Welcome to Episode 59 of iPullRank’s Rankable Podcast, where we discuss various hot topics in the world of SEO and digital marketing.
On Ep.59 of iPullRank‘s Rankable Podcast, Garrett Sussman hosted Will Critchlow, CEO of SearchPilot to discuss the topic “The Proper Way to Do SEO Testing”
Will Critchlow started the SEO consultancy, Distilled with Duncan Morris in 2005. They’ve been at the forefront of SEO split-testing methodologies and spun out the Optimization Delivery Network (ODN) platform SearchPilot (started in 2015).
Will’s a master at conducting split-test experiments, that can determine how changes to your website impact your rankings.
This week, Will Critchlow joined us to discuss the state of SEO split testing.
We also discussed:
- Different types of experiments you can conduct with SEO split testing
- What types of skills do you need to conduct an experiment
- The complexities of internal linking and why it’s essential to SEO (Check out Will’s recent post on the topic: How To Test Internal Links for SEO)
- Building a Diversity and Inclusion program (Check out SearchPilot’s Annual Diversity and Inclusion Transparency Report)
Transcript:
Garrett Sussman: [00:00:00] Okay, welcome. Hi everybody. Welcome to the Rankable podcast. I’m super excited. We are getting into the nitty-gritty of some complex, awesome SEO. My name is Garrett Sussman with iPullRank. Today I’m joined by Will Critchlow. Will is the master when it comes to SEO split testing, he has some really awesome thoughts about internal linking.
He’s the CEO of SearchPilot, which is actually the spinoff of his agency. Distilled which started back in, what was it? 2005, with Duncan Morris, they built it up and it was a very successful agency. And then, so SearchPilot was this software platform that allows you to do SEO testing. It uses some really complex mathematical algorithms to be able to do these accurate experiments.
So he’s going to tell us all about it. Thank you so much for joining me today, Will. Awesome.
Will Critchlow: Thanks for having me on, Garrett. It’s great to be here.
Garrett Sussman: So before we even dive in, you did share a fun fact with me and you educated me too about the land of whiskey, which first off, I didn’t even realize that the scotch whiskey spelled without an E American whiskey and UK whiskey with the EY. Tell me a little bit about your whiskey story. You had 200 bottles.
Will Critchlow: So we bought a cask. This is despite the name Distilled of the agency. This was nothing to do with the agency business. It was with Duncan, my co-founder. So, it was just the two of us and his dad, but it was an idea that I had when we were walking, my wife and I were out walking the West Highland Way.
I don’t know if anybody who’s listening knows this. It’s like a 90-mile walk that you go over multiple over like a week, you stay at camp in different places or stay in small places on the way. And one of the walks, one of the days is the, it’s like this the longest of the days.
And it’s across Rannoch Moor which if anybody’s seen Skyfall, the Bond movie, it’s that kind of Scottish, like, you’re crazy beautiful, but really forbidding landscape. We’re walking right the way across this moor and, uh, it’s, the rain is just driving in our faces and we were cold. We didn’t want stuff for lunch.
It is only like an 18 mile walk day or something. And, uh, anyway, it’s like mid-afternoon. We’d been walking in the rain all day and I just was like, I just really wish somebody would just show up right now and have a whisky because wouldn’t that be great. Anyway, I don’t know thought processes being what they are.
One thing led to another, and that led me to the idea of like, Hey, no whisky gets really expensive. It gets more and more expensive. I bet it’s a great investment. Why don’t we buy a cask of whisky? And we got home and a whisky and 15 years later ended up with 200 and something bottles. And, you know, it was, it was a fun journey.
Garrett Sussman: So how about, how much is it? What is it? Is it wasn’t worth the investment?
Will Critchlow: So we did make money. What we did with it was we actually kind of pre-sold it along the way. So people buying it mid-price, I guess like between the price that we were paying when we bought the original cask and what we figured would be worth a full price.
