As the founder of an SEO agency Distilled and a software product, SearchPilot, Will Critchlow has spent over 20 years bringing businesses more revenue through SEO.
Unlike influencers who proliferate buzzwords, Will tells us the hard truth: to increase revenue with SaaS SEO, you need to test everything patiently. And if you are a startup founder, there's a high chance SEO is not even for you.
In this episode script, we dive deep into the potential of SEO for startups and enterprises, separating fact from fiction in the SEO landscape, especially with the growing impact of AI.
Will’s journey in tech
Anastasiya: Will, why did you start in this domain? For me, even though I work in tech, SEO seems like a black box. I understand the expected results, but what happens inside? It feels like magic.
Will: Well, what we're doing now is trying to make it less magic and more scientific. The reason that I got into this area was, that I've always been interested in business generally. My dad ran his own company and I grew up around kind of small, small business, I guess. And my co-founder and I knew that we wanted to run a business together long before we knew what that business was going to be. And that's half the equation, I think. The other half is technology.
I was a teenager in the 90s. I remember the first time I had access to the internet, the first time that we got that at home, the real problems it was solving. My friend and I used to carry floppy disks into school every day to swap the work and the business ideas that we'd been doing back and forth. And when we could actually network our computers and send an email or send files directly, that was kind of revolutionary.
So I've always been a fan of technology, a user of technology, I guess. And where we started was just putting those things together, just realizing that this was in the late 90s, my co-founder and I started building websites. But then I think that was kind of even more on the technology side. And what pulled it back towards the business side was realizing that building the website wasn't enough. And that, if we wanted to help these, at the time they were small local businesses who were our customers, our clients, we needed to be able to help them get more visible, get more people finding their business online.
That led us to SEO because it was kind of the only game in town in 2005 when we were starting before the advent of Twitter, Facebook's newsfeed, and other forms of social media, leaving paid and organic search as the primary channels.
The impact of AI on search
Anastasiya: How do you see the market changing in the next few years? Predicting ten years ahead is challenging, but what about five? How will AI influence these changes?
Will: It's gonna be fun. I'm looking forward to living through that time. So a few interesting things.
Firstly, I don't think it's necessarily true that the rise of artificial intelligence agents or whatever we want to call them, our kind of personal assistants necessarily means that we will use Google less. It may well be that Google is a primary provider of these things. Probably my biggest prediction is that the rise of large language models and this kind of artificial intelligence agent chatbot experience, I think, in the geeky language of Clayton Christensen, this is a sustaining innovation rather than a disruptive one.
And I say this as a paying subscriber of OpenAI's ChatGPT, so I'm paying another provider for it at the moment, but I'm not convinced that this isn't just something that becomes part of Google and certainly part of a search engine in the broadest sense. So I don't think it's necessarily that much of a conflict.
Yuriy: Yeah, but we also see tools like Perplexity that try to compete with Google, so they do their own search. But it will also change perhaps that you will be doing agent search optimization. You will be optimizing not for people, but for chatbots, for copilots.
Will: Absolutely. Perplexity is great. I use that as well. And I think if you've literally yesterday, as we're recording this in it on the ninth of February is yesterday, Google announced the public availability of what it's now calling Gemini, right? What used to be Bard. And I had a little play with that. And I think that's very promising.
I don't think it's at all obvious that Google is technologically behind any of these other providers, what would make me wrong, I think, is if they can't figure out a way to make it compatible with their business model. Because the new upstarts, the OpenAI, and the perplexities and the like, don't have this constraint, which is a weird constraint of making hundreds of billions of dollars from search advertising. And that is a constraint on Google. Because if Google were suddenly to release something better for users, but cut off that stream of search advertising. Well, I mean, they're not gonna do that, right? They're gonna stop before they get to that point.
And so that's what would make it a disruptive innovation in the Christensen language. So I may be wrong, but that's part of it. I think actually that's my strongest prediction.
I think the one that's most relevant to my work and maybe to other folks who are operating in the SEO space is, I don't think websites are going away. If we're talking about the five-year, even 10-year horizon, I think websites will persist, websites still be a thing, visited by people as well as by smart agents, right? So yes, your robot might visit the website too, but I think you will also. And I think a lot of the prediction of the doom side of this looks a lot like the doom that we heard in 2007, 2008, when social media was on that kind of crazy up.
Do you remember when Twitter was just taking off and every day it was like, you know, this new person on Twitter, that new person on Twitter, this celebrity is doing this on Twitter, this news article broke on Twitter, whatever, all Twitter and Facebook newsfeed at a similar time. And there was a lot of prediction back then that that would, as social media grew, search volume would decline. And in fact, that's not what happened. They turned out to be complementary to one another. They both grew, and the search is up radically since then.
