Welcome to the second episode of Open Dialog, the podcast for collaborative SEOs and digital marketers. In each and every episode, we’ll be speaking with the best and brightest minds in SEO, digital marketing and beyond to find out how we can work more effectively, efficiently and productively with other teams, departments and clients.
In our second episode, DeepCrawl’s Sam Marsden spoke with Arnout Hellemans, Freelance SEO and Analytics Consultant at OnlineMarkethink.
Over the course of our conversation, Arnout shared some of his wide and varied experiences working with different development teams and senior stakeholders, telling us how he’s managed to leverage these to have meaningful impacts in large organisations.
A visual summary of this episode has been sketched out by the inimitable Katja Budnikov for your viewing pleasure. A full size image of the sketch notes can be found here.
If you enjoy this podcast, you can make sure you’re the first to hear about new episodes by signing up to our mailing list, and don’t forget to share using the hashtag #opendialog.
For those of you who prefer to read along, here’s a written recap of the episode.
Your Instagram is full of Oysters and beaches. How do you fit in time for marketing with this lavish life of luxury that you lead?
As an SEO you should always be experimenting and always be looking at different things and [thinking] how does something work. So, in this case, everybody who is following me on Instagram sees these pictures as part of a social experiment.
What I’m trying to do here is exactly what you just did. I portrayed something, an image of somebody who has this bizarre life. But I don’t. I just eat oysters once every two weeks or once every week. I just take a genuine picture. That’s all I do.
This is the way the human brain works. Basically, when it associates a person with something, it goes through all of the pictures they’ve last seen of that person. Since everything is so visual on Instagram and Facebook, that overwrites all of the other memories and suddenly you become ‘the oyster man’.
It’s part of an experiment I’m conducting and it was such a success that I never have to talk about boring stuff anymore. I can just talk about oysters and everybody wants to have oysters with me. It’s awesome!
How did you get into SEO and marketing?
I’ve always been on the Internet. When I graduated from my studies, which was in 2000. I saw something going really fast. I wanted a really fast Internet. I was downloading all kinds of games and software.
I found a job as a Product Manager/Owner at Telecom Italia, which was [aiming to provide] fast internet. I went for an interview and they said, “You can’t have this kind of job, but you can be a Support Engineer.” So I started off as a Support Engineer and it was boring, but it was also very interesting. Within two years, I was leading the development for the software we were using to deliver fast DSL lines, which was amazing. I got to fly to London every week on business class. It was ridiculous!
I did the same thing for France Telecom, and then I was approached by a friend of mine to set up a company like Rightmove in the Netherlands, but with more of a content focus. So there was all the houses for rent and everything, but there was also a big part where we would talk about how do you renovate? How do you get financing for it? Where can you find the best builder? How do you question a builder?
We spent £11 million in about a year and we fucked up everything in SEO. We had multiple H1 tags, no unique page titles. We asked for agencies to come up with a proposal, and one of them gave a proposal for £15k to do an audit and I thought, “How difficult can that be?”.
I started reading SEOmoz, their blog and the Beginner’s Guide to SEO. I started making [SEO] changes and I saw rankings go up. It was awesome! This was how I rolled into SEO. But [the property site] was owned by one of the bigger banks in the Netherlands and because of the financial crisis in 2008, they pulled the plug. They sold it for one Euro and I was still there.
The people who bought [the property site] put me in touch with a conversion and SEO agency. The founder of that agency put me in touch with two of the best Dutch SEOs. One is Joost de Valk, the guy behind the Yoast plugin and Peter van der Graaf who is an awesome SEO but a little black hat.
Then I started buying links from all of the famous link sellers back then and decided that I wanted to become part of the agency, and then they said, “No, you have to pay me a lot of money”. I said no and I started my own [agency]. Ever since then I’ve been freelancing, which was [since] 2010 or 2011.
What kind of what kind of clients do you work with now?
These days I’m a shareholder and working for Sitly, which is a babysitting platform in 10 countries. We’ve just launched in Brazil. And we have about a million unique visitors per month. I’ve also got Vintage Cash Cow which I’ve been involved with heavily for the past four years. They buy old stuff and resell it.
Apart from that, I’ve got a few clients in the Netherlands and I’m helping out one of the biggest financial institutions in the UK for about three days a week.
Who are you working with in this large financial organisation and what does your day to day look like?
I’ve been hired to help them improve their skills. That means training their internal people, but also to do sanity checks on whether everything is done properly, build safeguards and help them hire the right people. I’m an optimiser. I don’t have real goals, I just want to make everything better.
