and do a demo, but many people aren't ready to talk to someone yet. So you want to provide that self serve module as well. Like it makes logical sense. But the one CTA was a norm for a long time. The same thing happens with pricing for a very long time. Everyone did monthly pricing. That was the default. If you go back seven to eight years, then a handful of the top SaaS brands started to say, no, we're going to do default annual, but displayed monthly. Right. It almost seemed like a little bit of a gimmick at first because it's like, wait, is that. But now if you go just, you know, you run this experiment yourself, go pull a list of the top 100 SaaS companies, about 70% of them default annual displayed monthly. It is now the norm. And I think brands, they would rather get more people on annual. So they like to kind of push towards that and they're almost allowed to do it now, quote, unquote allowed because everyone else is doing it. That expectation is not like, oh, this is kind of out of what is common. So there's many areas like that that are evolving and that are changing over time. And I think the critical part is it's important to really understand when you're looking at that, you know, say, competitive landscape, you want to understand what is actually being tested versus what is just like common. Right. That's what we're kind of trying to get at because one of the things that happens all the time, we actually saw this happen with a Monday.com competitor, I won't say their name, but they did something that tons of companies do, which is they just went to their competitors manually. They looked at what their competitors websites were doing and then they went into their website meeting with the rest of their team and they said, hey, our major competitors doing X, unfortunately they copied the losing version that was being displayed on the specific day they went that was being shown to let's say 10%. I don't know the actual percentage, but let's just say 10% of people and that's what they copied because they didn't actually see what was winning, they just went and they just like copied a thing. So we're definitely not recommending for people to just blindly copy because you don't know if you're going to show up on the day that they're showing a winning or losing variant. And as someone who looks at a lot of top brands testing, I can tell you they run a lot of tests, right? Like many of these brands like are running 20, 30 tests in a given month. So you want to be cognizant of that.

Speaker B: Yeah, it's really interesting whether it's the case study example and like choosing the logos to show or the annual versus monthly pricing or you mentioned kind of the uh, showing the value props even before the logos. It feels like A lot of what we, as brands, ourselves included, are doing is we're doing what's best for us, and then we're testing what's best for the visitor or the customer. And it's like, that is what's winning, right? Is like we're so like, um, myopic in how we view what we want to accomplish, whether it's the most valuable path or the quickest path, but it's like we need to stop and pause, like what's actually best for the customer. And that's typically what's winning in these experiments. So really interesting there how you've kind of like the, the, the perspective at which you're providing on some of these tests.

Speaker A: A hundred percent. I also think we gotta be really careful of unintended signaling. So this happens a lot in other types of social proof. Like when someone says 4, 4 stars with 300 plus reviews. What I think people don't realize sometimes is that, like, people are using social proof to build trust. But the dynamic of that can sometimes, like, if I see four, four, I'm kind of like, ooh, if I go to a restaurant, I see 4.4. I'm like, that's okay. It's not like a huge negative, but it's almost leaning towards that. I don't see a 4.4 as positive. Right? That's how much just our system. Like, if I'm looking at a restaurant, if I see a four point, about seven, that's pretty good, right? The other thing is volume. If I see. I saw a website, big company recently, and they said 400 plus reviews. And I remember my first thought instantly was only 400. I would have expected they would have had thousands, tens of thousands, based on how big I thought this company was. So the other thing you want to be careful with is, especially as a startup, especially as a smaller company, a lot of times you're trying to use social proof to kind of show that you're fighting above your weight class, right? Like, we're, we're working with top brands, but if you inadvertently make yourself look smaller by including those numbers, you just want to be aware that that is something that can also happen. So I think that's the other thing that can sometimes happen with social proof. Like, we're very. Buyers are very kind of feelers up to being sold to. And there can oftentimes be unintended signals that occur with these kind of things as well. So something you want to be cognizant