And I think human search is going to continue to grow as we get this explosion in, as you say, agent search or whatever you want to call it.
Yuriy: Yeah, so I think that industry, at least for a few years, will grow because now there will be a lot of work to optimize for Gemini and all the bots so they will find your content. And there are already some tricks like writing more frequently asked questions sections because they usually tend to reply to questions. This kind of content helps you to be first in their output, even if you are on the search page on Google. And also optimizing for agents that can click through websites and just book tables or order some stuff. People will not trust and send their bots to finish orders, but you will want the bot to conduct the booking and then bring the user to your site to finish something. So maybe it would not be so important, everything on the website, but bringing this agent to your website would be important. Oh, okay. What about the products in the industry? Do you think that there is still a place for new search engine optimization software to come up now because of this AI change? Because I see a lot of products that help to write content, but what about other software that helps to get on the ranking?
Will: Yeah, I think I mean, I kind of broadly subscribe to the Andreessen-Horowitz view of Software Is Eating the World. We talked before about the agency model. These days, I'm running a software business, and happy and excited to be doing that. I think the opportunity for software is already huge, but it is growing and will continue to grow.
So, the short answer is yes, there's tons of opportunity out there. What I think is interesting though is what opportunities are going to come in what order and how soon. And I think there's a lot of buzz right now about these search generative experiences, whatever you call it, the embedding of chat into search results.
I think that may turn out to be an evolutionary dead end. I'm very bullish on large language models generally and on artificial intelligence, but specifically, chat interfaces in the search results. I'm not sure because I think that conversation feels slow quite often right now and inefficient. And so I think there are probably some evolutionary dead ends that software is gonna go down saying, hey, here's a solution to tracking how many mentions you get in SGE for different kinds of query, you know, the equivalent of the kind of tools that we've got in keyword land. Some of those will turn out to be useful, but I think many will not because I think for all of the disruptive power of SGE, and we've seen this with OpenAI's chat GPT integration into Bing, for example, it's fractions of a percent of the total search volume right now. And so if you're a search professional, you have to be aware of these things and you have to understand what's coming down the track, but what's working right now is still going to be a lot of the kind of humans searching in relatively traditional human ways.
So, yes, there's tons of opportunity, but I also think there's gonna be a lot of froth, a lot of noise, a lot of stuff built that is not necessarily long-term sustainable. That's fine. That's just good capitalism. But I do think there's gonna be quite a lot of that kind of froth and churn.
SEO for startups
Anastasiya: Maybe we can now talk about the application of SEO on the products. Recently, we had a client who approached us with an idea. And one of his riskiest assumptions was relying on traffic. And since this was an early-stage startup and they only had an idea, it's really hard to understand if they need to incorporate and think about SEO at the very beginning, starting day one, or if this is something they can work on later on. And how to make these decisions, because obviously, you may need to invest a lot of resources into creating this architecture and like proper keywords and maybe a lot of content that should be in this website or product, while you need to validate the idea first. This is one thing. And another, it's competition, because this product was in the domain of real estate, and there is a lot of it already. And how can you beat up the existing proposition and at least have your offerings and top results? So what would you advise this founder about how they can think of SEO at the very beginning? What should they do?
Will: It's a great question. It's also a very big question. Let's see how much we can do justice to it.
So I would say, first of all, there are two different kinds of risk embedded in what you're talking about there. One kind of risk is there is definitely demand, people are searching, and we would be a great answer if only we could rank. And if we were to rank for those things, we would get valuable traffic that would convert and we would do great. And the risk is, can we appear for this highly competitive search term? That's one kind of risk.
There's a totally different kind of risk, which is we're building something that is different enough that it isn't exactly something people are searching for right now. And a lot of startups are in that second camp. And I would say in the second camp, the answer is simpler. The answer is SEO is not your first channel. You're not gonna validate your product demand by trying to kind of capture and pivot search demand.
You know, people who are searching for kind of related things, and then you say, hey, but actually we have this 10X better solution but it's not what you were searching for. That does not happen on day one. That doesn't happen in year one. That has to be built quite a lot slower over time. And you're gonna be looking more at channels like PR and social and you have to be putting the idea in front of people who hadn't even considered that they were looking for your thing. So in that case, a simple SEO is not channel one for you.
In the first case, I would actually say it's. Still, mainly SEO is not a great channel for validating startup ideas.
So if you're operating in that kind of lean startup mindset of saying, you know, I'm going to look at all of these assumptions, pick the riskiest one, and knock that assumption down first. You're going to be a long way down the track if you're saying, well, we need to rank in organic search and get the majority of our search traffic, the majority of our business through organic rankings. And until we do that, we can't do anything else, that's a very, very difficult bootstrap problem. It is solvable. We've seen startups break into very established markets.
