Split testing is typically a term bandied about in the CRO world, not the SEO world. In this episode, you’re going to learn how to split test SEO. Running tests in parallel is not something that’s easily done with SEO. There are plenty of reasons for that. Each page has its own link profile, making it hard to do head-to-head comparisons. Not to mention, you don’t want to create duplicate content, so the copy on the pages need to be different too. Nevertheless, there are workarounds, which you are about to learn about. My guest is Will Critchlow. He’s the CEO of Distilled, a company he founded in 2005 with Duncan Morris. Distilled provides online marketing services from offices in London, New York and Seattle, hosts the SearchLove Conference series in the US and UK, produces the popular online training platform, DistilledU, and runs the SEO testing platform, DistilledODN.
Will, welcome to the show.
It’s great to be here. Thanks for having me.
Let’s talk about some of the cool stuff that you’ve created in these years. You’re a little different than your standard SEO agency. You have Software as a Service. You have your own conference that’s seen as one of the industry conferences to go to. Maybe we start there with what’s different about Distilled versus a regular agency that specializes in SEO?
Some of this stuff feels like chance and serendipity but it all ties back to the ways that we wanted to set out to be different. We didn’t set out to be different in having a conference or building out a product or anything else, but we wanted to connect to the business side, connect to effective work and connect to accountable work. Back when we got started, we felt a relative lack of that. I know that you felt some of this over the years as well. We wanted to do a lot of our own research. We wanted to do a lot of our own discovery of what worked. It’s some of those underlying values that led us to where we ended up, even though the journey might have seemed serendipitous and points along the way. In 2009 in London, we ran a conference in partnership with Moz or SEOMoz as they were back then. Their conference was called MozCon. It felt like an extension of the writing that we did and speaking at other people’s conferences. We thought it seemed fun and we sold out a couple hundred-person venue and all of a sudden, we had a business unit. We retrofitted that story a little bit to try and say, “How does all of this hang together?”
We subsequently launched DistilledU, our online training platform. We saw all of that whole strategy, everything from the blog, the writing, speaking at other people’s conferences, our own conference and that training platform. We saw all of that into this cohesive whole of over a long time period of five, seven-plus years. We felt that if we were discovering what worked, teaching and training others in what worked, that would differentiate us in the long run. We’re seeing that come to fruition finally. It’s super hard to differentiate a people-based business, a professional services business. Over the years, we’ve been very lucky to have some superbly talented colleagues and teammates who have built their own reputations. What we’ve always tried to do is encourage and support that while having it burnished the Distilled reputation along the way. Often even many years later when they’ve gone on to even bigger and more exciting things on their own behalf. That led through all that. That was the early 2010s opening up in conferences in the US as well to go with our offices in the US.
That was our point of differentiation. Our position in the market and the opportunity to get the word out about what we believed in and what we thought was a good way of operating. A few years ago, 2014, 2015, we were looking for the next step. It was a few years since we started running all that conferences and we still do. It’s up and running. We have these three offices and we are looking for, “How do we solidify that and take the next step?” We formed a small R&D thing. This was around the time that my co-founder, Duncan, stepped down from the CEO role. I took the CEO role years previously. We’d both been in all kinds of different executive positions within the business. He’d run the business for five years and he actually went back to his roots on the product side and worked in engineering. He was literally writing the software code along with one of the senior consultants in London, Tom Anthony. Tom had pitched the idea of an R&D team, Duncan was an engineer in this team. We get a surprising amount of free reign to cast around and do exciting things. The thing that ended up coming out of that was this SaaS platform that you mentioned, the Optimization Delivery Network. That was the drive for that quest for discovery, that curiosity we had of trying to figure out what could be a big thing, what fits with our values and purpose, and what could differentiate us and help us stand out in that market?
When I had my agency, Netconcepts, I had developed a platform as well. It was more serendipity and chance or luck. The universe is looking out for us. We didn’t have an R&D team. We didn’t have any structure around inventing anything. I had a client situation where it was extremely frustrating. That client situation was Kohl’s Department Stores in the US. I couldn’t convince them that we weren’t going to destroy their site with SEO. I thought, “Why don’t I use a proxy to pull in their pages real-time, add some SEO to those pages? I’ll show them what the SEO-ed version of their website will look like before I even have access to anything.” We never ended up getting access to their servers. We did build a lot of comfort with that. A few months later, I’m talking to another client, Northern Tool, and realizing that we’re not going to hit the deadline with this code freeze for the holidays. We need to rewrite the URLs. We need to add all this SEO stuff but they’re on an old version of IBM HTTP server that doesn’t support the regular rewrite rules. It does support proxy-based rewrite rules.
