Subscribe to RankUp on Spotify | Apple Podcasts | Google Podcasts | RSS
Technical SEO is crucial to organic success and a part of the SEO job description that many in the industry enjoy. Many technical SEOs have also found ways to use their skills to look into automation opportunities and rethink the ways in which different SEO tasks are carried out.
Luke Davis is one such person. A technical specialist here at Impression, Luke has also taught himself Python and applied it to a range of different activities. If you’re considering specialising in technical SEO yourself, or you’re simply interested in hearing other people’s views on how this area of the industry is developing, Luke’s thoughts and experiences are a fantastic place to start.
He joined us on the podcast to discuss his route into technical SEO, and what he sees as important in the field right now.
Like always, you can listen to every word of our conversation with Luke via the player at the top of this page, or by finding the RankUp podcast on your app of choice. You can also keep reading to see write-ups of some of our highlights from the conversation.
If you want to hear more from Luke, you can find him on Twitter at @LukeDavisSEO. As he mentions in the episode, Luke is also a prolific blogger – find his work at Sampleface, Cultrface, LOGiCFACE, Playrface, Distant Arcade and Pandog.
Twitter is also the best place to find the rest of the RankUp team: Edd (@EddJTW), Ben (@BenJGarry) and Liv (@seoliviamae).
Introducing Luke
Ben: What’s the overview of your career so far?
Luke: Going way back to 2000, I was living in America and my mum used to work at my elementary school library. As she worked after hours, I’d wait there for about three hours until 6 in the school computer lab, with access to the internet, getting an idea of what it was all about.
Back in England, I got into web development and coding sites in HTML and CSS – a very sort of rudimentary HTML back then in 2000/01 – and it’s carried on from there. I got into blogging around 2008 and that merged into my music tech degree at uni. That led me to a music label where I started getting into digital marketing a bit more.
Then I started my own blog – Sampleface – which is a music and music sampling blog. Once I left the label and finished my degree, that was when I really introduced myself to SEO, using the blog as a sandbox for different things I was learning and reading about.
In 2019, I finally got my first job as a copywriter before quickly moving into SEO properly, and the rest is history!
What do you do as a technical SEO specialist?
Ben: You’ve recently been promoted to technical SEO specialist here at Impression, so what does that mean for your day to day job?
Luke: It’s fundamentally less about writing and more about the inner workings of websites – about how they get crawled and indexed in Google (or other search engines) and presented to the user.
Now that I’m a specialist, there’s also a lot more strategy involved. There’s more task creation, and I’ve had to learn to delegate things that I was more accustomed to doing myself. That’s new for me because I’m very much used to doing tasks and now it’s the other way round.
Challenges of delegating work
Edd: Did you find it difficult to relinquish control of your tasks?
Luke: It’s very hard. I like the tasks and getting into the data, looking at things and finding answers and solutions, but when you need to work more on strategy you have less time for that. In time, and with all sorts of different colleagues to work with, it’ll get a lot easier.
You may end up finding that if you pass some tasks on to other people, they may be able to gain new insights that you’ve not seen before, which inevitably helps the strategy and the client a lot more than if you’d just focused on it yourself.
Why did you choose the specialist route?
Luke: My background has always been technical – I used to put it on my CV or cover letters all the time. It’s always been the case, especially with my humble beginnings on the web, I’ve just been more interested in that side of things and how it works.
Even though I’ve done a lot of writing as well, it made sense to move towards technical because it’s catered to my skills and my interests. I knew that I was already interested in that side of it, and in improving there more than in content.
It’s more aligned with the tangible and logical side of marketing, so I can actually see what isn’t working, and I can look into different reasons as to why. But writing is more subjective, and it’s not really aligned with how I work.
What should we be keeping an eye on in technical SEO right now?
Luke: I’m noticing a lot of site migrations with a lot of clients, and increased use of JavaScript frameworks. I think there’s a pivot towards PWAs (progressive web apps) and single page solutions, especially in places like SaaS or business solutions. Both come with their own respective challenges, and I’m personally having to deal with them for different clients.
For site migrations in particular, it can be quite daunting because you’re having to move everything that’s there on a site to a brand new place. It’s like moving house, settling down in new surroundings, and then having to reset everything for Google to pick it up and understand what it is and where it comes from. It’s having to reconnect everything in terms of authority and link equity.