I mean, you think about even when someone like Airbnb launched, those searches of places to stay in London or New York were incredibly competitive searches, even at the point when Airbnb launched, but they had such a radically different proposition that they were able to win over those search markets. But I would say they did that as a result. It wasn't like step one, we had to rank for places to stay in New York. And then we can build a business off the back of it. They had to kind of bootstrap their way there a little bit more.
So those opportunities are there, but I would say they're not there very often in a lean startup validate that as the first point area. And so we don't generally work at Search Pilot, we don't generally work with startups in their early stages, because for precisely that reason, you know, SEO is really the biggest channel for most established businesses online. But it's rarely where you get your first, you know, million visitors, or 100,000 visitors, or whatever it might be.
The tricky part, though, is if you are in one of those spaces like real estate, where organic search is clearly going to be your dominant channel in 10 years, you've got to start somewhere. And it's that kind of classic. I think it's a Chinese proverb that the best time to plant a tree was 50 years ago. But the second best time is now. And you kind of do need to start laying that groundwork for being the successful business you want to be, even though it may be a slow payoff. And that is very hard for a startup to prioritize.
So my advice in that situation, and I don't do this kind of consulting anymore, but my advice would be to get the foundational technical side of things right from the outset because it's broadly no more expensive to do it right than to do it wrong. So you might as well do it right. Get the website built in a good way, and get that technical foundation built the right way because there's no downside really. And then build out that, in many cases, you can build out that kind of content strategy and content development piece, again, in ways that are compatible with the long-term success you're seeking.
So you're gonna be building a lot of those landing pages for all kinds of reasons, they're going to end up being organic search landing pages. But maybe that's not the biggest source of traffic to them that maybe that's not the biggest channel in month one. But fairly quickly into many B2C startups' lifecycle, we see organic search overtaking and becoming the biggest channel at which point then yes, it's a flywheel, then you can really invest in it. But it's rarely going to be the way that you validate the hypothesis, I think.
Anastasiya: Yeah, because product managers sometimes rely on this fake door testing. They can launch the website and like to wait for people to come up and check how the conversions work, but they probably will need to rely on more advertising rather than organic traffic.
Will: Yep. I know I'm not a paid search expert, I used to be many years ago, but I'm not really these days. But I think that's a classic use case for paid search. If you know that the demand is there, people are doing the searching, and you need to validate that if they find your website, they're going to like it. That's what works perfectly even from day one, where in a way organic searches are too slow burn, even if it's in the long run better ROI, which hopefully is.
Common mistakes in SEO
Anastasiya: What are typical mistakes that companies sometimes make with their websites that when you come up, you see that everything is totally wrong and you need to fix everything?
Will: So these days we work with very successful websites. They're large websites with tons of visibility and tons of organic traffic, and we're helping them beat their competitors and get even further ahead.
So we don't tend to see fundamental mistakes so much these days. In my consulting years, we did, and you'd see everything from accidentally blocking the entire site from being crawled by Google to adding a staging publishing a staging site to the web and having no index, for example, and just literally saying to Google, we don't want to appear in your index.
Those kinds of things are easy to debug typically because traffic goes to basically zero.
What I would say we see more of these days is a disconnect between the realization that actual users are fun people, the actual humans who come on your website, notwithstanding the conversation we were having earlier about smart agents, is users use your website and they need to like it. And if, I think probably an underappreciated development in Google, particularly over the last five years, seven years is, that they care about those same things.
Google cares that visitors to your website like the experience. And they don't always, they're not always great at incorporating this into the algorithm, though they're improving at it. But I think as a direction of travel, as kind of saying, you know, what's going to be more true in five years than it was five years ago, I would put my money on saying, I think user experience as a broad ranking factor, not, I'm not talking about literally Google having a line in their algorithm saying, if user experience detail is good then, but more just the sites that rank are gonna have a better user experience and the ones that have a worse user experience are gonna perform worse.
So then you say, well, okay, fine, but every company wants to build a good user experience, don't they? I mean, anyway, but you realize actually there are a lot of terrible websites out there. So we see it with advertising just being overrunning the website, ad tech causing it to be slow to load, hard to use, and difficult and frustrating to actually find the content that you're looking for.
That's a classic that I see even on very large successful websites, site speed being part of that kind of thing. When we get down further into the details, it tends to be not major mistakes like that. Most of our experiments, most of the tests that we get to run are not as dramatic as saying. Hey, if we make your website radically better, does it perform better? We tend to be down in the tweaks. So our job these days is finding the small ways that you can continually improve.
Balancing UX and SEO
Yuriy: Well, how would you recommend finding a balance? For example, the website is very rich in media. They need to have a lot of images, and videos, and it objectively needs to load longer. And then Google says the speed is not so good. You are not green, you are yellow, for example. So would you prefer to always be in a green area and not have so much media or lower quality media? Or it's okay if you prefer to focus on the user experience.
Will: It's a balance. I mean, if ultimately, you focus too much on the user experience to the detriment of the channels that bring you those users, then you're not going to have any users to have an experience with. And so you do need to think holistically about this. And I think that the key is thinking holistically.
If genuinely it needs multimedia content, as you say, I'm aging, show my age by calling it multimedia content probably, but you know, you talk about CD-ROMs next that you do need that. Now, do you need to make your site slow loading on every page? And I would say in many cases, there are a lot of smart technical moves you can make so that only the user who wants to consume the video has to load the video, for example. Those kinds of sensible decisions along the way make sense.
The way we approach a lot of this kind of thinking, is we talk about full funnel testing, which is not something that I hear a great deal about in the industry. So we combine our core businesses and SEO testing. So testing whether a particular change is gonna result in more search visibility. But we can also run a conversion rate test.
We can also run user experience tests based on saying, you know, we'd like, we want this change to cause our users to be happier to do more of the things that we're hoping they're gonna do on the website. And when we combine those things, we call that a full funnel test. So we're saying, ‘Okay, we have a hypothesis, we're going to make this particular change to these landing pages. And we hope that that is going to both increase search visibility and improve user experience on those pages. And we can measure both those things independently and in conjunction and say, what's the net impact?’
And occasionally you do see an effect in different directions. Occasionally you do see that a change has a positive user experience impact, but a negative visibility impact, for example. And what we counsel our customers to do in those situations is not just to kind of throw your hands up and say, I guess we can't, you know, tough, we have to compromise then. And it's more to say that's a sign that you need to keep iterating. You need to keep looking for ways to gather the best of both worlds or at least mitigate the downsides. And so a classic example of this is we see it with eCommerce websites on what they call in the industry PLPs, right?
Product listing pages outside the industry, they often call them category pages or subcategory pages. And there's a lot of discussion and debate about how what should be on those pages beyond literally just a list of products. Do you want to describe the category? Do you want other imagery that isn't just the specific list of products that are on there? What kind of rich information might you want on there? And that's an area where we've seen some of these tests that go in different directions, where you might say, ‘Look, putting a paragraph of text at the top of the page means it ranks better, but it converts worse.’
Those are the kind of classic situations where you need to kind of iterate and say, well, is that true even if we improve the quality of the content? Does the ranking benefit stay even if we move that content further down the page? You can iterate through different avenues. And that's not to say they're always aligned, but I think, again, coming back to the direction of travel, we anticipate those things being more aligned in the future than they are in the past. So the more that we can get towards great user experience, the more that we can believe, if not trust, that's gonna help us in visibility terms over time as well.
A/B testing strategies for companies of all sizes
Yuriy: So if you're a huge corporation, perhaps you have a team that is iterating daily and trying to improve everything. If you're a smaller company and you still want to run A/B tests, how often should you do it? Is it like when you are doing major redesigns and you try to check if the new design is better than the previous one? Or like it's a quarterly approach or it's how would you recommend to work with that?
Will: I think it really comes down to the scale of your operation in terms of how big is the website right now. How many people are visiting it? How scaled up is that web operation?
The more traffic it's getting right now, the more opportunity and value there is in testing. The smaller you are, the earlier you are on that journey, and the more you have to take, you know, opinionated, hopefully intelligent guesses about what the market's going to want, because running the test would take too long, or maybe isn't even possible to reach statistical significance.
I can talk about this from my own experience. We relaunched our own searchpilot.com website last summer, and we couldn't test, we didn't test it at all. We had, well, we showed it, we did some kind of qualitative subjective user tests to say, showing it to some of our customers and prospects and saying, does this answer your questions? Does this work for you? But as a small B2B SaaS website, we don't have enough traffic to run our own kind of tests, nor really a conversion rate test. And we're actually doing some conversion rate work this year to iterate and improve. But we just had to take a fairly opinionated view that says, well, look, we think this is better than what we had before. We think there's value in pushing on.
Conversely, I think if you have reached that scale and, you know, we're talking quite often to not necessarily very early stage startups, but that kind of scale up stage. Many of them, in my opinion, should be investing much, much more in that team, that software that supports that testing program, that external support, whatever it might be, because I think this is a massively under-invested channel.