It starts with SEO, but it’s not purely SEO. It’s also about data integrity because, in my opinion, we all should be very data-driven. You also need to have the right data to act upon, which in a lot of cases isn’t the case because of double measurement, not setting the filters correctly or not measuring the right goals. I’ve seen companies spend £200k a month without measuring goals and then they say we want to rank higher. No, you first need to fix the basics.
Some people will also hire me to do A/B tests. In a lot of cases, I’ll come in and say you shouldn’t be A/B testing, you should just decrease your load time. You need to fix the basics.
What are you looking for when helping this company to hire new staff? Are they SEOs?
No, it’s more Experience Managers, which essentially means a single point of contact that will do testing, analytics and the actual content. One of the things we’ve changed is that previously [job descriptions] would state at least five years of relevant industry experience. One of the things I’m seeing is if you want to change a big financial, pharmaceutical or online supermarket, you shouldn’t hire the people that you’re competing with because they’re not doing it any better. What you need is people that are open to change, because that’s what you’re trying to achieve.
One of the things that we did was we removed that [criteria], which opened it up to a whole bunch of new people who were open to applying. Years of experience in the same industry doesn’t help you to move forward.
One of the people we hired, used to be a teacher and somebody else we hired came from a whole different industry. What you see is that these people have a different mindset. Especially in finance and insurance, people are very reluctant to make changes there, because of the nature of the industry. What you want is different people that are able to make these changes and are open to challenge what’s there because otherwise, it won’t move forward.
Was the company that had no goals set up, this financial company?
No, that was multiple others. I’ve had discussions [with clients saying], “We’re a big brand! Our direct traffic is 80%. ” Uh, is it? The thing is that direct traffic in Google Analytics is anything that it can’t bucket into anything else. It’s like the shit hole of Analytics. If these assumptions are still floating around, how the heck can you improve stuff?
I think there are multiple things here. One of the biggest ones is that a lot of people when you work in a relatively large organization, you want to change all kinds of stuff. When you’ve tried to make changes and there was just barrier after barrier after barrier, at some point, you give up. If you haven’t been in other organizations where it could be done, you accept it’s going to be like that.
With one of the companies I used to consult with, I came in and said, “Why is caching set to 15 minutes?”. They were like, “Well, if we change something, it won’t be visible for our users”. I’m like, “What the heck! You can do that with hashing or with ETags. There are multiple ways to do it and then you can cache it for three months”. They were like, ”Well, the developers said we couldn’t do it”. This was a huge website with millions of visitors, and I think it’s it’s purely because of a lack of understanding and challenging.
The other thing is that developers are open to change as long as you take them along and build a relationship [with them]. You can’t just go to somebody [and ask them to do something]. They’re going to retreat, and they’re going to go defensive. Building the relationship is extremely important.
I’ve spoken with a few agency owners and if their staff do SEO work for a client, they spend one or two days per week in the client’s office. They’re going to go there, sit there and build a relationship, have dinner with them, or go to drinks with them. Build a relationship based on something else, because that will gradually progress into a nice working relationship.
From an agency’s perspective, if you get embedded within a company and you deliver results, are they going to kick you out? It’s also way better for your people being able to sit with a client and work with them. That’s a benefit of working somewhere, not just sitting in the office and firing off emails and Slack messages. They actually get to sit with new people and get new inspiration. They see things happening, they understand where the hooks are. There’s so much to gain.
Does the client still need an agency if people are basically part of the organisation?
No, they’re only coming in for one or two days or they might do it once every two weeks. I think when you’re too deep into your own products, you tend to lose focus, and you tend to not be on top of your game. Whereas if you hire an agency and they work on different projects, they see different things they can reuse.
I was once told that I should just become the travel SEO guy in Europe but I didn’t want to because then I would become a one-trick pony. I want to work on multi-faceted search on ecommerce sites but I also want to do local search because I can learn from everything. I can progress my knowledge, and I can understand what’s needed. Whereas if I would just be working on travel sites, I would just be reusing stuff on a new client. How much fun is in that?
Have you got any disaster stories you can share with us?
I’ve got a pretty big one. I can’t name the client, but we were asked to rebuild a whole platform. That was already underway, and at that time there were eight developers. They used a CMS called Sitecore. I was a bit wary because I don’t really like Sitecore. It’s still around, it’s extremely expensive and it’s on .NET. The ease of use is relatively okay for the back-end, but on the front end it becomes quite clunky in a lot of cases, but not all cases.
[The client] was using [Sitecore] for data storage and the front-end was a React app. When I started, there were eight really good freelance developers working on it. The whole website was a 300 kilobyte JavaScript file and it would run the whole website.