“Why not apply the thing I had just played with and created a few months earlier as proxy-based optimization platform, which is just for demonstration purposes earlier that year and use it in a production environment?” We did. That was the beginning in 2003 of our own SEO CDN-type solution, Gravity Stream. There was a lot of serendipity there. The reason why we got acquired back in 2010 was of the technology platform. It generated over half of our revenue. It’s pretty exciting and I never would have imagined an agency that would mostly be a Software as a Service company but there you go. Did you have any serendipitous moments like that or some huge opportunity with the platform or a new service offering or something presented itself like that?
I remember watching it from afar through that era of you being so far ahead of the game with that technology back then. I know many people looked at it in a similar era but you had it live working in production. There was a deliberate phase of ours where we knew that there were some things we wanted to do around this angle. It fits so well with our core purpose. If I remember correctly, yours was deployed into the racks, you were actually putting physical servers.We want to connect to the business side, connect to effective work and connect to accountable work. Click To Tweet
We had our own servers but we had them add some proxy-based rewrite rules to their HTTPD.conf.
Part of the question for us was whether we could build this scale and get it performing in the cloud. Obviously, that was something that enabled us to scale quickly and scale up to serving the primary traffic of some pretty substantial websites, even very early in the product and engineering side of our development. There was a bunch of deliberate engineering working on that side. Probably the big serendipity stuff came earlier. It was the partnership with Moz, which dates back to those that era when Rand and the team at SEOMoz were still consulting before they’d raised the venture capital. Initially, it was just us hanging out as everybody did back then pre-Twitter in blog and building a name for ourselves, answering questions and helping people out. That turned into referrals of work and in particular European work, UK and Continental Europe. It was a couple of those big breakthroughs early on where we got great results on some early introductions that Rand made to some European companies. That word got back and we are a safe pair of hands, we could be trusted and we’re doing good work. Everything else flowed from there. That was the reason we ended up in the US.
It’s pretty amazing what Rand and Moz can do to a company to transform it and take it to another level just from introductions and joint project. They certainly have transformed my life and my business. The book that Rand and I started together that ended up becoming The Art of SEO was hugely transformative to every aspect of my business, career and life. I still get tons of referrals from the Moz recommended list, which you’re on as well. It’s pretty amazing what Rand has done for others as well as for himself and for the core company that he had founded. Of course, he’s moved on to SparkToro. How did this all come about? If you’re chiming at the time as SEOMoz’ blog being helpful and so forth, did he contact you? Did you guys meet up at a conference or something? What was that magic moment?
The very first magic moment, it’s what felt like bravery. It was a funny thing to say looking back on it. It was that moment of, “What do we have to lose?” Rebecca Kelley, she used to work at Moz. She was making a trip to the UK and she was traveling with her boyfriend. She was going to go work at Fresh Egg. I forget some of the details but she was doing a learning exchange. She was traveling through London and she’d never been to the UK before. We’ve never met in person or interacted with any of the SEOMoz team. We just dropped her an email and said, “We’d love to show you London if you have any time in your schedule.” She did and we all get on really well, a couple of our team and two of them. At that time, we were tiny, only a handful of people on our team and we were just getting started. They seemed like rock stars and even looking back, she was a member of staff. It felt like some crazy thing to reach out and say, “We’re going to be in the same city. Do you want to catch up?” It seems funny to talk but in those terms, but I think certainly for anyone who is at that stage in their career, I found that these people are not as inaccessible as you think. You can reach out to pretty much anyone and these days it’s actually often easy. You can chat with people on Twitter, you can send them a direct message. Many people will metaphorically take that call. It’s surprisingly doable.
Let me just jump in and say that I’ve heard that one connection, one introduction, one relationship can cut off five years of a hard slog from your career path. It’s like a shortcut. It’s having a wormhole right into five, ten years into your career.
It’s the future. I don’t know how many people are at this stage in their career. At that point this was 2005, 2006, we’d done nothing in business, this is our first company. We were learning this whole thing as we went along. We have never even managed any people in our previous jobs and previous experience. We used to go to networking events. I remember at that time there was a big trend of business speed introductions, like speed dating but for people to meet other entrepreneurs. I met very few people through those things that I’m so in touch with. You have this fake view of networking like networking is a different thing you do. Whereas the thing that’s been valuable to me is making industry friends, people who work in the same space. Understand the same things, are passionate about the same things, excited about the same things and see the world through a similar lens to me. There are many things that I disagree with some of these guys about. You can have those exciting and interesting conversations just like you would with your real-world friends if you’d like. Those relationships have been so valuable over the years. They’ve been a source of business, they’ve been sources of inspiration, support, advice and so much more.