Speaker B: of, unintended signals reminds me of one of my Favorite, uh, stats or sayings is that everybody who confuses causation and correlation dies. Um, which I think is a good lesson for everybody involved. Uh, okay, so there's another, uh, study that you have on the Duo Works website, which we'll link in the show notes and I encourage everybody to go check out because there's some really interesting stuff there. AI, I think it's a record. We're almost 17 minutes in this podcast and we haven't mentioned AI yet. So, uh, kudos to us. But your guys, um, data is showing that AI messaging is actually losing, but brands are using it anyway, which I. There's some, um, maybe some reading between the lines there that needs to happen. But why don't we first just talk about kind uh, of the data that you saw, and then I'm going to share some data from our own findings as well that I think is parallel to what you all found.

Speaker A: Yeah, for sure. So I want to make a couple clarifications. First, I want to make a disclaimer that the AI landscape is changing incredibly fast. Right? So any trends, any things we can detect are evolving. So recently, what the data has shown, which is, by the way, has been echoed by a lot of other parties outside, is that generic AI language, AI powered AI first. All these kind of AI native are basically just filler. They don't really work, they don't really drive, they're not highly persuasive. Instead, if you take a look at, like, Zendesk's pricing page, they say something to the effect of our agents provide 24. 7 coverage for, like, customer support. That type of language is winning because it's tied to the outcome. So what you want to think of is if you're using AI language, nobody cares about the name of your AI thing, right? They care about the output. So I also shared a test recently from hootsuite, the social scheduling tool, and they tested two versions. One was like, meet Owly AI, our new AI bot that, you know, whatever. And the other one was just based on what that AI agent did, right? And no surprise, the one that was focused on the outcome one, whereas the one that was just kind of focused on, like, we're introducing this new thing, Lost. And we saw the exact same thing happen with godaddy. Godaddy did this big campaign around, like, I think it's called, like, Arrow Aero, which was their whole new AI thing. And their first iteration was all about, like, meet Arrow, right? And the focal point was all around that. Whereas they changed their messaging to, say Something they still talk about Arrow, but they focus on like get uh, domains for like $0.01 or like they have some core output thing first and the agent is just supporting and facilitating that thing. So I guess at uh, a high level, what I would say is avoid generic AI language and then also try to make sure that you tie in the outcome. But then I also want to get back to the question that Jonathan kind of teed up, which is like some people are testing and the AI version is performing better, but it still gets cut. And why is that? I think we're existing in a complex time where people all want to be heavily associated with AI. Anyone who follows the venture markets knows how much funds have been shifting to pushing as much investment as they possibly can into AI companies because they see the replets, they see the lovables, they see these companies that are just, you know, in one year going 0 to 50 million and they're looking for that next huge kind of breakthrough company. And so there's a lot of funds. And so a lot of times what happens is an executive steps in and says, well, I don't care if it's losing, like this is what we need for the optics because I'm going to go try to do a funding round and I need you to back me up. I need our content to make us look a certain way. So it's a complex thing and how I would have teams navigate that is when you're thinking about making these changes, get your stakeholders involved before. So if the founder is really going after a funding round and he needs that, maybe the idea is, hey, we'll absolutely still include AI messaging, but maybe you convince them that you're going to put it in a slightly different spot other than the hero. Because this is what the data says, right? Or you say, let's run this test and if this is what the data says, can we do this by getting that permission upfront before you go into the test, then you have the ability to more effectively implement the higher converting changes versus if you're like, hey, we ran this test and this is what it says, but you're on a parallel track to what some um, board member, investor or executive ones. That's when you get the thing where it's like, oh, this version without video actually converted 3.2% better, but we have to use video because some other stakeholder, you know, is, is, is kind of tied to it. So I think that is definitely an important takeaway on that front.