And we auctioned in some bottles and, I think some of them went for a reasonable amount of money, and more importantly, I ended up with a few dozen bottles. I’m in a lifetime supply. 15-year-old cask strength, Arran whisky.
Garrett Sussman: That’s the world of SEOs, right? It’s like, we all have ideas. We all buy domain names with always the good intentions and some pan out. And some really just never do like how, man, how’d you get into SEO? How’d you get into specifically SEO split testing? What was that journey?
Will Critchlow: Well, so that journey, it’s got a whole bunch of different bits to it. So we started out, it’s not. it was design and development, I guess, back in the nineties. Duncan and I went to high school together. We played basketball at age 11. So what we call High School, which actually before I think US High School starts and, in the late nineties, we would probably have 16 and 17.
We were casting around for businesses that we could start that were more interesting than, waiting tables or whatever, other kinds of…
Garrett Sussman: Buying minidiscs for your minidisc player?
Will Critchlow: Yeah, there’s another story there, but, we were catching up and stuff and we ended up building some basic HTML and building some small business websites for local companies.
And then when we finished university, we thought about going straight into it, but we wanted to move to London. Couldn’t afford to do both things at once. So we went and got jobs in the city and two and a half years later, we said, “Hey, you know? We’re going to start a business together.”
Cause we should do it now while we’re renting before we have kids, you know, all that kind of life stuff. So we did it. We’d been casting about on a big idea, you know, that anybody who was around in that .com era remembers the big idea. You know, that it’s like actually most of my businesses now.
Right. Eventually, but it just took so long. It’s like it’s pet food, but it’s delivered. And in 1996, that was a terrible idea. Of course. Now it’s a multi-billion dollar, uh, multi multi-billion dollar.
Garrett Sussman: ECommerce in the nineties, like was not yet a thing. You were scared to put your credit card into the computer.
Will Critchlow: Right. I mean, so much, like the fulfillment wasn’t there anyway. So one day, we thought we needed this big idea to start a business. And then we were like, we want stuck on me anyway, even though we don’t have the big idea, I think it went the.com bubble burst by this point. So we went back to like, why don’t we do something profitable?
You know, crazy talk. And so we started. We just went back to the web design development stuff. By this point, Duncan had got a computer science degree. So, all of the actual hard work fell to him and my job was getting out there, selling it and doing everything else, running the business.
Right? All the finance administration, legal stuff, all that kind thing. That was 2005. We spent a year or two building, smaller, small business websites. We built our own CMS back when WordPress wasn’t that great for, you know, running anything but a blog. We built this super simple, super stripped-down kind of thing.
Anyway, after having built dozens of these small business websites, we realized that what everybody really wanted, the differentiated thing was getting more traffic to them because, you could literally, in the part in Southwest London we lived, you could throw a stone and hit a web developer. But there weren’t many people specializing in digital marketing.
And so that then ended up combining, especially in a way I was almost first interested in paid search, because back then there wasn’t really anything else, right? There was no social to speak of. And so it’s really all search and either paid or organic… And I loved the auction theory of paid search because I’d actually studied auction theory at university. That’s a whole other rabbit hole that we can go down.
And if you want to properly get me geeking out for an hour and this thing would, nobody would be listening by the end of it.
Garrett Sussman: We’ll save that for our NFT conversation next.
Will Critchlow: So I was really into the business model behind hiding that side of things, but for small local business clients, SEO was where it was at. Right. And so we initially, it was just a case of helping them get visible, making sure that they were found for the obvious things. And then gradually we realized there was a lot more to it than that. A lot of opportunity in that.
And after practicing essentially for free for, you know, like we were just doing it to help out our website class initially. And then we gradually started getting more into it. I got involved with the Moz community, SEOMoz, as it was called back then. Eventually started selling as an established service, which was around the time my brother got involved with businesses and things as well.
And the rest was history. It basically was perfect for me to geek out. It had the right combination of business impact, technical detail. Stuff that was changing. The stuff that we can research data and analysis, and all those things were things that I’m still excited about 15 years later.