To share one of my magic moments with Rand specifically, I had never spoken to him before and we were on a panel together. It was the first SMX Advanced Conference, the Give It Up Panel. This was an idea that I had given to Danny Sullivan, “You should have a panel where we share our best secrets and let’s make sure that those secrets stay in the room.” It became an embargo instead of a code of silence. It’s 30 days then all the secrets are out but I still shared my best stuff that I had figured out during my own R&D. Rand was really impressed apparently because a few weeks later, I see him in the speaker lounge at SCS Toronto. He comes up to me and gives me a big hug. I had never spoken to him before and here he comes with his yellow shoes, gives me a huge hug and tells me, “You brought it. That was incredible at the Give It Up Panel, that was amazing.” We had a great conversation. In that conversation, we decided to do a book together with O’Reilly as our publisher.
While in that conversation, we sent a message to Danny. He happened to be at the Food Camp Conference, the friend of O’Reilly conference. We went and asked some friends there, “O’Reilly, would you like Rand and Stephan as authors on an SEO book?” They’re like, “Yes, of course.” Within a couple of days, we had our publisher. If you would have told me a decade ago that O’Reilly would be my publisher, I would have been, “No way.” I was so enamored by O’Reilly since the early ‘90s. I learned how to program and everything from O’Reilly books. I have three books with O’Reilly and I have Rand to thank for it and that hug. What happened after you hosted Rebecca, you showed her around the city? What happened next and to turn this into an amazing partnership?
That emboldened us and with this newfound bravery, we reached out. The next time Rand was in London, he was on a trip to SCS London. He was traveling with Gillian, his mom, and a couple other folks. We invite them to dinner. We went to a restaurant called Boisdale, which is near Victoria in London. It’s a Scottish themed restaurant. It’s game, haggis, whiskey and so forth. Gillian still talks about the soup that she had there. They loved it. It was this authentic experience that I never found anything like that in the States. They wouldn’t even let us pay for dinner. We invited them but we ended splitting with the check. We made friends. We realized that we had thought the same way about so many of the knotty challenges that clients faced. That’s always been pretty decisive on these things once he’s figured out that you know your stuff. They started sending us referrals very quickly after that. It was one of those early projects we were working for a company called Sofa.com, which was an introduction for Rand. He passed it to us because it was specifically international, which was an area that we knew a lot more about and worked in more languages.
Europe always has that international flavor. It was this specific problem and the founder of Sofa.com, Pat Reeves who sadly died far too young a couple years ago, Pat had emailed Rand and said they bought this domain, Sofa.com. They’d spent a fortune on it and it ranked number one for the query sofa if you search from the US. They bought it from an American company and it was nowhere to be found if you searched in the UK. This turned out to be our testing ground. We eventually solved this problem. This is in an era where Google didn’t acknowledge that international existed and had all kinds of weird funny bugs in it. Eventually, we changed it. It worked and overnight you could see the traffic switch from US-centric to UK-centric and they never looked back. That company did phenomenally well. Pat himself was one of the biggest sources of business for us. He referred us to the dozens of people. More importantly, he thanked Rand and said he couldn’t recommend us highly enough. That was the beginning of the partnership. Rand referred us to the UK and European business. We had a couple of big projects for big logo clients and then also we landed a retainer around that point. That was more money that we’d ever seen in our life. The rest is then history. As they moved out of consulting and focused on their software side, we ended up opening up a Seattle office. Rob from our team went out there to open that office and he’s still there. We invited Rebecca in London and then taking Rand, family and friends out for dinner.Making industry friends, people who work in the same space, is valuable. Click To Tweet
You guys are still in Seattle, that’s where Moz is located. Are you still working closely with Moz even though Rand has left?
We know a bunch of the folks over there. We don’t get business from them in the same way we did in the early days. When we first set up in Seattle, we had a landing page on their website that they were no longer offering. They obviously get consulting inquiries they’re well-known for not offering that anymore. We still get the personal inquiries and interactions via the members of the team. We do bits of product feedback. We still blog over there and answer Q&A and do a bunch of other stuff. It’s ongoing but I don’t think it’s as critical to either business as it was in 2011.
