E-mail:
jsem@patrikpilous.cz
I had been looking forward to SEOloger 2023 ever since last year’s edition, which was my first. That’s despite SEO not being my main focus at the moment. It’s a relaxed event made up mainly of networking, beer, talks, a swimming pool, table football and walks. Everyone mixes it up the way they like and picks what they want from the programme. Nobody pushes anyone anywhere. And that’s great.
Disclaimer: This article is pieced together from my notes, so some parts may not make sense without context, and the terminology may not always be accurate.
SEOloger takes place every year at the Skalský dvůr hotel, a compromise location in the countryside between Prague and Brno. This year I again came away with a very positive experience and, of course, plenty of insights on what to focus on. Above all, you can discuss your own problems with people who understand them and can offer their perspective. So what did I take away?
Outside the talks:
- Tips on work habits
- AI, marketing and SEO consultants: what are the prospects?
- How the world of domains is changing
- How to use ChatGPT
- How SEO was done when I was at primary school
- No SEO consultant has favourite developers they would recommend
- Can you make money from dropshipping in the Czech Republic?
Credit for organising the event: Pavel Ungr
Photo credit: Petr Krauz (an amateur photographer with professional gear).
SaaS SEO
The most valuable talk for me was Michal Pecánek’s on SEO in B2B SaaS. I have been working intensively on B2B over the past year, and Michal made several very interesting points. In recent years there has been strong demand for SaaS marketing, and also for SEO consultants who specialise specifically in SaaS. Compared with e-commerce, it has many specifics, so this specialisation looks quite promising.
SaaS companies are often very wealthy and sometimes burn money without much thought. These companies are growth-oriented and need to deliver performance. Michal has worked at CDN77 and Ahrefs, so he showed some examples from these very projects. There were also several examples from HubSpot, for instance lead magnets such as “Marketing template plan”. HubSpot does this well. It gives you a template and then pushes quite hard for the lead and the conversion. If you’re interested in learning more, I recommend clicking through their landing pages for these terms.
In B2B, CLV is often huge. If you calculate your cost per lead properly, sometimes a single paying customer a month is enough to cover a senior SEO specialist on the project.
In B2B, deals take much longer to close, so the attribution model in GA is not suitable. Even 90 days is not enough.
In B2B, only 3% are currently ready to buy and 7% are open to buying. There was an interesting chart on this.
The pSEO strategy: only a few parts of the content change on these pages, while the core stays the same. I have tried this technique before, but it certainly isn’t suitable everywhere, and it feels to me like a slightly outdated technique. But of course, in certain cases it can work well.
I was surprised by how much content creation costs for SaaS projects, and links too. A page or article for $2,000 is common. These websites can’t really afford below-average or non-expert content. Strong expertise is needed, which a copywriter can of course never fully deliver, so someone from the product team needs to be involved.
B2B SaaS companies have valuable assets they often don’t use: the data they hold on users and their behaviour. Original content like this can then be cited and used as part of a linkbaiting technique.
Other undeveloped notes from the talk:
- Distributing content to other platforms and networks. Having a strategy for it. Signals from social media are an important factor.
- Social signals are also good for Google Discover.
- Regularly publishing high-quality content, updating it, working with content. The classics.
- ChatGPT can perceive and recommend. In a survey of where people had heard about Ahrefs, a significant number of people mentioned ChatGPT
- With a B2B marketing strategy, you need to allow for up to 20 people being involved in the buying process
Content types
On larger websites, it’s important to have defined content types and page templates. Viliam Kanianský gave a talk on this topic. Having your page templates perfectly mapped can help with the clarity of the website, and you can see which types of content search engines and users like, and also which content isn’t performing well.
From an SEO specialist’s point of view, for example, I know that on some websites categories or tags get almost no traffic, while elsewhere they get a great deal. It’s then ideal to analyse the possible causes.
You can respond to this by working on the speed of page templates, responsiveness, how elements are displayed, user interaction and structured data. Content types can also be labelled and segmented in GA4.
BEER BREAK.
Search engines and the future
This talk by Sarah Presch only confirmed to me how different buying trends are across the world. In China, people enjoy shopping more, there are no shop assistants, and it’s entertainment for them. The Chinese aren’t afraid (they have no other option) to hand over data in exchange for better service. China is therefore currently ahead of Europe in connecting traditional shopping with e-commerce.
In 2016, Jack Ma, the former executive chairman of Alibaba, introduced the “New Retail” concept as a key strategic priority for e-commerce businesses. It is an integrated retail model that combines offline, online, logistics and data to improve the customer experience and make everything faster and simpler.
Other points from the talk:
- Bing vs Google: Google doesn’t support metadata, while Bing does.
- In China, the homepage title must contain the exact company name as registered with the authorities.
- Registration requires a Chinese phone number, which is a major barrier to entry for new companies. You need registration and an ICP licence.
After the talk there was a debate on whether or not Europe will follow China’s example.
AI from a legal perspective
With the arrival of AI, the talk addressed the line between inspiration and plagiarism, and also between what counts as a copyrighted work created with AI tools and what certainly doesn’t.
As I’m fairly legally ignorant, I took quite a lot of notes from this talk, including some very obvious ones.
All photos are copyrighted works and need to be treated as such. They must always be created by a human. A photo taken by a monkey is not a copyrighted work. Not everything in our camera is ours: the copyright belongs to whoever takes the photo.
- With AI generators, it depends on how much creativity you put into the prompt. Three words are not enough. But a half-page prompt may already be sufficient.
- You bear full responsibility for whatever you generate.
- If you put a particular designer’s style into your prompt, that’s already copyright infringement. You’re copying them.
More technical talks
Martin Žatkovič gave a talk on JavaScript and SEO. My main takeaway from it was the technical classification of websites. On a JS website, it’s important to check that the key elements, such as headings, links and images, are visible in the returned code. The second more technical talk was by Marek Prokop on R. Instead of that talk, though, I had an after-lunch nap. But Vojta, for example, even if he doesn’t look it in the photo below, was thrilled by this talk and plans to put it into practice. As I said, everyone found something for themselves. So see you again next year. CHEERS!
Other reports from SEOloger:
