BrightonSEO September 2019 Round Up.

Earlier this month we were at one of our favourite search conferences, BrightonSEO. We were also celebrating our 10th anniversary as a business, so were giving away cakes. 🧁 BrightonSEO never fails to disappoint and I have written a recap of the nine interesting talks I attended.

I’ve included all the links to the slides where possible, so please feel free to check those out or contact the speakers directly, if you have any questions about their presentations.



Living in a Mobile-First Indexing World


– Frank Vitovitch (@FrankieVSEO)

Frank’s talk was naturally focused around the importance of optimising for mobile search (as well as desktop) nowadays. He started by explaining that in November 2016, Google announced that MFI (Mobile First Index) was coming. He then went through some of Botify’s recent research data that suggests that most of the world wide web are still focused on desktop indexing.

The advice was essentially the same as before – focus on ensuring that your content is crawlable by the search engine bots.

So what is the current state now? They analysed 34.5 billion URLs and 68 billion Googlebot visits over 498 domains (from a real mix of sites – big, small, enterprise – classified, publishers and retailers). They then looked at the Crawl Ratio – the percentage of (indexable) pages crawled by Google within a 30 day period – and compared the number of desktop vs mobile hits.

One of the conclusions from this exercise was that 63% of sites had moved over to the MFI in 2019, compared to 9.6% in 2018, so there was traction and sites belonging to publishers were leading the pack (73%).

They noticed a massive spike in July 2018 in particular, as Google started moving sites over to the MFI. However, this died down and Google’s attention (desktop vs mobile) is roughly 50/50 now and the actual device-based traffic a typical site gets matches that.

They also drew a few other key takeaways:


  • Responsive websites were moved over more frequently than non-responsive sites.
  • Small sites were moved at a greater rate (maybe because Google was allowing larger sites to have more time to prepare 🤷🏼‍♂️).
  • Websites with more content (words/page – ~1000) were moved over more frequently than sites with thin content, but it wasn’t as weighted towards publishers’ websites as they originally expected it to have been.
  • More ‘technically clean’ sites (sites with correctly implemented canonical tags e.g.) were moved over more often.
  • Page speed wasn’t an overly conclusive indicator. However, if a site was slow (> ~2s page speed) they were less likely to be moved over.
  • Nothing to suggest the results were regionalised.
  • On almost every instance, there is a HUGE spike in bot hits when a site is being moved over. So this is a really good indicator that Google is up to something – even if it takes 30 days to send you that email to confirm.
  • Some sites showed a decrease in bot visits after the move, though. This was possibly because of ‘low content similarity’ – as opposed to ‘content parallelism‘.
  • In another example, poor internal linking and content were possible causes of the site not being moved over.

All in all, this was a good talk, full of good data and interesting insights. Good job, Frank! 👍🏻



Tracking Growth with AMP Analytics


 – Solenn Boulic

Solenn is a technical specialist at Google and gave an insightful talk about AMP.

Speed is crucially important to conversions. You have a 2x effect on conversions if you can get a page to load in <5s.

Common complaints on mobile devices:


  • Pages not scrolling
  • Ads blocking content
  • Slow loading

In order to answer the question, “Is my AMP page performing better than my canonical?”, Solenn’s suggestion was:


  • Measure
  • A/B Test
  • Report

The solution they came up with at Google was AMP Analytics. You can install this by importing the Javascript component in the header of your page.

One of the common causes of issues was pixel marketing. AMP Analytics allows for this. Third party cookies are now taboo, though, due to privacy concerns – the AMP Linker was a solution for this. They also came out with an AMP Tag testing tool. So that would be measuring sorted.

What are you testing for? Performance? Draft and experiment. Test if pages convert better if you implement certain changes. <amp-experiment>

Reporting – Amp Analytics includes integration with GDS (Google Data Studio). They even provide some pre-created templates! Nice! 👍🏻

I can’t find her slides anywhere, at the moment, but if you’re curious about how AMP Analytics works, they also provide a tutorial. This was a good talk and should have been useful for those SEOs who love to get into the weeds with technical data.



Tips for Optimising for Google Discover – Lucinda Wood (@incorgnito_mode)

Luci likes dogs (esp corgis) and archery. 🐶🐩🐕🏹And #GreatBritishBakeOff. 🥧Oh, and the Queen ♕ (think it has something to do with the Corgis)! She also has a clever Twitter handle – gotta like that!