Let’s start talking about this platform that you’ve developed, DistilledODN. How are you able to split test SEO, which is not normally something that you can do? You have to typically run these tests in serial instead of in parallel. You can’t say, “I’m going to test two different versions of my title tag on my home page in parallel. You only have one home page.” How does that work?
You hit the nail on the head of the key difference between this and the kind of split testing that people might be used to. If you’ve run conversion like optimization tests, for example, using a platform like Optimizely, you’ve been able to do things like split test changes on your homepage because you are splitting the traffic. You’re showing one version to half your audience and a different version to the other half of your audience. The way our platform works is it’s targeting larger websites. The way it works is instead of splitting the audience, we split the pages. We’re testing kinds of change rather than specific title tags. For example, if we’re running on an eCommerce site and they might have a couple of thousand products, you come up with a hypothesis that a particular change to the product page layout is going to be beneficial from an SEO perspective. The way our platform works is that it applies that change to some percentage of those product pages. The easiest way to imagine it is making a change to half the product pages and leaving half the product pages unchanged. We apply a similar set of mathematical techniques to those that you would use to evaluate, a test to see if the pages that we’ve changed are outperforming relative to what you’d expect given the performance of the pages that we haven’t changed.
Instead of splitting the audience, you’re splitting the pages. The critically important stuff is that these changes are completely transparent in the sense that users and search sliders get the exact same experience. Googlebot is being served the exact same thing that users are being served, which are some pages updates and some pages not. That side of it is relatively simple to understand though it’s hard to implement in practice generally without a platform like ours. Most CMS, for example, won’t let you arbitrarily change the structure of half your product pages. Part of what our platform does is enable us to make those changes and then the other part is the mathematical analysis. It’s some statistical techniques plus some neural network machine learning stuff to evaluate the benefit of the change and come up with a chart that can say, “This change resulted in this change of traffic.”
You don’t want to run afoul of Google’s guidelines, which specifically referring to here the fact you don’t want to cloak. It’s serving a different version of a page or set of pages to Googlebot than you serve to the users. You’re completely white hat in that regard. You’re also doing all this analysis and you talked about briefly machine learning. How are you incorporating machine learning or AI into your algorithms? How will you be incorporating that stuff in the future into the platform?
The biggest challenge is that the control in variant pages don’t necessarily perform identically. With conversion optimization, if you’ve got a big enough audience, you can assume that the conversion rate of your control group is statistically similar to the conversion rate of your variant group until you make the change. The challenge with trafficked pages is that pages are more different to one another than users are from each other. Your highest traffic page isn’t two or three times as visited as your least trafficked page. It might be 10,000 times of traffic. The techniques need to be a little more sophisticated. In particular, rather than comparing the traffic to the control pages directly to the traffic and to the variant pages, what we need to do instead is we use the control group to make a forecast of what would have happened to the traffic, the variant pages had we done nothing. What we actually end up doing is comparing the actual variant performance to the forecast variant performance. We use the control group to make a forecast incorporate seasonality, site-wide changes and competitor action with Google algorithm updates, all of these external factors that might mean that the traffic would have changed. All of that is captured by the control group and factored into the forecast. That’s the background.
We started out building this forecast using a technique or an algorithm piece of software called CausalImpact. If you’re into statistics, it’s a stochastic time series model that lets you forecast into the future based on historical data and control group data. Google actually developed some of this stuff. They use it for evaluating paid search, click-through rates. That’s the basis of all of our model. If you meet online, you’ll see the people at Airbnb and Pinterest have talked about similar approaches. What we’ve been evaluating are neural network-based algorithms that do a very similar thing. They take into account the historical data, the data from the control group and this neural network outputs forecast of what would have happened had we made no change. We can get a confidence interval on whether the actual real observed data is statistically significantly better or worse than expected, and hence whether the change has been beneficial or not.
For those who are not that savvy about statistics, could you define what’s stochastic means?