You know, I mentioned that organic search is still in 2024 the biggest channel, certainly for new customers, the biggest channel for the vast majority of websites that we see. And it's also the biggest growth channel for most of them, right? Most of the people are not number one in their market and the number one player is significantly bigger. And so there is a ton of growth opportunity and headroom. And yet it's the least understood channel.
You know, you think about how much data you have about your email marketing, about your paid search, about your social advertising and it's way more typically, and you understand, ‘Yes, I could spend more money over here and that will result in more business over there in ways that organic search, historically at least, and this is obviously what we're trying to change, but in ways that organic search has not done and has not been able to do.’
So yeah, my answer is kind of twofold, I guess. For a lot of folks, the answer is you can't really test those things and you've got to just build the best plan you can from comparative data and best practices and research. And then once you reach a certain size and scale, you should be testing as much as you can. And yes, that does mean investing more and building out that kind of testing culture and testing experimentation program, not just doing ad hoc when you have a major redesign, for example.
Yuriy: If I understand correctly, it's based on the number of daily visitors, perhaps around 1000 people a day on a page.
Will: That's when we say that it's possible to run an SEO test roughly at the kind of a thousand organic sessions per day to the site section. So not to an individual page necessarily, but to a group of pages that have a similar page structure, whether that's product pages or destination pages. Whether it's sensible to be testing is much more of a business model question. How valuable are those users? How much is an extra 10 % of traffic worth to you? Is it worth staffing out a team designed to hunt that down?
Because if you're a multi-billion dollar eCommerce retailer with a 50 % gross margin, you should stuff that team out, I think. That's something you should be investing heavily in. If you're a media publisher who makes a few dollars CPM and you have tens of thousands of visitors, then no, you can't afford to be stuffing out that kind of team and you're gonna have to grow the business first before you're able to really run those rigorous tests.
Yuriy: It's because you need somebody in your team who will do it. Yes?
Will: Yes, I mean, you can obviously outsource some of it, you can work with agencies, you can work with partners, work with software vendors. But all of those things come with costs. And whether it's your costs in your time, costs in staff costs, or stock costs with suppliers and vendors, then yeah, this is not cheap. But the opportunity is massive for so many of the folks with large websites and scalable sections that make them real money.
Yuriy: Should you ask your designers to learn about A/B testing? Should you ask your marketing manager to learn about A/B testing? Or it's a profession and there should be people who are focusing only on that probably a scale question as well.
Will: I personally think it's something that everybody should be interested in. Everybody who touches any of those areas should be curious, should be interested. You mentioned, in fact, this is part of your origin story. I think that part of the fun of design is making stuff that works, not just stuff that looks pretty.
So I think every designer, every product owner, every marketing manager, should want to be effective. Effectiveness comes from understanding an element of the data. And that doesn't mean you have to be a statistician, but understanding the concepts, at least. I think whether that's something, whether the experiments are run by those people, or whether they're run by dedicated data professionals, is really a question of scale and specialization. And, you know, are you able to justify having that as an individual, a team, a department? We see the whole gamut of those things.
But I wouldn't let the kind of statistical side of things put you off from running those experiments. I think even in a small company, folks can use, whether it's like Mailchimp's A -B testing on your email side of things or simple website experiments, you can do those things even if you're not a statistician.
And I think actually when you read about this on the internet, you see, in my opinion, too many statisticians who get too hung up on the fact that this is, it is a very complex area if you really want to understand it down to the kind of academic level. But in a small business, that's probably not the priority.
And we have a saying at Search Pilot, we talk a lot about how we're doing business, not science. And what I mean by that is if you're, you work for a pharmaceutical company and you've got a new heart attack, a drug that you're hoping is going to work, then of course, you need to run that to like, amazing levels of statistical confidence and scientific accuracy, and really rigorous statistical tests, because the cost of getting it wrong is so high.
In business, for the most part, this is not life and death, and actually, speed is important. And being directionally confident is more important than being absolutely certain every time. And so, we spend a lot of time saying our experimentation programs and our tests should inform your ability to move in broadly the right direction faster than your competitors. And if you can move with more confidence be right more often and move quicker, then that's how you can outperform the competition.
From agency to product: lessons for a successful transition
Anastasiya: So right now you're leading the team of specialists and you have a product that helps businesses to improve their SEO. So maybe we can talk about the transition from being a service company to launching the product and transitioning to a product company. So maybe you can tell, because it involves the change in the organization structure, hiring people. Like it's launching an entirely different business. So what was your journey here? Like, was it a very slow transition to this product and you first started on launching the POC and working on some concepts of the product as a kind of pet project with your team, or you straight away like launched it as a product idea and validated it and launched it to the market.
Will: Everything takes longer than you think. I think that's the, I forget who is, there's a law named after someone, isn't there, that everything takes longer than you think, even when you factored in the fact that everything takes longer than you think.