So that’s how we got it.
Garrett Sussman: No, and it’s fun. And so it’s interesting because you did, you did sell Distilled to Brain Labs, but now you put your focus really on this, this SEO testing. And obviously you still dabble in everything, but like, What is the draw with SEO testing? When did SEO testing really come to the forefront?
And how do you perceive it? Like in the modern SEO world?
Will Critchlow: Yes. So the history is that we… the platform that became SearchPilot, as you mentioned at the outset, it started out as a project within Distilled. So, in 2014, 2015, somewhere around that kind of time, a few folks in our team formed an R and D team or research and development team within this building.
And actually, Duncan, my co-founder, and I were part of that team. Some other names people might know, so Tom Anthony led the drive for a lot of this stuff. That initial group of essentially engineers, set out to start building things that helped us differentiate the agency may make us better, make our client work better to help our clients be successful.
Help our consultants do better work and those kinds of things. And they, they’ve built a few different went in a few different directions, exploring different things. And by around 2016 or 2017, we knew that they were just focused on the one, which was, this, this platform was what became SearchPilot.
And it started as being technically an R and D question of, could we build a [00:10:00] platform that deployed like a proxy layer, right? Like, like a CDN, right? Like on-time delivery network into our entire customer’s web stack. Could it be fast performance, secure, all those kinds of things and modify the HTML.
So essentially a CDN that can modify the contents of the page. That was the first thing, the kind of technical R and D. And then it was kind of saying, okay, well, what we could do with that, cause either we could, we could just make changes, right. We’ve just fix SEO fixes. But also at the same time, a few of us that it came from a few different directions at once.
But I remember a conversation with, Tom Anthony about how folks like Etsy and Pinterest. I forget who was doing right back then, but there were, there were some articles published. This concept of SEO split testing, which is, you know, many people have been written or talked about user experience testing, right?
Conversion rate optimization. Classic A/B testing that people are familiar with. When you, you make two versions of a page and you show half your audience, one version, and half the other version. I mean, a lot, a lot of subtleties to it, but [00:11:00] that’s the basics. This approach that they were pioneering at some of the big tech companies was a little different and it was about saying, okay, there’s only really in this context, one user, we care about is Googlebot.
And what we care about the end-users, but there’s one user we can measure. And we can’t show Googlebot two versions of the same page. That won’t work. Doesn’t work, technically. And even if we could, somehow there’d be duplication. And so what the way it works on a large website, you take a site section, make a change and you’re testing, really template changes.
And so what you do is you pick half the pages of the templates that make the change and the other half the pages in the site section and leave them unchanged as the control group. And this was something, that some of the big tech companies were doing, and it’s quite complex and in particular, there’s a bunch of stages that are difficult.
So you need to, you can’t just randomly allocate those pages. You have to separate them into two statistically similar groups.
Garrett Sussman: Right.
Will Critchlow: That’s difficult. Then you’ve got to make changes to half the pages in the site section, which most content management systems don’t let you do. Then you’ve got to ingest all the analytics data and [00:12:00] do a bunch of statistical work to figure out did your change work.
Did, did it have an impact, but, but that’s the core of it. And we realized we could do this with this platform. And, so we set out building it basically and tried to validate the, um, the, the business idea that this was something that the people needed. And you asked how it fits into SEO.
Our vision was this could solve the two biggest problems that we had encountered as consultants over the preceding decade, which were, that, all of our clients struggled with getting things done. Right? So even if they knew what to do, getting it actually implemented out into the real world on their website was super tough.
And proving value and being able to know which things worked, which ones didn’t, which ones you had to roll back, what the right thing to do was, and those just blew away every other challenge that we had, those were the challenges. And we could solve both of those problems for our customers if they deployed this stuff.
And so, yeah, first I think first paying customer 2017 ish kind of era. We’ve turned that group [00:13:00] that had just been techies, the kind of R and D group into a business unit in 2018. And that business unit is what it had. So P and L and that kind of stuff. And that was what we spun out into its own independent company at the same time as the Brain Labs deal.