It’s a fancy word that means random is probably the closest normal language stuff. I would recommend not going down the rabbit holes if you don’t have a background on it unless you’re particularly interested in it. The important thing is that what we’re doing here is controlling for all of these outside inputs. We can detect if an uplift is caused by seasonality, volume changing, site-wide changes, algorithm updates or competitor actions or any of these things. What we’re doing is isolating the change essentially. We’re saying that the only thing that we’re looking at here is the uplift caused by the hypothesis that we had, the change that we made to these pages, and we’re putting confidence info on that. What we’re saying is that there’s only a very small chance that what we’re looking at here is random, that these pages just went up on their own. A much greater likelihood that this is down to the change that we made. In the last decade or so in consulting, it’s been generally very hard to be able to get that management-friendly data chart that literally goes up and says, “This much of the performance is down to that specific technical change that was made to the site on that date.”We need to ensure that we keep our costs under control and all the rest of it. Click To Tweet
Is your pricing usage-based? Is it performance-based? How do you price?
We’re a small company that isn’t venture-backed. We need to ensure that we keep our costs under control and all the rest of it. The way that we’re working is that we take a very close look at the websites that we’re going to be pairing. We put together a fixed price based on that. It’s a monthly SaaS-type fee. It is based on a variety of things that essentially affect that costs. It’s the traffic and the bandwidth that is going to be used. We work with teams to put pricing together. We are aiming at the enterprise. We’re talking thousands a month for the platform. What we find is this requires a reasonably large website to work on and decent levels of traffic and all rest of it, that the places where it’s effective are the places where it’s worth putting our time into designing great tests and what’s worth having the budget for this software platform to evaluate them. We are not going to taunt into that part of the market but we’ve had some really great results with these and it’s a ton of fun.
One thing that was very innovative at the time for us was to create our pay-per-performance pricing model for Gravity Stream for my proxy-based platform. It was based on a cost-per-click. This is whatever you’d end up spending on pay per click, maybe it’ll be $0.50 or whatever, “That’s a bargain I want to buy as much of that organic traffic as possible through Gravity Stream and not spend it over on AdWords.” Have you considered some performance pricing model?
We haven’t deployed one. We don’t have any partnerships like that but it’s definitely something that we’ll consider in the future.
If somebody cannot afford or it doesn’t make sense for them to use this platform because they don’t have enough traffic or they don’t have enough scale in terms of pages of their site, what do you recommend as far as testing SEO for a smaller site or a smaller budget type of company?
There are a few different angles. One angle is user testing. It’s increasingly true that Google, in particular, their technology is matching their rhetoric. For years they’ve said they’re all about giving value to users and showing good results to users. It’s increasingly true that you can improve your search performance by improving the attractiveness of your website to your users. That’s not 100% true. There are plenty of things you can do for users that screw up your site performance. When we’re talking about within the field of sensible things you might change for an SEO benefit, you can test those in a more CRM way. That’s one angle. If you don’t have the traffic for that, you can start with more qualitative research so interviews and understanding what your users want. Understanding how they’re using, hearing those anecdotes from users, which can be in person or it can be electronically. You can email some of your users or you could put up a pop-up survey on the website. There are many different ways of implementing this.
Distilled’s own website isn’t particularly amenable to SEO testing. We’re not an eCommerce platform or a large media site or whatever. One of the impactful things we did was actually on our conference pages, we had a little exit survey that said, “What information were you looking for? Did you find what you’re looking for? Was there anything missing?” We realized we had essentially only put the conference schedule, the times or different sessions and those things, we would only publish that once we had speakers confirmed. What we discovered from the survey was actually our users wanted to know that way earlier in the process. There were people who were happy booking a ticket before they knew who all the speakers were if they knew what kind of topics are going to get covered. They knew what time of day, everything was going to start and ended. They just needed to know the logistical information. In our heads, that had always been coupled to the editorial calendar, who’s speaking when.
It was insightful to us to realize that actually our intent’s slightly different to the way we are producing this and it was valuable to say, “There will be eight sessions starting at [9:15]. It will end at [10:00] AM,” and publish all that information, even we didn’t have all the speakers confirmed yet. That changed our process a little bit. That came out of the qualitative use of research. I give a bunch of talks these days about how best practice isn’t enough, which is to say that the more Black-Box that Google gets and the more artificial intelligence, the machine learning is at their end, the more that we need to be testing. You can’t just assume anymore that things work the way they always have or something will generalize or apply across industries.
You can’t test your own stuff, it’s still better than nothing to be relying on the output of other people’s tests or what I call molecular laboratory tests. You can still run your own tests to say, “Does X, Y or Z work in a laboratory environment, a test domain or whatever else?” You can also start with the news, we publish information about what tests that work and what did not work for many of our customers. You can apply those things. I would encourage to say at that scale, step back a little bit. The place where these on-page tweaks and technical side tweaks work best is at scale. The biggest uplift from those kinds of things come when you’re already getting this huge amount of organic traffic. You are already performing very well across tens of thousands of pages on your site. If you are running a fifteen, twenty-page website, it’s very unlikely the biggest benefit is going to come from making one small tweak to one of those pages or even ten of those pages. You’re much more likely to get incremental gains from publishing new content on topics that people search for and are interested in. That comes out of the whole market research, keyword research side of things and looking for gaps in the market.