There are a lot of pitfalls that can come with losing traffic and potentially not getting it back, so you have to be prepared for that. I definitely recommend using a checklist or something of that nature until you get more comfortable with it and develop a sixth sense of how things can change.
As for frameworks, whether you’re technically astute or not, I’d recommend reading some of the documentation around them if you know which one you’re working with, and then looking around the staging site (if the client has one) to get a feel for how things are, particularly in the back end.
You want to familiarise yourself with these things so that when issues arise, you’ve got some kind of standing, and you’re not in the dark too much. The devs will have a significant amount of control over the site because they’re the ones that manage it, but you want to be able to understand and express what you need doing if you have tasks that involve their work or that might change things in the framework.
It’s important to get an understanding of these frameworks and the systems behind them because while they may look good and they may offer some speed optimisations, it’s important to understand that if they’re not done properly, they can go the other way. So it’s crucial to keep on top of these things and make sure that they don’t!
Using Python in SEO
Edd: We had Ruth Everett on the podcast last year and you’ve said that she was a big influence on you learning Python. When she was with us, she discussed some of the applications for Python in SEO, but how have you personally been using it in your work?
Luke: The very first thing that I built and have used in work since is a SERP scraper. When I discovered that you could scrape SERP data with Python, I knew I wanted to do it! I learned the different modules that you need, like BeautifulSoup, the Requests module and so on, and then learnt how to take the data that was there in Google Search and put it into a readable format.
I also learnt about APIs. One of the fundamental things that Python has exposed me to is APIs and how to communicate with them. You can do that with any programme, but learning Python and being able to do it has been really important for me.
That SERP scraper was a bit of a labour of love because sometimes things wouldn’t work, and I’d spend hours getting headaches looking at it, then having to sleep on it, then waking up and going back to it and hopefully having a eureka moment…that happened quite a lot when I was building it!
I was able to use it in a work setting at my last job in terms of creating distinct briefs for content, rather than just writing about topics and telling people to use a keyword a certain number of times, which obviously doesn’t work anymore. I could use the scraper to find out what was being written about certain topics, then create briefs off the back of that with usable data. The whole process has been really useful to me in learning Python and helping clients.
In other areas, like my blogs, I’ve been using natural language processing to understand the blog content that I’ve written, like using natural English language classification to find ways of understanding the text, then mapping it to certain tags and categories to improve internal linking, which is something I’ve been doing for the last year or so. That’s been really interesting, too, and has exposed me to machine learning, which is a whole other thing!
A couple of examples of automation in SEO
Edd: Python is one tool that can help automate certain tasks, but also in the past few years we’ve seen more discussion around the introduction of AI and content automation, too. So where do you feel like the biggest opportunities for automation lie in SEO at the moment?
Luke: I think there are two main areas for automation: the manual tasks that are done often – that we might say waste time that could be better used for strategy, content etc. – and for it to complement some of the more specific work that will still need human beings to carry out, rather than leaving those tasks to AI completely.
These areas include tasks like writing meta descriptions. You could automate those with different scripts or tools that use Python to gain an understanding of a text and generate a meta description for you, which is really useful. You can then add a human touch to it if you want to.
For me, with one of my clients, I’ve been marrying the two: using templated meta descriptions and an automation script that inputs certain words or a brand name into three different sets of metadata, and automates it across different products. In this one example, there were 100 different types of a product, and having to write meta descriptions for all of those would have been very time-consuming.
Automating that made sense, but it still had the human element of the templates being written initially. Rather than full autotext, it was semi-automated, which I think worked really well. Full automation would have been quicker, but this way is also more readable, which is the key thing to make sure a user clicks on the search listing.
Meta descriptions are a great example of where to use automation because it’s one of those manual tasks that takes time that could be better used elsewhere in places that need the human element. Reporting is another good place for automation. If you’re constantly having to input numbers, you can get Python to do it for you, which could be really helpful.
Join the conversation
To hear all of the content from Luke’s interview, listen to the podcast episode using the player at the top of this page, or find the RankUp SEO Podcast channel on your podcast app of choice.
Edd and Ben will be back soon with a new episode of the RankUp podcast. In the meantime, you can find us on Twitter at @BenJGarry and @EddJTW.
If you’re interested in being a guest on the show, please reach out to us on Twitter or via email.