So we had a goal, a dream initially, or as an agency of being technology-enabled, technology-enhanced. And I think, I mean, think that's very common. I think many agencies look for this as a way of improving their margins, their client retention, and ultimately their valuation. And so we formed an R&D team, research, and development, just kind of looking for ways that we could be a better agency in the early days. And we've built a few things that helped at the margin. But the idea that would become SearchPilot was just one of those things. And it was a much bigger bet.
But we took an approach a little bit like you described with your client of trying to knock down the biggest questions one by one. And yes, that did take some time.
So we had a hack week initially of saying just literally technologically, can we build a thing that is fast enough and operates in the way that we want it to operate to build the eventual product, which knocked down the kind of technology hurdle. It then took a lot longer to knock down the market demand hurdle, because this is basically an entirely new sector, an entirely new area. And, you know, we went from first revenue a few years ago to it was still a few years before we spun SearchPilot out as an independent business.
So the transition basically went to an R&D project, first revenue, then we turned it into a business unit within the agency. So it had its own P&L. So the finances kept track of it as a group. It had its own staff allocated to it. And we put a general manager in charge of it. In fact, the guy who's now the COO of the whole business was the VP in charge of that kind of division. And we had a couple of years operating like that, growing it within the agency. And the vision kind of evolved over time. And then we had this opportunity that came along to spin it out and turn it into its own business, which was mainly enabled by the fact that we had this acquisition interest in the agency.
So Distilled, the agency was bought ultimately by a company called Brainlabs. And Brainlabs integrated Distilled into that and is doing very successfully. The Brainlabs SEO division is what has grown out of the old Distilled team. And the fact that they were basically interested in exactly the bit that we as founders were most ready to move on from. They wanted the agency and we wanted the software. And so there was an obvious deal to be done there.
And that's, it obviously took us a little while to navigate through it, but that's essentially where we ended up as they continue to run the agency and the evolution of that old distilled team. And we took the software business. They helped with some finance to get it up and running in that first year, which we've now paid back and bought out. So we're back to independent search pilots now an independent bootstrap standalone software business. And yes, the evolution goes R&D project, first revenue, business unit, then spin out. And then finally, independent, standing on its own two feet startup.
Differences between product and service culture
Anastasiya: And since you have experienced both worlds, what would you say are the biggest differences in this product culture and service culture? Like differences in mindset or maybe motivation of people working there?
Will: Yeah, well, so it's interesting actually that we've got a lot of the team that has made that transition. So we spun out with, I think it was 12 people from the old distilled team. That may not have included me, I may not be counting myself in that. Anyway, 12 to 14 people in that kind of range. And the majority of those people are still with Search Pilot. So a lot of our team has actually gone on that journey with us, which is excellent, superb, really. Proud and pleased with that.
We've also tried to, perhaps unusually for a software company, retain some elements of the service culture that I think often, and we hear this in our customer feedback, as we have a professional services team. We don't want to build an agency again, but we do want our customers to get that kind of very personal, very hands-on customer success experience. And so we've retained a lot of that, and I think that's actually a differentiator of Search Pilot, one of our competitive advantages.
The biggest changes are, there are some, two big changes to the business model that I think make a radical difference. One is the nature of what recurring revenue looks like and retention and churn and upsell and those kinds of areas, which, I think every agency tries to make their revenue as recurring as possible. But with the best will in the world, it's always less recurring than recurring software revenue. So I think that dynamic of operating in a world of metrics around churn and net revenue retention and so forth, we've had to really level up our understanding of the financial metrics that go around that.
And there's a surprising unlock of just the way that cash flow works. So, you know, an agency typically can be cash flow positive very early, but broadly speaking, collects the money as the work goes along, right? You bill monthly, typically. Software is not bought like that. I think probably a legacy of the fact that it used to be a physical, shrink-wrapped delivery of software, it's paid for in chunks. And so, we typically get to bill most of our customers on at least annual contracts, a number of them are on two or three-year contracts. And they typically pay annually in advance. And that unlocks a kind of funding model that is not available for agency growth.
And even though we've not taken outside investment, effectively our customers are investors and we have to treat them that way. And so that enables us to grow faster than we'd be able to otherwise. I think everything else is more evolutionary. So we're investing a lot at the moment in improving the way that we think about product development and ensuring that it is, ensuring that we're building the things that our customers most want and need and that we're building those things in the right way.
I think that's probably one of our biggest learning journeys is it's incredibly easy to get sucked into building things that don't actually help the customer. And they sound like a good idea, they look like a good idea, and nobody uses them. And that's, I think, that's our biggest current learning area where I think we're improving most rapidly in that area, which is quite exciting.