That company decided, probably because of [internal] politics, to scale up the number of developers. So within half a year, there were 120 devs working there and that creates problems. Getting things actioned wasn’t a problem, but doing it efficiently was virtually impossible.
Fast forward a year and that 300 kilobytes that was running the whole website became 3.7 megabytes of JavaScript. If I were to crawl that using Chromium with my desktop crawler, my laptop would almost melt. It was that heavy because it needed to execute 3.7 megabytes of JavaScript, which is ridiculous.
I did an audit with one of the developers, and we reckoned we could bring it back to 480 kilobytes without losing any functionality. This is what happens if you’re trying to scale under time pressure, because all of these developers will reuse something they already used. They will call libraries multiple times and outdated ones. It’s really messy. This is a big example of how not to set up your development.
We managed to bring it back down to about a megabyte but that is a huge difference. Then I moved on because they brought a permanent SEO in. I think the main problem was that they had the wrong priorities. They hired more people because if you have more people in your organization, you seem more important.
Another one where things get really messy is when you decide upon using a certain platform without having looked at what you actually need. I’ve seen that happen multiple times. For one big insurer in the Netherlands, they were building a big platform and halfway through that process (about 6 or 7 years ago), I asked, “Why are we using a CMS which cost £300k and is loading my e-learning modules in 28 seconds. Why the heck are we doing this?” I saw that in the first proof of concept, and they said, “Well, because we chose this, so now we need to use it”.
Fast forward nine months, and I went into a meeting asking if they would mind giving me £25k and we’ll rebuild it in WordPress. I did and it was way faster. Bear in mind, we’d spent £750k on the previous platform, and now we rebuilt it for £25k. Now it ranked and did everything right.
The people that made these decisions had a lack of knowledge. A salesperson comes in and says that their CMS is used by all of these big financial institutions so it’s the best for you. In a lot of these organizations there’s a [strict] hierarchy, so you’re not going to question your boss and you’re not going to question the boss above that.
Does it take an external voice for people in these large organisations to listen
Ideally not, because they should be consulted before that. In a lot of cases, that’s not the case, so this is what happens. It’s about shaking things up and trying to get them reasoning. This is why I think if I want to make an impact, I want to be hired with C-level back up.
The way I look at SEO is that there are three pillars. One is the technical part of the website, what the website’s built like, how does crawling work and everything. Then we’ve got content and the popularity part, which is the link building, the creative content, the outreach and the PR.
The biggest problem in old companies that want to transform digitally is that when you have a bricks and mortar store, you had store and category managers who provided the layout of the store and the products. These are the first two parts, the technical and the content. Then there was a third one which is marketing, which would make sure people found the store and buy things.
Automatically when you think about hiring an SEO, it’s the third one [that’s focused on] because it’s about getting people in. The problem is the first two, which are the store and category managers, are in a different department from marketing. So if I come in and I want to make changes to the way this website is built, I’m not hired by the right stakeholder. I’m hired in marketing.
Marketing now has to transform my message or repackage and sell it to a stakeholder that needs to prioritise it. But they don’t know what or how to prioritise so a lot of it gets lost in translation. This is why some agencies are now moving to pure technical SEO, PR and content agencies. So you see a kind of a split.
What I saw when I started with this big financial institution was a backlog of SEO points, but they were definitely not prioritized. They were sent in a year ago, but nobody was responsible internally. The part of the company that created the backlog and prioritized stuff didn’t know about SEO.
It takes someone to come in and explain the importance and build up the relationships, understand and explain and give them tools. If you want to look at code wastage or how responsive images work, then you have to show them in person.
What I just did in a project is I went and got one of the people managing the website to change all the imagery to an optimized format. We lost about 50% pageload, which equalled to a two-second faster webpage. That change alone was worth the money. There are some changes where you need to show the results, and then you’ll get buy-in.
Who are you showing the results to?
It’s broad [who I share results with]. I just become very energetic and enthusiastic when I get these results. When I can use a tool like GTmetrix or Lighthouse to show that my score or my page size has decreased by this much, which equals to so many seconds. This is also why you want to have the measurement in place to benchmark yourself.
Another great way of doing it is, I built dashboards using the Chrome UX Report. What I did was in Google Data Studio I created three reports on Time to Interactive for all of the biggest competitors and challengers in the financial market. Just to show them you’re as good as the nearest old financial institution, but all the challengers are twice as fast.
There was a Revenue Impact Calculator as well?