Speaker B: Yeah, there's a lot going on there and I I like the, I mean right up front you say, you know, don't just use like AI driven, AI powered. I think like there's a, I don't know, uh, Brad Feld, one of the, the people in vc. I read a lot of what he writes and he had a recent post where like, and he actually it was just a conversation with uh, Chad gbt he was having around like is AI just software? Like you know, in the early days I imagine everybody's website said, you know, we are the software for CRM or the software for email or what. And it's like. And now it's just become like that AI is just in everything. Like so like AI driven really doesn't mean much. But some of the parallels in what you're seeing in like what um, is working around AI is similar to what we've seen. I, I've shared this anecdote before, you know, ah, data from our own platform in what was it, November of 22nd I think if that's right, uh, if I'm not getting my ears confused. When Gen AI first kind of became a thing that was publicly accessible through OpenAI and ChatGPT, we kept expecting to see on our platform like AI becoming one of the, if not the most popular topic in all of the millions of resources we have on our platform. And it just wasn't, you know, I remember sitting down with our product team and being like do we need to adjust algorithms? Do we need to adjust our product to show what is trending differently? And it was like no, like everything seems to be running. It just that content wasn't resonating. And then I remember it was March, you know, five months later, all of a sudden it was the fifth most popular topic. Third by April is our most popular topic. And what happened in that time frame is the content that we were kind of sourcing and adding to our platform shifted from what does AI mean for the future of B2B SaaS, what does AI mean for jobs? To how to use AI to build a better email cadence, how to use AI to accelerate the sales process, how to use AI to do XYZ. We just actually published the top 20 resources from uh, the first half of 2025 on our platform. This is like a week or two ago. About 40% of the resources, I think it was eight or nine pieces were AI thematically or about AI. Uh, but it was how to use AI in M partnerships, how to use AI in email cadences, how to use AI marketing workflow templates. Like it was less like thought leadership about AI and actually how to use AI to be better at your job if you're the reader of that piece of content. I just think that's. We as marketers get so enamored with being thought leaders and experts and making sure we're talking about the trending, coolest topics that sometimes we forget that what people really want to know is how to do their job more effectively or efficiently, uh, when they're learning about AI, uh, 100%.

Speaker A: I think that all resonates. And the other thing I would say is as much as we can avoid general benefits, I think is also really important. I was talking with someone about this difference between like, benefits versus capabilities, right? Like benefits being these general takeaways. We can save time, we can increase revenue, we can do whatever versus capabilities being something more. Similar to what I share with Zendesk, where It's like we provide 247 coverage on your support. Like, uh, that is a very specific thing that someone can relate to. And so I think it's tempting with AI to also get in that bubble. And I've seen that countless times where it's like, you can use this thing to like save money in your, or save time in your week or save whatever. And it's like, what would be better is for you to explain exactly what you are doing. That is saving people time, right? And to have that be the actual thing. So as much as we can move people towards that, again, mirroring your experience, I think they're going to get better outcomes.

Speaker B: Man, this is also fascinating. I could, we could have an entire follow up episode just on, uh, that topic. Maybe we will. But there's, uh, a bit of irony in this question. Kasey, I always ask, and even on our previous episode, we always ask all of our guests, what's an unpopular opinion that you have as it pertains to B2B marketing? But you might even have a unique lens on this question, given what do what works does. But, uh, I'll tee it up to you anyways and let you run with it.

Speaker A: I have a lot of unpopular takes, um, so it's tough to narrow them down. But one that has been top of mind for me recently is we've seen big changes in traffic, right? Like SEO is fundamentally changing. And as a byproduct, there's been this kind of narrative out there that everyone has been saying, like, traffic is dead, traffic is not important, traffic. You know, insert whatever type of thing as related to traffic. And from my own experience and from a lot of brands that I work with that are not heavily search based. Right. So if you're not getting a lot of your traffic from search, like, we drive like 30% of our pipeline from LinkedIn. We drive a bunch of pipeline from webinars, we drive them from podcasts, we drive them from news, like all these different, more kind of direct channels. In those cases, I actually see a very strong correlation between all the core parts of the funnel. When I have months where my traffic is dramatically higher, my leads are also higher, my demos are also higher, and my sales are also higher. So it's like connected to this thought that people say, like, is viral important? Like, do you have to go viral? And of course you don't have to go viral. But my experience when I wrote a post about clay logos, which I shared today, that got over 500,000 views, that was our highest demo week ever, most leads joined ever and most closed businesses in conjunction with a post ever. And that was because of a viral post. Because look, that's doing 10 20x, right? What any kind of normal post that I would put out would do. So this is not necessarily, this is kind of dependent on your business.