When, when the rest of the company was acquired by. Brain Labs.
Garrett Sussman: Right. Oh, it’s such a major achievement, cause you just listed out all the technical challenges in the first place, let alone, you’re working with all the variability of Google that’s so typically outside of the control of SEO is like, you know, you still see like anytime there’s an algorithm update or there are new competitors entering the field, you know, that can really impact the results that you’re seeing.
You’re tied with your forecast model and the mathematic algorithms that your team has built. You account for all of these variables that, cause you have to be able to say with some amount of statistical significance with confidence that these experiments are coming to some sort of conclusion and you still have to always be like adjusting it as well.
So people can’t just do that [00:14:00] themselves. Your tool is very unique in that respect, right? Yeah.
Will Critchlow: So, so all those things are complicated. And we figured if we, if we could just build it once and put a lot of our energy and effort into making that a scalable platform, then that could be available to a wide range of customers and all of those improvements that we would put through the platform.
Yeah, available to all of them, the classic kind of SaaS business model. And, yeah, because there’s a control group because there are pages that we haven’t changed, we get statistical confidence. We can measure the uplift and we can kind of rule out all of the other possible causes of those changes, whether it be seasonality site-wide changes, Google algorithm, updates, things your competitors have done.
And all of those things are taken care of by the control group. So we can narrow it down to the SEO changes that you made.
Garrett Sussman: So you mentioned Etsy and you mentioned Pinterest, and we know with CRO, there’s always, you need a certain amount of traffic to be able to even conduct an experiment that gets the results that can be statistically significant, [00:15:00] who is SEO split testing for like who can actually use this and implement this to get results that are reliable.
Will Critchlow: Yeah. So we focused, on the bigger end, and that’s quite deliberate for a whole variety of reasons, but, so most of our, our mission is the really big websites. The rule of thumb with the kind of at the bottom end of, of where it becomes statistically valid, I guess, is when you would likely to be able to get statistically valid results is somewhere in the region of a thousand organic visits per session per day to the site section in question.
So excluding the homepage typically, right? So, the kinds of sites, this works for. It could be e-commerce. It could be travel, pre-COVID. It could be jobs, real estate, local businesses with enough, if they’ve got enough location, like a national business with locate or retail outlets , any of those kinds of websites that are scalable, that have a lot of number of pages in the same template.
And as long as that group of pages gets that kind of a thousand organic sessions a day kind of level, then you can run statistical significance with tests on it.
Garrett Sussman: Cool. Okay, so you have your type of websites that you can test this on. What are the type of experiments that you run?
I mean, I can imagine there are different levels of complexity of experiments. What, what do you usually recommend? Where do you start?
Will Critchlow: Yeah. I mean, obviously, it depends on what they’ve already worked on and what their, what their kind of priorities are, but it does run the full gamut from super simple stuff, like a title tag change, to I guess mid-level complexity will be something like a purely on page change, but more difficult, like implementing schema, for example, is something that is typically, maybe a little bit trickier from an engineering perspective or to build the test.
And then probably the hardest test, which is some stuff that I’ve been writing about recently is internal linking tests. And the reason that that’s difficult is mainly on the measurement side because the challenge with internal linking tests of course is a link is from it’s between two pages.
And so with a schema change, you make the change to a page and you expect to see the results in performance [00:17:00] of that page. Whereas with an internal linking test, you can see changes to the source page, change to the destination page, or even changes to pages that are neither of those things that are just now their links are a little bit more diluted because you’ve added some more links to the page.
Um, so it’s, it’s kind of complicated.
Garrett Sussman: I’m actually really excited. So that’s a perfect segue into internal linking, which you actually just wrote an article that came out yesterday where you kind of got some more of your thoughts out there. You’d written an article back in, in 2017 about it. And it’s a really complex subject.