Key takeaways from Luci’s talk:


  • Google is snooping on you.
  • GSC (Google Search Console) shows you traffic from Google Discover now. Compare it to your traditional sources of traffic.
  • Google Discover is like having a rank of -1. Advice is sparse from Google, though, on how to optimise for it.
  • She had some really cool corgi animations at this point in her talk. Hey, art is important!
  • Relevancy is more important than search volume. Think about topics rather than just keywords.
  • Focus on your audience’s lifestylesyougov‘s profiler is helpful for this – don’t know what this is? It’s this👇🏻


yougov’s profiler tool


  • Login to your Google Analytics account and take a look under the affinity section.
  • Site Search can help with identifying areas of missing content (as it records what actual users of your site are searching for)
  • Look at Featured Snippets (what we call ‘Answer Boxes’ at Authoritas)
  • and People Also Ask results (hey, Luci, we do all that here automatically take a look at our FAQ Explorer tool, for instance!)
  • Always consider Search intent
  • and use Google Autosuggest (hey, we do that too!)
  • Ensure your headlines are concise
  • Write clear content
  • Use Eye-catching (large) images that are aligned to your users’ interests
  • Avoid using paragraphs that are too long or wordy
  • Don’t use potentially alienating content (fake news, clickbait or misleading headlines)
  • Use AMP (optional, but really useful and recommended)
  • Complete a form to give Google permission to use your hi-res images
  • Build loyalty (good user experience, customer service, quick download speeds, shareable content, repeat visitors, gamification, VIP exclusivity)
  • Have a mobile-friendly site (check parity between desktop and mobile)
  • Pages have to be indexed in order to be discoverable
  • You also have to comply with Google news policies
  • There is no need for structured data, though!
  • Check the performance of your Discovery Ads insights to check what is working.

In conclusion, Luci said, there are 800m users to go after, so forget keywords – your audience and your content relevancy are crucial factors and personalisation is key! Generate new content for your users (trending, seasonal and evergreen) and publish regularly!

This was a great talk. Engaging, funny and informative. In fact, you could’ve said it was the dog’s bol…! 😀



A Deep Dive Into The SEO Strategy Of The UK’s 80 Largest E-Commerce Sites


Kenneth Christensen

Kenneth referenced the title of a lot of talks asking whether SEO was dead. It wasn’t, he said; it’s just changing.


  • They tested the sites belonging to 80 companies and looked at 50 factors.
  • SEO had a positive effect on revenue and profit. Traffic, keywords, visibility.
  • They looked at site performance, Links, Technical SEO, Page Speed and On-Page SEO

How do companies increase their visibility?


  • By ranking for more keywords
  • Improving their existing positions for ranking keywords
  • Improving their link profiles

How were the winners winning?


  • By engaging with their audience (podcasts, webinars, frequent blog posts)


Technical SEO

Under the banner of Technical SEO, some of the things they looked at were:


  • HTTPS implementation
  • Mobile friendliness
  • Redirects
  • Schema markup
  • NoIndex tags
  • Hreflang

Page speed is important – ASOS’s mobile site still showed a white loading page after 3 seconds. Competitors were fully-loaded in that time.


On-Page SEO

The easy bit, right? None of the 80 companies Kenneth looked at were doing it right. Video content was particularly an issue.

UX (User Experience). There is a clear correlation between conversion rates and SEO.



  • Pricing (include price matching)
  • Social proof (Ebay has some good examples of this)
  • Scarcity (“last one available! Get it now!”)

So, in summary, Kenneth recommended the following:


  1. Do the right things
  2. Engage
  3. Speed up your website!
  4. Focus on UX and the customer journey



Practical E-Commerce Tips to Drive Up SEO ROI & Save Time – Ben Barker (@Ben_Barker1989)

Ben started his talk on how there are often a lack of good communications between teams and it’s common to find that there is a mutual lack of understanding of what others do within the business. I quite agree and that’s probably a very common issue across most industries. You also experience careless attitudes and encounter people who think they know better than others. Yep. All these factors (there were others – see his slides) can mean:


  • Products don’t get launched on time
  • The right products don’t get marketed at the right time (and you miss out on profitable opportunities as a result)
  • People waste time on the wrong things
  • Commercial objectives are disregarded

Possible Remedies:

Ben encouraged use of the following:


  • Having daily meetings. Even short ones – over a coffee. ☕️Talk!
  • Building relationships! Ben found his girlfriend through this process! Did she know you were going to mention this during your presentation, Ben? 😄
  • Working with buying teams and finding out what their forecasts are – other teams know things you don’t (product performance, seasonality, marginswhat is selling overall in the market, even if it’s a product that’s not selling well for you)! I thought this last point was a really good one. Buyers talk to a lot of people in your industry, whatever it is and can be useful sources of information.
  • Combining a product roadmap with forecasting adds up to a successful channel strategy
  • Working with your Visual Merchandisers as what they do can often make or break your campaigns
  • Planning early for Christmas (late launches can equal lost sales) – Xmas can add 50% more revenue for some companies; get your content and ads signed off early, as this helps to manage expectations. Early ad campaigns can help but will affect your CPA (cost per acquisition) – work on retargeting correctly.
  • Testing Google Shopping with zero ad spend (see his slides for how to do this).
  • Work on dynamic search campaigns – add your top performing keywords into these campaigns as negative (excluded) keywords
  • Making use of the available project management tools in order to help with intra-team communication (Trello, e.g. – but it has to be used)
  • Ensuring you have interaction across multiple management levels, whilst reporting in the right way to the right level of manager. Yes, the hiding sheets analogy you used did make sense, Ben! 😁

There was plenty of good observations in Ben’s talk, so it’s well worth a peruse, if you have time and Ben has thoughtfully added some extra slides to the Slideshare version above to give it all some context for the people who weren’t there on the day. Good job, Ben! 👍🏻



How to Build Perfect Product Landing Pages – Alexandre Sigoigne (@siGauss)

Alexandre started out by highlighting how the SERPs is “crazy competitive” right now. With lots of paid and monetised results, together with the impact of Universal Search results over the past few years, overall visibility of organic results is only 40%! 😬


  • He recommended focusing on optimising product pages as they are the best optimisation targets for improving your conversions.
  • 34% of your visitors don’t visit your product pages
  • Blog pages have greater visibility than product pages in Google
  • 3% is a good conversion rate on a product page

You should conduct the following types of analysis:


  • Keyword intent analysis
  • Review analysis
  • SERPs analysis
  • A Technical SEO audit
  • SXO

For Keyword Intent analysis, group/tag your keywords and analyse the demand (Search Volume) for different combinations.

Reviews help your product pages convert at 58% higher than normal, so they’re well worth delving into!

For SERP Analysis, pay particularly close attention to People Also Ask (PAA) results. Hello!!! We do that here (see our FAQ Explorer tool)!

Don’t work for nothing. Write the right kind of landing pages to target the right kind of queries.

They analysed 100k keywords and looked at all the image results that were ranking. As a result of this, they came up with the following image-related tips:

Get the images technically right.

Show the product in use in your images.

Videos can increase conversions by 144%, but remember to use video sitemaps.

Technical SEO


  • Use rich snippets
  • Image alt attributes
  • URL format (avoid complex ones)
  • Optimise internal linking (MyPOSEO use OnCrawl)



  • 50% of traffic is mobile
  • 40% of purchases during Black Friday were on mobile
  • Think fast! Load time!
  • Optimise our product pages (changing the main image on your product can increase your conversion rates by 150%)
  • Use Hotjar, etc

One of the follow up questions to Alexandre was a good one – “When you add URLs to a video sitemap, what do you include? The landing page or the Youtube page?” Alexandre said you should add your site’s landing page to your video sitemap, not the page on your Youtube channel.
Slides from Bjorn Darko’s talk “From Market to Insights to Strategy” – delivered at Brighton SEO on Sep 13, 2019


From Market to Insights to Strategy – Bjorn Darko (@bjoerndarko)

Key takeaways from Bjorn’s talk:


  • SEO data helps you to focus on important products and beat competitors.
  • Get a view on your entire marketplace.
  • Identify user intent, seasonality and trends.
  • Market research is a $46bn industry but they only touch a small proportion of people. At this point, the session became interactive as Bjorn asked a member of the audience how he had conducted market research (2,000 people in this chap’s case). Bjorn’s point was that SEO data can touch a much larger amount of people as search queries = demand. So they used 1m keywords for a well-known sports brand as the basis of their research.
  • E-Commerce sites use larger files than other sites.
  • Focus on your top funnel and convert at the bottom.
  • Focus on SERPs integrations and come up with the correct media type.
  • Inventory & Category Management
  • Prioritize your entire marketing campaigns, based on user intent
  • Media Strategy (choose your partners wisely) – check for shopping cart use and magazines
  • This data gets executive attention, drives revenue and helps to manage stock.
  • Get out of your silo and engage!