Speaker B: Right.

Speaker A: And can there be vanity metrics? Yes. But I would just caution people a little bit. I mean, this is probably good general advice for anything is dead type of things that get out there. But I do think that with a lot of these issues, there's a lot of nuance. Right. And you want to be really careful with that nuance. Like the same thing. We won't go down this rabbit hole for time. But with influencers, I spent a lot of time, 18 months, doing influencer work, deployed it across 19 different influencers. In general, B2B influencer deals are very hard to track and to make effective. Right. What you don't hear, and I do hear, not only from my own experience, but a lot of people I work with on the inside of these organizations is they see them as a boss, they don't renew. It is very, very hard to get good influencer deals that work, whether you're going macro or micro. So it's the same kind of thing. You'll hear like, everyone's like, oh, the new name of the game is B2B influencers. Can they work? Absolutely, they can work. We've had some phenomenal relationships that have been really effective. But just like with everything else, there's a lot of nuance. And if you jump into that and you just say, okay, everyone's saying influencers, this guy's got good follower count. 100,000 followers. We're going to fork over 5k. He's going to do this great post and then you get crickets. And part of that is your expectation. You maybe came in thinking, they're going to do this post. It's going to give me all this lead flow and whatever, whatever. And that wasn't really how things work in that context. From an outside party whose audience isn't, you know. Anyways, I digress. As we said before, we could have lots of sub conversations. But those are two. I know I, I did two instead of one, but those are two that I'd be aware of.

Speaker B: I love it. We, uh, we relied on influencers a lot in our early days. I would always caution people, people, it's volatile. You'll have one that's a home run and, and you'll have another one that's a complete whiff. And you just kind of have to accept both, uh, when doing influencer marketing, I know that's kind of the hazard of relying on that channel and I love your website visitor kind of example there. I. It's similar to a take I've had in the past where like people always like, oh, marketing vanity metrics like website visitors. And I'm like, website visitors is not a vanity metric. If you know the conversion rates beneath that. Right. How many people hit this page? How many people convert? Like, then it is absolutely like a critical metric. But I think sometimes people just like use it like, oh, visitors doesn't count. I'm kind of like. Or doesn't matter. I'm kind of like, that's always to me a signal that people maybe aren't doing the right optimization or kind of the conversion rate, uh, analysis that they should be beneath that. It's interesting for sure.

Speaker A: 100% agree.

Speaker B: Well Casey, if people want to learn more from you or learn more about Duo works, what's the best way to do that?

Speaker A: Yeah, so I'm super active on LinkedIn post pretty much every day. So if you're interested, I share tests, I share kind of unique things that websites are doing, all sorts of stuff like that on my LinkedIn. So Casey Hill, the one who works at do what Works and then if you guys go to do what Works IO, you can go to the website and if you're curious in seeing the platform and seeing your competitors, we can hop on a demo and we can show you your competitors. We also run a newsletter. So every Thursday we do these really in depth analysis is where we take lots of our learnings from tests and we kind of go through websites. We give you insights and data. We share actual A B tests. So if you want to just kind of consume more on that front, head to the website, opt in there. And we send that every Thursday.

Speaker B: So very cool. We will get all of those links in the show notes. I'm going to go update our website based on our conversation here today. Kasey, thanks so much for joining another episode of the Content Cocktail Hour. Until next time, same time, same place. Cheers. Thank you for joining the Content Cocktail Hour powered by Audience plus. If you want to see more episodes or more resources curated for your role, join us@, uh, audienceplus.com. see you next time. Same time, same place. Cheers.