I mean, you’ve even said yourself, there are still things that you feel like still confuse you about internal linking. Where do you currently stand on it? Like how important is internal linking when it comes to impacting rankings and SEO?,
Will Critchlow: I mean, I think it’s, I would say it’s probably been one of our most consistently positive test types.
It is it consistently moves the needle. I would say it’s [00:18:00] probably not… if you ask me, what’s the single cleaned of test that has had the biggest impact. That’s more likely to be something like titles. And that’s just because they affect both ranking and click-through rate. And we’ve seen that we’ve seen nothing that messes a site up faster than getting titles wrong.
That’s the, there’s a whole other avenue of stories in that direction, but some of the most experienced SEO consultants I know have come up with SEO titles that, yeah, a hypothesis about titles, they thought was strong and it’s been like minus 30% organic traffic or something. So a title of the biggest, biggest impact, but internal linking is consistently positive.
And so I think, um, it’s, it’s still a big deal and it’s a big deal because large websites, it affects not only affects crawling, right? It’s the crawl path. It’s the ability to. re-crawl. and re-index long-tail pages more easily, more quickly distributed that kind of authority. It’s also, it’s [00:19:00] also an authority metric that you know, we don’t know what’s going on inside Google’s black box, but it does appear that some kind of concept of a PageRank-esque thing is still in the mix.
And that they’re never going to say that only applies to external links. Internal links are always a thing in that model. And so it for sure distributes authority around the site and can lead to ways of shaping that and pushing your most important pages. more strongly to the forefront.
Garrett Sussman: It would scare the crap out of me, to be honest, even mess with internal links the way you talk about it, because to your point, there are the three factors of how, what can be, you know, impacted the source, the destination, and then just ones that aren’t related, but in the ecosystem.
Yeah. Like how do you, and obviously Page Rank, you know, differs between, you know, the homepage is going to have much more impact than everything else. How do you approach SEO split-testing internal linking? That seems like complexity on top, on top of [00:20:00] complexity, but it seems really important before you go and make any changes because you can’t anticipate what’s going to happen.
How do you approach that?
Will Critchlow: I mean, in terms of designing the test. You can approach it from a few different angles. I think you can, you can there are rules of thumb and best practices. That you could apply. Obviously, you don’t know if they’re going to be right, but reducing the number of clicks away from the homepage of longer tail pages, better internally interlinking between pages at the same level.
There’s a lot of kind of general rules and a couple of hypotheses you might come up with in that kind of way. Or you can get more scientific about it. And that could be just running a crawl and finding know unloved sections or the opposite finding, looking more through analytics and ranking data and trying to say which are our biggest pages, which are our biggest opportunity pages. So, you know, stuff ranking bottom, half the first page, that, that kind of stuff. And just looking at ways to distribute link authority around to those places or even further. And this is some stuff that I’ve, you mentioned that the thing I wrote in 2017, we dabbled with this, honestly, it’s not, it’s still more kind of R & D than anything else, but around, running your own restrictive algorithms.
So running a kind of internal PageRank kind of concept on your own, on a crawl of your own data. And then I guess potentially in that rate. So, so rolling out a theoretical change, and then measuring the impact of that on the internal PageRank side, before you push it out to the, you know, the real world, one more notes on that, I seen quite a few articles about internal PageRank that just run page rank on the, on a raw crawl. And I think that can get misleading results if you actually do that. And it’s not too hard to do so you can run a call using anything from Screaming Frog to a cloud-based crawler and get a list of all your internal links or even on a pretty large site you can do a sampling of that. And then Python, R, you know, a bunch of programs just have built in the ability to run PageRank. The problem is if you’ve just run a naive Page Rank on an internal crawl data, it’ll throw out things like, you know, your most important pages, your terms and conditions. [00:22:00] And that’s because it’s linked to every page on your website.
But, so that’s misleading, it doesn’t give you kind of the real picture. So what I think you need to do, if you’re actually going to go down this route is seed the internal PageRank with some kind of external authority metric data. And, you know, that could be something like Ahrefs data, Moz data, or even just tracking data.