From agency deals to enterprise software success
Yuriy: What about B2B sales? I know that you moved to a segment of bigger corporations. So at the beginning, you started trying to serve different kinds of clients. Now you're focusing on a specific niche. Was it hard to focus on this kind of sales? Because in the agency business perhaps you had a different sales experience than selling products. Or it's very similar and it helped you.
Will: It's similar. Actually, I think it depends on who you're comparing to. But I think compared to many software startups, the agency sales model helped us specifically because even a small agency project is quite expensive software. You know, if you think about, I don't know, a 40, 50,000-pound dollar agency project, which is not that big, right? Even SearchPilot spent that order of magnitude on a new website, a new CRM.
We've done projects that size and we're tiny compared to our target market. When you think about the cost of software, I think a lot of software startups are scared of charging more than $100 a month or something. There's this idea that because the computer's doing it, it should be cheap somehow. And that was not a hang-up we ever had. We've made plenty of mistakes, but I think charging too little is not one of them.
And so our experience of selling five and six-figure agency deals helped with our experience of doing enterprise, well, it's quite an enterprise software sales, right? If you're talking about a six-figure software deal, that's quite an enterprise. We haven't yet done our first seven-figure one, but I hope that that… That comes along soon.
What I think is different is that there are a lot more stakeholders involved in a large software sale. So it's not the monetary size that is different, because we'd sold hundreds of thousands of consulting work. It's that when you buy an agency or consultancy, they just really deliver to their stakeholder. They just deliver to the person whose budget it is. If you're delivering software, especially with infrastructure software, right? So our software deploys into our customers' web stack.
That means all of a sudden, it's not just the SEO team, and the marketing team, you've got the product team, the engineering team, the infrastructure team, the InfoSec team, the legal team, the finance team, literally, it could be, you could have 10 stakeholders, each of whom can kill a deal, and all of whom have to be inside before you can close the deal. So sales cycles are longer, and deals fall apart in annoying ways at late stages. And I think we just have to continue to get better at that.
The model that I've been most excited about recently is, I recently read the book, The Challenger Customer, which is the follow-on to The Challenger Sale. And it talks about, how we as business owners and as technologists, we get very excited about talking about the way that things are gonna be great if you take our product, right? Here's how incredible Search Pilot is, here's how wonderfully your SEO testing program could run.
The hypothesis of the book, backed by a load of research, is actually it's more powerful to show them the pain that they're in right now. They actually need to understand better that it's really tough if you're not running SEO tests. If you're running a major website, with tons of traffic and you're not doing SEO testing or you're not doing it scientifically or you're not doing it with an easy-to-use product that means your SEO team can run those experiments, you're actually in a very difficult, dangerous, challenging position. And I'm on this learning journey, but that's something that I've been learning about recently and enjoying. Also highlighting the opportunities, like you are here and you can reach this level with improved process or additional service. Yeah?
Hiring and empowering talent
Yuriy: Yeah, for me, it sounds like a crime that you are stealing money that you could earn from your team and from your product. Yeah, because you're losing the opportunities. Nice. Regarding building the product, yeah, so you already had some technical staff, technical people who are working on the agency and then they switched to the product. What about other people, maybe some other hires, some expertise you had to find for your product? For example, designers, did you have designers on your team? If not, how did you build design expertise? In B2B sales, they say that maybe design is not the number one differentiator as of now, at least maybe a few years ago. Do you think that the user experience of your product helps you to differentiate on the market or it's mostly still about features?
Will: That's a great question. I think actually it's an area where we're hoping to see big strides in the future. And we probably haven't invested as much as we would have had to if we were building a B2C product or, because I think broadly speaking, that's right. That, especially enterprise products are bought for their capabilities more than they're bought for their user experience and their design. So we used to have designers in the team in the agency days.
We actually don't have a full-on dedicated design resource in the software company. Now, we work with outside freelancers and contract designers, and we obviously have some design capabilities, particularly in our front-end engineers. But we've been through a few cycles of working on specific projects with outside designers, our whole public-facing website was built by an agency. And so we've bought in those capabilities more as we've needed them.
I think actually the things that we've tended to hire, so we have recently been on a journey of hiring more experienced folks into our engineering and product team. So where I think as an agency, we were very good at taking newer, less experienced, early-career folks and giving them the time and space to develop and become phenomenal contributors. And that's still quite a big thing in our professional services team, because we really know how to develop an SEO consultant, right? We've been doing that for decades.