It’s still there. It’s in Think with Google. I don’t really like it, it’s oversimplified. Say you’re an insurer and you’ve got 10 insurance products, one of them has a lifetime value of £200k, another one of £10k, another one of £1k.
The way the calculation works is to give me an average value on that. The way we’ve now approached it is that we looked at our conversion rates, at our current rankings, and then we looked at how much can we go up and down and what the impact on organic traffic will be.
If you create a business case, what you want to do is have a worst-case scenario, which is still positive, but you want to under-promise and over-deliver. What I usually do is the worst-case scenario divided by two or three and keep that as the worst-case scenario. But the most positive one, I’ll just keep it. In a lot of these cases, we’re talking about millions that is being left on the table.
For instance, if you want to optimize page size, you can shave off two megabytes per page even in hosting. If you’ve got tens of millions of sessions a month, that equals money. If that [optimisation] can already pay for it, then that’s the only thing you need to do because that is a saving you can have. Then the added value comes on top of it. So there are different ways of approaching this.
Do you need to create your own model to calculate revenue impact?
Yeah or you take [Google’s calculations] and you calculate based on an average, and then you divide it by two or three just to be on the safe side. You can tell them we’ve used this tool but to take out any of your doubts on it we’ve divided it by three to be on the safe side and we still leave so much [money] on the table.
Nothing is as difficult as doing this in SEO, because I don’t know what’s going to happen. Say you’re an online travel agency. If Google moves deeper into flights and hotels, I can have a very cool business case but if they take away all my organic ways of ranking, then what’s going to happen? So always try to be on the safe side when you do these business calculations.
Have you got any SEO success stories you can share with us?
[For a client] we did a quick fix because they had a lot of problems with internal tracking parameters that would show up. What somebody previously did was block them in robots.txt. So if the non-navigational links had tracking parameters, they were blocked from crawling in the robots.txt.
In a way, it’s good to do it if you don’t have a canonical in place. If you have a canonical in place, you should have removed it because you don’t want to have lots of URLs indexed. This website had instead of 10k URLs, about 120k URLs indexed. So we fixed it. We got the canonical in place and then we fixed the parameter handling. We removed the blocking from the robots.txt, crawl rates went through the roof and rankings got a 20-30% uplift.
How did you go about getting these fixes actioned?
This is trust. In a lot of cases, technical people that want to change stuff but don’t have 60% of the knowledge. If you help them make a better case and you help them solve some of their problems. I just tweeted something today, “How do you become an expert in your niche?” Just help people for free and don’t expect anything in return.
So I helped this person in the company with some of the questions he had, and I provided them and helped him, listened to his problems, provided him with some feedback. I listened to his frustrations, which were pretty much the same as mine and then I asked if we could make this change and he was like, “Sure, let’s just do it.”
When he left the company to work for someone else, he didn’t want to leave but was made redundant. He reached out to me as one of the first people. This is because you have a relationship. So having a personal relationship with people who can make these changes and trusting them is key. That’s the way to do it.
One of the other ones I’ve seen is when people do site migrations but they fuck up the 301s. You see it everywhere. What I see agencies do is say, “Here’s all your broken links, fix them”. What these companies want is you to do that little extra and say these links are broken, this is where you should point them and these are the pages you should recreate or bring back up because they’re very valuable.
Why do you think agencies only give recommendations
I think it’s a lack of knowledge of how companies work. Everybody’s busy. If you give them things that are partly finished, they’re not going to finish it off for you. They’re too busy with their stakeholders and other meetings.
When I’m at big companies, something I really want to change is that half of the people have meetings 90% of the day. If you want to get stuff done, make it extremely simple saying, “Here’s a file. So just put this in your htaccess. Here’s a guide on how to do it”. Ideally, you put it on your own blog on how to do it, so you reference them on your own blog and they might send it forward. There are so many things you can do.
I’ve just had it with one client at a university in the UK that basically said, “Yeah, we know we have broken links, but we would love you guys to just come up with which ones should be reclaimed, updated or redirected because I don’t have time to go through this list. I don’t even know whether it’s worth it. That’s why I hire people like you”. That’s just one of the examples. Make it very tangible and easy for them. People are willing to do stuff as long as it doesn’t suck up their time.
How do you choose a good client to work with on an SEO project?
I have my own process for that because, in the beginning, the biggest pitfall is that you just want to earn money. You take on everything you can get. That pivots you away from you want to do in some cases.
Probably every single agency owner I’ve spoken with has gone through the process of needing to fire a client or who doesn’t know it, but if you ask them, “Do you have someone who drains your energy?”, [they’ll agree].