It’s just something that gives some concept of which are your big pages. And then, if you feed that in, you can kind of pre-weight, a Page Rank algorithm with, like starting values and then run the iterative algorithm. And I think that gives better results. Like I said, this is a little bit experimental, so you can do those sorts of things.
And then ultimately, at some point, you’re gonna want to put it out to the real world and you going to want to test the, was that a good idea? It was all that work I did worth it. And this is where you get into, as you said, all those kinds of complexities. So the way we think about that is I mentioned splitting a site section to control and the variables before.
What you need to do is not only do that but then also come up with what we call a measurement section, which is the group of pages where you might expect to see the result. So that’s in this case, the destinations of the new links that have been put in and then compare that to a lookalike control, which is like a group of pages that are similar to those pages, but didn’t get any links.
And so now you’re measuring the impact on the pages where you placed the links, the impact on the pages that received the links, both of them control against pages that haven’t had those changes put in that’s as far as we’ve gone so far, but we have an idea for going to the next stage, which has done another measurement group, which is a control group of pages that are outside of the set of pages that could have been linked to, but that’ll link to from the same page, same source pages.
This is getting super hard to explain, but essentially looking for the dilution effects, right. Looking for the possible negative effects on other page types. From the fact that you’ve added new pages into new links into certain templates.
Garrett Sussman: No. I mean, it’s fascinating. And it makes me think, especially for the enterprise brands that [00:24:00] are running these types of experiments, like for SearchPilot like, do you have a, like, almost like a concierge service of people who like run the experience in-house or who is the type of person at the enterprise?
What role would they have that could competently run these types of complex experiments? Like, do they need to be a technical SEO with a background and significant mathematics and statistics? Or is it, can it be someone who’s just, you know, maybe an SEO who doesn’t run that, but they, you know, are working with your team.
Like how does that work?
Will Critchlow: We work in a whole bunch of different ways. I would say. Looking at the folks who are good at this on our team, as well as it in-house, you don’t totally need to be a, you know, math statistics, genius, like our platform can come to handle a lot of that complexity. Cool. We do need SEO capabilities.
So it turns out that coming up with a good hypothesis is hard and that is that’s an SEO skill. So the biggest thing is probably SEO experience [00:25:00] to come up with ideas that might be good. That’s probably the biggest differentiator. In addition to that, a little bit of coding skill is useful, HTML, basic JavaScript, CSS, that kinda stuff, to be able to ideally implement this stuff local that we try and make that as easy as possible.
And then yes, we do have the kind of what you call the concierge service. We call it Autopilot, which is, like run it for you kind of thing. Because having spun out an agency, we know how to do that. Customer service kind of side of things and, impact. So I would say almost all of that team are still folks who work in the agency world.
Who came across, to SearchPilot, and then we started obviously hiring and training into that team as well. And yeah, I think that’s a really fun area to spend that time working because it’s a little bit like being an SEO consultant, but with a little bit more control over the customer’s website and a little bit more credit when things go, right. I guess you might, you, you get, you got the nice chart that goes up and you get to show that to your client’s boss.
Garrett Sussman: And [00:26:00] I love that. I mean, and it seems fun. I mean, not like experiments are fun because it’s like, you know, yes, there’s a lot at stake, but you get to see, you know, whether you’re right, whether you’re wrong, I’m curious, kind of tailing off, of the subject though, in terms of, you mentioned best practices and, and you, you come up with these, these experiments, these hypotheses, To what extent do you, how do you even think about best practices versus like just Google documentation?
Cause we know Google is not always transparent versus your own personal experience, despite the fact that everything’s constantly changing. Like how, how do you come up with like hypotheses in the first place? Is it just intuition? Is it, where does that stem from for you?
Will Critchlow: A lot of experience there’s into that.
I think. We’re trying to make it a little bit more. Process-driven I suppose. So that’s a combination of capturing test history. So, so we, you know, we can have a list of tests that we’ve run in the past that we can tell how they’ve gone or whatever else, um, [00:27:00] some checklists and that kind of stuff. Right.