It’s publishing exceptional content that can really stand apart from what anybody else has talked about in that area. There are plenty of downsides to social media. One of the upsides is you can come from nowhere. If you publish the stuff that is genuinely better than anything that’s out there right now, you can gain that visibility and you can gain authority and people will link to it. We’re finding in general many large dominant-type sites that an extra few links don’t generally move the needle. In highly competitive verticals, there are a lot of other factors that are as important in the mix. If you’re just starting out and you’re running a small website, in particular, you’re running a website that maybe hasn’t been around for decades. There’s a very good chance you need more authority. Publishing some of that get big ticket stuff that really gets you on the radar of influencers and gets you some of that link opt-in and general authority is much more to move the needle.Relationships are so valuable. They are a source of business, of inspiration, support, advice, and so much more. Click To Tweet
Is there a platform that you would recommend for smaller folks to do some of these user tests? For example, there’s UsabilityHub.com, there’s UserTesting.com. Do you have a favorite that you would recommend?
I’ve found that end of the market has a whole bunch of things that work plenty well enough. There are free tools these days, the Google Optimizer element to things if you’re going to user-centric split testing. I also really like some of the qualitative feedback somewhere in between things like Five Second Test where you’re just getting focused on specific pieces of feedback and the narrative that overlays that. If you’re a small business, I wouldn’t go over on spending money on this stuff. You can do a ton of it with a Google phone and email some of your customers and contacts.
Just to do a pop-up, which can be done very inexpensively too using OptinMonster or whatever. If somebody wants to take it to another level and do some testing or some focus groups but in a facilitated way that they don’t have to reinvent the wheel. Have you tried out UsabilityHub or UserTesting or some of those other platforms?
I know some of the team have but I haven’t actually sat in on one of those. Have you done that yourself? I’m very curious about the other person’s experience of it.
I’ve heard good things about UsabilityHub, and something that’s not specifically UserTesting for usability or user experience but more just conversion rate optimization. There is Optimizely and there is VWO, Visual Website Optimizer. I’ve heard really good things about both but in particular, Optimizely has been raved about by clients and industry people that I know. Is that your experience?
Yes, it is. Although I know that they’ve been focusing on more on the enterprise end of the market. I think their pricing has become a little inaccessible for small to midsize companies. We just wanted to do the Google conversion testing tool, which is less fully-featured but free.
It used to be called Google Website Optimizer and then it became part of Google Analytics, Content Experiments as what they called inside Google Analytics. Then they rebranded again to Google Optimizer.
It’s probably part of a 360Suite somewhere in there, there is a paid version. I don’t know. I feel like sometimes Google’s branding on this stuff is getting more Microsoft like with now.
Let’s move on to the Distilled University or DistilledU and how that fits into the model. From my experience at least, I have six online courses and I don’t attract the same type of user there for my courses than I do for my consulting. That’s a problem if I’m trying to use the online courses as a lead generator or as part of my funnel to get consulting clients for SEO engagements. Are you finding that to be a similar situation where you’re not attracting the ideal target audience for consulting or for agency work versus basically educating your competitors or other agencies?
We certainly do our fair share of educating our competitors for sure. However, the reasons it fits into our overarching strategy are because of a couple of different elements. A big part of the reason that we built it was trying our team. By making it public facing and paid for, it was both a motivator, an excuse and a reason to keep it updated and invest in it. I have a read a blog post at some point about the history of the initial pitch that I made to our team that we should build this thing. A big part of it was we should be building the best resources as a knowledge capture game. We do give it away to a bunch of our competitors, but there’s also a lot of attrition. People move jobs a lot in our industry and I think at every tech space.