How do you prevent the drainers from becoming your clients? You do a pre-audit and get access to their Google Analytics, Search Console, you look at broken links, a crawl. You get a simple idea of where the biggest points are. Then, what I tell them is that we can do a session. In the first session I usually only do Analytics. I audit their Analytics set up. Are they measuring goals, using e-commerce? Then I do nothing on that site for about a month. Reason being, I basically want to prove that what we’ve done has increased their bottom line.
The other thing is, I would need something setting the baseline because I’m going to measure form validations or that kind of stuff. So I need dev resources. So that first session sets the stage for the way of working.
I then ask, “How much time do you need to get all of your stuff in order?” I’ve had clients that told me that they needed a month, and after two weeks, they called me and said they were done so we can move on. I’ve also had clients where that first process was supposed to take four weeks, but it took eight. So I tell them you can have an extension once, but if you can’t get it done, we’re not made for each other.
In a lot of cases, it’s those two pillars where you want to make an impact, this is where I focus most of my work. I don’t do outreach. I help them with creative ideas. I help them with recommending agencies that are really good at that. But I just can’t do it. I don’t like it. I don’t enjoy it. I focus on the first two pillars, but if I want to make an impact on the first two pillars, then I need to be hired the right way. I think the biggest one is to fire your first client. Tell them we’re not made for each other.
I was just chatting to Ross from Type A, he told me, “I’m going to fire a client. I’m done. ” It’s a positive thing because he came out and he said they were really, really happy and wanted to increase and do all of this stuff. I guess that also boils down to having the right communication with the other company, the company you work for.
If you are in a big company, or in a small company, and you have an agency running your PPC, you want to log into Google Ads and go into the change history. You’re probably paying them £1k or £2k a month. Look at how many changes have been made in the past month. In a lot of cases, what you’ll see is that they’re not producing. This also has to do with the fact that you don’t speak to them and pay the invoice.
I run a reasonable amount of Google Ads campaigns, and I spend about half a day a week just going through them, managing them, updating negatives, setting up new tests and stuff like that.
Is this a problem with an agency’s ability to embed themselves in the client’s organisation?
No, [the problem is] getting hired by the right stakeholders. So if you focus on CRO, you shouldn’t be hired by the marketing person. If you do technical SEO, you want to be hired by the responsible level within the company. You don’t want to be hired by marketing. If you do PR and creative content, definitely get hired by marketing. Then you’re on the same level, on the same page and you can deliver without having to sell anything.
Try to sell technical stuff to a marketer and make sure they translate it to technical people in the right way. It’s definitely not easy. [Things getting] lost in translation is a big problem.
I think the really cool stuff about this is, and I haven’t only seen this with DeepCrawl, but also other tool suppliers, are the way technical people talk about stuff is not the way some of your users do. Using the right language for searching and for creating content is so massively important because [searchers] have a problem and they’re gonna type in the problem in Google. Then you should supply the answer to the problem, whether it be an FAQ snippet or a featured snippet or in different ways. But it can’t be that they go to the page and read about it and they can’t find it. That’s way more important than anything else.
Is there anything you’d like to shamelessly promote?
So my shameless plug. I don’t have any tools, but if you want to work with me, hire me, get in touch. Just Google Arnout Hellemans and you’ll find me if you think you need help with optimizing.
In terms of tools, I think one of my favourites is Franz. Since we all get notifications all over the place, Franz is a skinned interface for all of your communication. It does Slack, Facebook Messenger, Telegram, Whatsapp and everything in one desktop application. It uses Chrome to render a page, but you’ve got it in a nice overview. It’s really handy to work with and saves me a ton of time.
For Android, I use Slim Social for Facebook. It has Messenger in the app, so no different but it doesn’t have access to your phonebook but you can still use all the functionality. It’s basically, Facebook’s mobile web page, but in an app which is pretty cool. I would recommend everybody to root their phones. Then a third one would be Google Maps because I travel a lot, especially their AR navigation.
I am working on a tool but it’s not there yet. It’s no competition for DeepCrawl by the way. I want to make something that monitors site speed, but in an actionable way. What I miss in quite a few tools, is that in my work I’m too crowded and I’ve got too much to do, so I only want alerts when they’re actually actionable. The same goes for Google Analytics you can set certain bandwidths of things. I want to build something like that. I also want to make it free to use apart from if you’re really happy using it and then charge a small fee.
Be the first to hear about new episodes
A massive thank you to Arnout for being our second guest on Open Dialog and teaching us so much about his wide and varied experiences working as a consultant. You can find more episodes of Open Dialog here on the DeepCrawl Blog and make sure to be the first to find out about new episodes by joining our mailing list.