You know, have you thought about testing all of these different elements on the page? That’s the kind of thing you can, you can build a checklist around, but I think ultimately there is a creativity element in there as well. And, and we called that autopilot earlier. That’s definitely human-powered autopilot.
This is not Elon Musk style, self-driving cars. Like we, we haven’t, we don’t revisit it for a long time, but this is a computer-driven task. A computer is a tool, but there’s still a kind of human creativity element in there to figure out ideas that might work. Even if it’s not a list of ideas that will work.
Right. It is just a list of hypotheses. So, yeah, it’s a question of, of working through all those kinds of things, but it looks at if you think about the work of an SEO consultant who doesn’t have access to testing capabilities, or he’s working with a website isn’t suitable for that, or a page on the website.
So even on these massive websites, they still have a homepage and you still might want to make changes to the homepage for SEO reasons. And you can’t test that in a controlled way [00:28:00] because there’s only one homepage. So you fall back on all of the tools and techniques that you would use if you were, if you were in any of those situations.
And it’s just that at the end of it, instead of saying, you know, dear client, we think you should do this. It’s like, Hey, here’s my hypothesis. Now we can test it. It’s the same process building up
Garrett Sussman: And it just speaks to the, I think to the point of like, whether you’re an agency or you’re a consultant having the experience is so key.
Like having had, like, seeing so many different types of websites, like, you know, after all these years, you and Duncan and building all these websites and doing SEO, it’s like, You know, so much better firsthand than someone who might just be in house working. Like they might know they’re their own website inside and out, but in terms of best practices, that’s really hard to
There are pros and
Will Critchlow: cons of the different…
I’ve seen a lot of people take the journey all the way through, from one to the other, from agency to in-house, and back again. Uh, and there are different pros and cons, but certainly, the best thing I [00:29:00] think about the agency world is, is exactly that. It’s the combination of, well, not only that you get to see a lot, but also you’re surrounded by colleagues who are seeing a lot.
And I think when I’ve spoken to folks, who’ve been in the house, even when they joined significant size teams, some people are a one-man band in house, but even if they’ve got a decent size team, that whole team is just looking at the same website all the time. And yeah, the agency side benefit is definitely, you get to see a lot, but also you’re surrounded by people who are getting to see a lot.
So you hear all those stories. Slack is the thing, right. What we’re trying to do with the certified professional services team is about just the level above that. It’s the same thing. It’s just with data, uh, in more data , more statistics.
Garrett Sussman: It’s so cool. And, and, and I, and I, I, the coolest thing is I love how you and your team, your, your, like, continuing to innovate it doesn’t stop.
It always goes on. I know we’re about at time, we’re not even getting to robots.txt files and your advice there. What have you back. But one thing I did want to ask is, is important. You know, one thing that we [00:30:00] represent at iPullRank, I know you represent at SearchPilot is your diversity and inclusion program.
Um, you have this, this program that you’ve been doing for the last two years, um, last few years, uh, that you’re continuing iterating on the process. Can you speak a little to that and why, what it is, and why it’s important to you?
Will Critchlow: Yeah. Well, so in going back to the beginning, this is something that I read a little bit about this.
I didn’t realize the importance soon enough in Distilled. If I’m being brutally honest, I’m critical of myself. I think, um, you know, it was started by two white guys who are literally the exact same age. Um, you know, we came, we, we grew up walking distance from each other’s houses. Like you can’t get less diverse.
Um, we played basketball, you know, like we’re into the same things. So we’re from the same place. Uh, never mind. We’re not diverse in any sense. And then, yeah, as I said, it wasn’t, it wasn’t something I realized the importance of in terms of building [00:31:00] strong teams, in terms of resilience of the teams, in terms of also the self-reinforcing aspect of it.