On the lead gen side, it’s much more a really long game. People who get that start learning from DistilledU, maybe we’ve got the agency side learning from DistilledU, then end up in-house somewhere, then get more senior, then maybe buy more seat ticket. Then one day, they have a consulting need or a need for an agency and Distilled is high up their list of choices. Five, six years in, we’ve seen that play out. We’ve seen people at the major household name and multinational brands who are now running SEO budgets to spend who say, “I got my start learning from DistilledU.” It works sometimes but I’m not sure it’s easy to build an ROI model around that. It is lead gen for SearchLove. About half our SearchLove audience is technically our competitors. It’s people coming from other agencies, other consultancies, and DistilledU is a lead gen to that. Also, we use it the other way around. We bundle seats to access DistilledU into our consulting work. We are educating our client teams on what we do and what we’re up to, why we are doing it and helping them build and grow that capabilities.Publishing exceptional content can make you really stand apart from what anybody else has talked about in that area. Click To Tweet
You definitely have the long game in mind and that’s awesome. It’s very difficult to implement long-game strategies when you have investors, in particular, venture capital because you have to keep meeting your numbers every quarter. Have you thought about at some point you might take on venture capital or you’re just like, “No, that’s going to change where we focus in a negative way enough that it’s not worth it?”
I think we’re pretty clear-eyed about the pros and cons. We’ve got enough friends now. A decade ago, I didn’t know anything about this. I didn’t know the benefits of taking the money, the limits of doing it bootstrapped, I didn’t know the downsides of taking the money and all those things you just listed. This was the way we knew, this was the way we could figure out. We didn’t have access to capital, and so we got started and we bootstrapped. Now, I think we’re pretty clear-eyed about it. We have enough friends and acquaintances and we’ve read a ton.
I think there could come a time where we felt that our personal objectives are aligned with what an investor would be looking for, which is that swing for the fences or nothing outcome I think in a classic venture capital world. I think that would probably only happen in a case where we as founders already felt like we’d taken some of the money off the table or done well from the original thing. We’ve been stocking away this thing for thirteen years now, bootstrapping and building it up. It’s done very well for us and I feel very lucky to be in a position where I would now bet this company on the tossed coin saying, “Maybe it lands heads with $100 million. Maybe it lands tails and we’ve thrown it all away.” We’d have to align the incentives a little bit. I can see that happening one day. I could see a more different kind of investment. There are some investors who come along who are not necessarily looking for that venture capital type, return-the-fund investment, but are looking for more of doubling the money where we can be much more aligned. Never say never. We’re funding what we’re doing out of cashflow operations. Running part of the business, that’s the game we’re playing.
Where does the SearchLove Conference fit in? Is this a big part of your business or is it just one of many add-ons? Their conference is their thing. I’m thinking, for example, so much of what Content Marketing Institute had created is around their Annual Conference Content Marketing World. They have a magazine and they have online courses, they have a blog and everything. Content Marketing World is their crown jewel and that’s where I would imagine the bulk of their revenue came from. Now, of course, they’ve been sold to UBM. I’m curious, is that a huge part of the company? Is SearchLove that big or an industry conference that is an add-on to a big agency?
It’s a little bit more than 10% of our revenue. It’s a relatively small part in the financials. It’s a much bigger impact on our brand and positioning. I do wonder what would happen with more focus. That is one area where I could imagine something more focused and bundled up the conference asset, the training asset and some of the publishing would be a fascinating business in its own right. I’m sure it could be substantially bigger. Focus is not something I’m great at. I think a lot of people who start businesses probably have these issues. I’m always off looking at the next shiny thing. This is an area where I’m trying to stay disciplined in saying that we’re focusing on the work that we’re doing for our clients and customers through our client work and our software work. The conferences support that and a crucial part of it, but we are staying focused on that objective right now.
I’ve got an advice regarding my Shiny Object Syndrome that I read The ONE Thing. Have you heard of that book?
I’ve heard it but I haven’t read it. I think I probably should.
The other book that I’ve gotten recommended to me is Essentialism.
I’ve heard of it but I haven’t actually read it. I think it comes from curiosity. What I like is learning and what drives me is discovery. That’s why I need to surround myself with people who are great operators because I’m not great at doing the same thing day in and day out, week in, week out, month in, month out. I find, for example, other kinds of people who like having a thing they do every month, having a thing they do every week, having a thing they do every quarter. That’s not what drives me. I need them to do that. What I love are learning and discovery. When it manifests itself positively, that looks like figuring out how to do something that we need and then delegating and handing it off and making it a cool part to somebody else’s job. I worked for a good chunk on learning how to do enterprise software sales. That was something we’ve not organizationally really done before, selling a bunch of big deals. Then one of my colleagues, Craig, who now runs that the ODN Team and is leading the sales into that. He owed me better than I was at that, but he’s building off some of that background work that was that I did as well.