What we've been finding recently is that actually there's a big, it's very powerful to hire engineers in particular, who can hit the ground running, know what they're doing, have done this before. And I would say our last few hires, front end, back end, DevOps infrastructure, have all been more experienced folks who have been senior somewhere else. And that is a new thing, actually, that's a new area of learning, as well as that, I think, with the way that with software being higher margin, higher retention, etc. If we can bring somebody in who can be materially better, the business model can afford to hire those more senior, more experienced folks a bit more. So I suspect that trend is going to continue. And as we add design and UX into the team. No doubt we'll go looking for folks who've been there and done that.
Yuriy: Yes, that's interesting because part of the value proposition of agency for their teammates is helping them to grow. But for a product, usually, you don't have so much time to develop your team. Your main business is building the product. And in this case, you would want to focus mostly on people who already know how to do stuff and be more competitive than other teams.
Will: Yes, we talked earlier about the pros and cons of the agency model and the software model for the owner or the founder, the senior team. I think it's also true for employees. I think agencies are wonderful places to get a start and they're great places to accelerate your early career. And then some folks find they love that work and they want to become senior contributors. But I think for a lot of people, a career path that makes sense is agency early on and then moving either in-house or product or wherever as they gain more experience.
Yuriy: Or create your own product. The best case scenario and my dream is that most of the people who work at Cieden in the end will fund their own product or become co-founders at least.
Will: So I mean, for those who have that dream, I don't think that's for everyone, but definitely. And when I look back over the years, naturally, it's why we built it right into the mission and vision of Search Pilot. So we have our mission, which is what the product is seeking to do essentially, which is to prove the value of SEO for the world's biggest websites. But we have a second part of our purpose, which is the vision of creating, and proving that you can create a great place for folks to work from diverse backgrounds within a high-performance technology business. And we made that a kind of top-level thing. That's not in our values, that's actually in our purpose.
Because looking back over what was at the time when we sold it, 15 years of Distilled, everything else comes and goes. But the things that you really remember are those little times when you made a difference to a human being, to a person's life. You gave them an opportunity, hired them, developed them, and gave them a platform to grow from. And I, yeah, I mean, those are all the stories that you look back on and remember, whether they go on to do their own thing, whether they stick with you and you're still working with them decades later, or they go and they become a CMO somewhere, you know, there are many different paths and all of those can be really exciting and really fulfilling.
Featurism: website personalization
Yuriy: We have a section called Futurism about vision and ideas. What are your thoughts on the future of AI-generated website content? Imagine creating a site that uses AI to personalize content based on user demographics, weather, and time. For example, a young mother from Berlin looking for a stroller could receive tailored recommendations on parks, transport options, and relevant products. This would involve dynamic, user-specific pages that adapt in real-time. However, could this level of personalization create challenges for Google in understanding and indexing such content? Do you think websites will evolve into highly personalized applications in the future?
Will: I'm sure some will. I think my zoomed-out prediction is these are the kinds of predictions that are hardest to make. If you look at early predictions of what the internet or mobile phones would be used for, the details rarely match the imagined version. Some of these capabilities will likely emerge, but whether they will be used in those specific ways is uncertain.
I'm fascinated by the life cycle of ideas, especially considering that the current generation of large language models is trained primarily on human-generated content. The next generation will have AI-generated content in its training set, which could lead to an inflection point. There's a risk of a "garbage in, garbage out" scenario if the models aren't good enough but still get used as training data, creating a dangerous feedback loop.
Personalization is significant, and many of our eCommerce customers already leverage it. However, people are often less different than we assume. Within a large enough cohort, like new mothers in the UK, personalization might not be as impactful. While some degree of customization will exist, the differences may be minimal for things like reading news or shopping for a laptop.
Reflecting on the future of search, changes may occur slower than anticipated. While new startups will explore these advanced personalization technologies, the typical eCommerce experience—browsing, adding to cart, and checking out—will likely remain common in the next five years. I'm excited by the technology and its potential, but I don't foresee an overnight transformation.
Anastasiya: Yeah, there is some adoption of technology. Like there are always some early adopters, but the majority is still lagging. So yeah.
Will: Of course, yeah, I mean, totally the way and I think some of that is accidental, some of it's deliberate. Some people want to continue doing it the old way. Some people are scared of the new way. All of these things are fine. And personally, I'm quite excited to use the new stuff, but I also realize that I'm not everyone and I'm not necessarily typical.
Anastasiya: Yeah, thank you very much for your amazing insights and your perspective of the future, and your expertise on SEO and how this specialty works with the product team, how it can improve the product. So we were really happy having you today.
Will: It was a really fun conversation. Thank you very much for hosting and for a really insightful set of questions. I think it's been, we've gone down quite a few exciting and interesting different avenues there and it's been really fun.
That's all for today. Thanks for reading. If you think your colleague or friend could use these tips, go ahead and share this episode with them. Cheers!