Once, once a team looks a certain way or thinks a certain way, it could be, it becomes self-reinforcing and it’s much harder to introduce diversity later. And, um, we knew that we were starting with, uh, the handicap can search by that of spinning out of Distilled. So the senior team has worked together for a decade now, um, the SearchPilot crew, and, um, so we’re, we’re kind of, we have some work still to undo from the fact that we do not realize the importance of building diversity into the, into the infrastructure of everything we did.
So yeah, we knew it was important. We built it into the right, into the core values of SearchPilot that we, uh, we want to capture the strengths of a diverse team in, um, in all senses. And yeah. So, the particular program that you were referencing was that I decided that one of the ways that we were going to do this [00:32:00] was a transparency and accountability report.
And so we, for two years now written a public document that, uh, builds off of, uh, survey data. And, um, so like, um, uh, what’s the word like, um, self-reported characteristics and so forth from the team, uh, obviously aggregated and anonymized and so forth as well as, um, um, behavioral type stuff. So how people are feeling about the inclusion side, I guess, you know, I feel welcomed SearchPilot, answers to those kinds of questions scored out of a score at five or scored out of 10. And for the first time this year, we also included recruitment stats in that. So, um, we had, uh, an optional anonymous survey that went out to applicants for our jobs after they’d applied that said, Hey, you know, we want to hold ourselves accountable for this stuff.
It’s totally anonymous. It’s totally disconnected from your hiring. But if you would like to complete this information, we’d love to catch her up. And so we reported a little bit on, on [00:33:00] those things as well. And, um, you know, we, we still got a long way to go and we’ve been very transparent about that in the, in the writeup, but I hope that by gradually, um, holding ourselves to account to it, showing that we’re working on the inclusion side, And then putting this effort in, on the everywhere from recruitment through to, uh, like promotion and Intel developments and those kinds of things that we can, um, yeah, we can end up with, with that more robust, uh, diverse and inclusive organization.
Garrett Sussman: I think it’s awesome. I think it’s incredibly commendable and I think it’s really important. So I thought it was really important to continue to spread the word on that. I love that if people want to find that we’ll have it in the recap of the blog post, like a link to that transparency document, because I think it’s something that other organizations should take and learn from and, and include.
Will Critchlow: Before we have that link up, it’s just so you can go to searchpilot.com/jobs. And, um, it’s linked to from there that among other places, the about page.
Garrett Sussman: That’s awesome. Will, you’re your wealth of knowledge. You’re a pleasure [00:34:00] to chat with. I really appreciate you taking the time with me today. Um, if anyone wants to find you, what’s the best way to get in touch, where do you hang out online
Will Critchlow: Twitter is definitely the best place to ask me Robot.txt questions. Um, and, uh, I and of course, anything else that people want to geek out about? See I’m @willcritchlow on Twitter and pretty easily accessible that the DM’s are open. Um, and on the more, uh, I guess, work side, um, and @SearchPilot on Twitter is the place to follow for, um, all about test results and taking part in our little polls, where we ask people to guess what the results of different tests are gonna be.
And if you’re a searchpilot.com and sign up for the email lists where again, we email out those test results and we’ve just started releasing exclusively to the email subscribe. Video behind the scenes discussions between me and the consultants who actually build a test or, analyze the test and answer all those questions that we had that we know you have about, Hey, you know, like what about this thing?
Or was it the same on all devices or did you see the same thing? The whole page types? And, uh, [00:35:00] we I’d love for people to get involved in those conversations as well.
Garrett Sussman: I love it. Those tests are awesome by the way. Thank you so much for being my guest . Will, we really appreciate you.
Will Critchlow: Thank you for having me on it’s been a, it’s been a blast.
Great talking to you. This is Garrett Sussman ofiPullRank for the Rankable podcast. Next week, we are being joined to talk tech SEO e-commerce with Kristina Azarenko. Looking forward to that Will Critchlow find him online. Start SEO, AB testing. Hopefully, you have enough traffic and the web pages to do it.
Thanks again. Will, we are signing off.
Leave a Comment