Back to SearchLove, what happened next in terms of the planning for the next conference? Do you have a conference organizer? Do you have a team that decides who the speakers are or do you just fill the agenda with your own team as the speakers? How does that work?
There will always be some people from my company onstage but it’s definitely not all our team. We’re pretty rigorous in that selection process. We have one dedicated team member. Lindsay has been in that role for years now. She’s the kind of person who won’t have lost count. She will know exactly how many conferences she’s run over the years and I don’t. Lindsay runs over the programming and logistics of the events. That part is with me and other members of the senior team who help with the specialist industry knowledge and speaker evaluation sides. She will do a ton of research and shortlisting and so forth. Then we will help with the selection in terms of making sure they have rounded out editorial view, the set of talks holds together well. We’ve covered the topics we want to cover and so forth. Evaluating speaker quality and capability. Typically between me and maybe one or two other senior team members, we will do speak prep calls with every single speaker who speaks at our event, which I’m amazed more events don’t do. I’ve barely ever had that happen to me when I’m speaking at a third-party event. We try and do it for everybody who speaks fast and help them understand what our audience is expecting and tips and tweaks and polishing up to help them bring their A game.
That is very unusual. The only conference who’s gone over my presentation beforehand, I can only think of NEMOA.
It’s super rare but I’m amazed because I know it helps me when I run my presentation through this feedback internally. We’ve seen the benefit it makes to our speakers and the quality of talks that we got at our events. It’s work, but it’s a super high return on effort is what we found.
You have a SearchLove Conference coming up in San Diego. When is that?
That will be in the spring. We’re moving locations. Previously, we’ve been a little bit further away from downtown. We’re still in this beautiful resort-type setting. If you come from any of the places where we have offices, if you’re in Seattle, New York or London, then this is a fantastic time to visit Southern California. I think that’s true for many parts of the states as well. It will be on the March 4th and 5th 2019. We’re at the Kona Kai Resort, which is a hotel resort. It will be just SearchLove, which is fantastic. We’re looking forward to that. It’s always one of our destination events where I think people bring their families and stay for the weekend and whatever else. We bring the same level of rigor to the preparation and the same level of talks. Also, it’s just a great location to be in and a fantastic time to be visiting.
If our audience wanted to work with your agency, Distilled, if they wanted to sign up for SearchLove, if they wanted to sign up for your DistilledU online training, where would we send them to?
Distilled.net for everything. Hopefully, we’ve done our job pretty self-explanatory from there. Distilled.net/events for the conferences which are in San Diego in March and Boston in June and London in October. Distilled.net/U for the training and university platform. Distilled.net/odn for the Optimization Delivery Network. They can chat with me on Twitter, @WillCritchlow. I can always put people in the right direction and answer any other questions that they may have.
Thank you, Will. It was a great pleasure to have you on the show and a lot of fun. I’m sure our audience will now be able to do something with this knowledge like maybe run some tests.
Your Checklist of Actions to Take
☑ Take that first step and reach out to experts. One way to grow is by surrounding myself with likeminded people who I can collaborate with.
☑ Build good working relationships. Aim to have mutual respect and open communication.
☑ Be willing to try different techniques in testing the data. Will say to never assume anymore that things work the way they always have.
☑ Improve my site performance by understanding what my users want. Send email or create a pop-up survey to get valuable inputs from my clients.
☑ Stay abreast of the latest news and information on SEO split testing. Will publishes this information on their website at DistilledODN.
☑ Slowly but surely generate new content that stimulates people’s interest. Do market and keyword research and identify gaps.
☑ Don’t underestimate the power of social media. This is where I can gain visibility and authority by publishing genuine and outstanding content especially if I’m just starting out.
☑ Make use of different testing platforms such as UsabilityHub.com and UserTesting.com. Will mentions the Five Second Test if you’re getting focused on specific pieces of feedback and the narrative that overlays that.
☑ Don’t overspend on testing platforms but maximize Google phone and email to reach out to my clients.
☑ Attend Will’s SearchLove Conference on March 4th & 5th, 2019 and learn from the industry leaders on how I can uplevel my online marketing.
About Will Critchlow
Will Critchlow is CEO of Distilled – a company he founded in 2005 with Duncan Morris. Distilled provides online marketing services from offices in London, New York, and Seattle, hosts the SearchLove conference series in the US and UK, produces the popular online training platform DistilledU, and runs the SEO split-testing platform DistilledODN.