How Machine Learning Insights Change The Game For Enterprise SEO


– Chris Liversidge (@liversidge)

Machine Learning has been a hot topic for some years now, so I was all ears for this one. Chris started his presentation by pointing out that his agency drove £366m of organic uplift last year! 😲 Er, wow!

Some key points Chris covered:


  • Work out what is an appropriate use of machine learning in relation to SEO and understand what it is not
  • ML (Machine Learning) is not AI (Artificial Intelligence).
  • “ML is all about understanding connections we do not understand.

How does ML affect SEO?

Use Case #1 (Retailer): how might SEO affect bricks and mortar footfall? But what about the reverse? How does a Black Friday promotion affect what you should be doing as an SEO? This was a good use case for ML.

CDPs (Customer Data Platforms) try to answer the question ‘What are people doing outside my channel?’ Chris said there were lots of accuracy issues with cross device analysis – lot of gaps – and it was very hard to reach a single point of truth. Can ML help? Er, No, according to Chris.

Chris covered the typical purchasing process – Awareness / Consideration / Purchase / Service / Loyalty

Check out his slides to understand what I’m talking about here and why my next notes simply said:

See / Think / Do / Care / (Expansion)

And if you’re wondering what OOH on his slides means, it’s ‘Out of Home’.

Chris brought up the subject of Random forests! This is a topic in itself!

He said they focused on a big client where most of its traffic was unidentifiable in terms of geographical origin.

Sales attribution, Chris said, is broken! 😲 In his opinion, it’s completely useless because there is so much data that is invisible to your analytics software. Data quality is terrible now. But you can clean it up. They worked out that in this case, £13.7m of revenue was being attributed to the wrong channel.

With ML, you can use predictive models to see how campaigns are affecting results. They calculated that an SEO budget increase of just 5% would generate £3m-£5m of incremental revenue.

Chris also touched on the issue of channel cannibalisation.

There was a lot to take in with this one, but if you want to go through in much more detail than I was able to get down at the time, you can view Chris’ presentation here.



How Pagerank Really Works


– Dixon Jones (@Dixon_Jones)

Dixon started by talking about the t-shirt he was wearing – it had the formula for PageRank printed on it. Apparently, these are a bit of a collector’s item in SEO circles now. 😀

So, does PR (Page Rank) still get used? Yes, according to Gary IIlyes.

You can view Dixon’s presentation here:

Another – John Mueller – said it was finished 2006 and made a joke about Google now using cane sugar link juice instead.

According to Dixon, “Pagerank is an early versions of Machine Learning.”

Dixon then went on to explain how the formula worked.

Some other key takeaways:


  • You cannot use PR at a domain level. You need to go down to the granular level with it (i.e. page level).
  • A strong page on a weak website is still a strong page.
  • By the same token, a weak page on a strong website is still a weak page.
  • Every signal is small (a minute bit of data)
  • Every piece is prone to error or opinion
  • But at scale, the error decreases, so your confidence increases
  • There is no direct correlation between the number of links and the amount of Page Rank a URL enjoys.
  • Reasonable Surfer may augment the maths. Not all links are equal! With PR every link was considered equal, but Google correctly (probably) decided this was wrong (maybe ones in the nav or at the top of the page were worth more), so they now re-weight the links before they apply the maths. So Reasonable Surfer lays on top of the PR model.

Dixon also talked about how Google Dance used to take place every month, when all of the rankings would start juggling around the SERPs over the next 24 to 36 hours because of the need, at the time, for Google to use a static version of the web. This wasn’t scalable, but doesn’t happen now, so updates to PR are much quicker.

Page Rank was a key development in the history of SEO and it was good to hear someone explain it in such detail. If you’re interested in juggling some numbers yourself, Dixon said he’d also send a copy of this spreadsheet containing the formula for Page Rank to anyone who wants it. Contact him on Twitter or LinkedIn, if you’re interested.

And that’s a wrap, another great BrightonSEO for 2019. We will be back again in April, but before then come and see us at International Search Summit Barcelona on Thursday, November 14th and SMX Paris on November 18th and 19th.

Leave a Reply

Your email address will not be published.

Fill out this field
Fill out this field
Please enter a valid email address.