It took a lengthy and spirit-shattering encounter with my lifelong nemesis GWR to reach my destination, but following a satisfying slumber in the lovely hotel, I was up Brighton early on the morning of Friday the 27th of April to attend (for the first time) the biggest assembly of SEO fanatics you’re likely to see anywhere in the world.
In truth, I had no idea what to expect from BrightonSEO 2018. I knew there would be talks, and which ones I’d be attending (we split our efforts as a team so everything would be covered), but it was a whole new world for me. I’d never even visited the town before.
When the day began with drizzle and the terrifying pressure of hundreds of people struggling ahead to search for their printed tickets, I was mildly concerned — but no fear! Aside from the constantly heaving crowds, the disappearance of SEMRush’s non-alcoholic smoothies after an all-too-brief morning period, and a keynote that ran on a teensy bit long, it was smooth sailing.
If your search query is “What talks did you attend?”, then let’s get to your SERPs, starting with my personal featured snippets and featuring explanations, notes, observations, and anything else that springs to mind. (Note: results are sorted semi-arbitrarily by my personal Al-gorithm.)
Diving into HTTP/2 – a Guide for SEOs — Tom Anthony
Storming in as my highlight, this talk on the nature and significance of HTTP/2 struck me for a couple of reasons. Firstly, it provided valuable and actionable insight on a topic with which I wasn’t intimately familiar, and secondly, it was built on a strong and enthusiastic truck analogy.
Plain old HTTP consists of trucks going back and forth along TCP roads, and HTTPS is a tunnel around them that stops people looking at the contents. But each truck only serves one request, and each road can only take one truck at a time. HTTP/2 allows multiple trucks on one road, making it much faster to deal with a ton of small requests without slowing everything down.
And you don’t even have to implement HTTP/2 on your server to benefit from it. If you use a CDN such as Cloudflare, you can grant your users the speed of HTTP/2 without making any other changes to your setup.
Great talk, charismatic speaker, interesting topic, and truck cartoons. Yes.
Flexible Sampling: Now, Then, and Next — Taneth Evans
Bereft of any amusing theme, this talk was simply rich with fascinating information. Selling content as a product online, Evans explained, is a hefty challenge. Journalists can’t allow users to read all of their content for free (they need money to live, apparently), but Google won’t rank content it can’t see and has long disliked it when it can see content that others can’t.
Through the introduction of their ‘Flexible Sampling’ policy, Google has now shown that it is willing to compromise. Googlebot can be whitelisted for a publisher at the CDN level, and that publisher can then set a monthly allowance for free articles for each user, though they’re not technically obliged to allow any at all.
As a result of this, The Times rapidly found their content crawlable and competing for rankings, which prompted a fevered rush to go from being entirely apathetic about SEO, to viewing it as a real priority. They then had to handle the task of determining ranking attribution factors in an industry that is incredibly variable and unpredictable.
While I knew about Google’s disdain for paywalls, I didn’t know the details of the situation, and I found this talk very engrossing. It’s absolutely something I’ll continue to think more about.
Diagnosing Common Hreflang Tag Issues on Page and in Sitemaps — Emily Mace
Beset with Lego references (Mace really loves Lego, I’m reliably informed), this talk was a great look into an area of SEO that I hadn’t even come close to before in the form of regional tags. For large countries with an international presence, targeting content by country and making use of the right languages is important— but a lot of them do it very poorly, as it sounds.
Canonical tags must match Hreflang tags, sites in different languages must be consistent with each other and use reciprocal links, and any Hreflang tags must have live URLs with no redirects. It’s also extremely important not to mix up (or invent) language codes and country codes. There are businesses, Mace explained, that handle their regional tags so poorly than they end up seeing several different versions of their site compete for ranking places within one sector.
It piqued my interest to learn that so many top companies make such basic mistakes on something that important. All in all, a compelling effort.
Advanced and Practical Structured Data with Schema.org — Alexis K Sanders
Schema markup is something that has interested me for a fair while, so there wasn’t too much in this talk as far as giant revelations go, but it’s always good to be refreshed and reminded that small errors are incredibly common in markup, so you always need to validate it.
As well as demonstrating how a study she’d carried out had shown that sites set up for rich results improve CTR on aggregate, Sanders noted that JSON-LD is not currently viable for Merchant info. She touched upon how the ‘SpeakableSpecification’ type can point us in the right direction when it comes to optimising content for voice search.
How to Identify Search Intent — Anna Corbett
Search intent is always an interesting topic, because it isn’t just a matter of rigidly delineating phrasing structures. It also involves wild variation from country to country and person to person. To learn about user intent for particular keywords, we need only look at the results generated by Google and draw inferences from what they serve.
Do results include plenty of FAQ highlights? Google thinks the user is after information. Do they include action buttons? Google thinks the user wants to take action. I hadn’t really considered just how much we can glean from Google’s existing interpretation of user intent, so this is something I’ll be paying more attention to in the future.
Perfect Keyword Strategy — Stephan Spencer
Stephan Spencer is a very smooth and confident individual, so this talk barrelled ahead at a brisk pace. The only reason it didn’t make my highlights is that I was mostly familiar with the ground it covered. (Also, he gave away books, and I didn’t get one, so I’m holding a grudge :p)
Spencer noted that keyword research should look for keywords that are relevant, popular, and attainable, avoiding vanity keywords that would only drain resources to no effect. It’s important to separate keywords by user search type and map content thematically instead of by current or preferred site architecture.
How to Do Ecommerce Keyword Research at Huge Scale — Patrick Reinhart & Tom Smits
As the last talk of the day, this caught me at a point of low energy, but I found much to like in their discussion of SEO-friendly title generation for ecommerce using categories and modifiers. It’s important to map back to human intent to get rid of useless terms, they noted, and identify only sensical combinations to cut down a giant keyword list.
I also like their PIE approach. Protect page 1 keywords, improve your page 2 keywords, and expand your other content. I would have taken more from it were I the owner of a large commerce store, but perhaps one day!
Optimising for Search Bots — Fili Wiese
In this talk, Wiese noted that search bots are extremely conservative and won’t push traffic in directions they consider unwise. To cover all bases and account for what bots might not be able to interpret, businesses should aim for ‘Progressive Enhancement’ — supplying the HTML, then the CSS, then the JS, etc.
Something I found quite thought-provoking was his suggestion to delay the soon-unavoidable migration to HTTPS and optimise all site content in order to make the most of the opportunity to get a fresh SEO start. That’s a really smart way to engage with a change that many dread.
SEO In a Corporate Environment — Liraz Postan
In her position at Outbrain, Postan has had to deal with a lot of stakeholders and struggle with internal politics, and her solution has been to step back to ‘Internal Education’: explaining what SEO is about and why it has value at a basic level through international presentations and the distribution of digestible resources.
To make best use of value data, she said, you should turn it into interesting storytelling, creating optimised content mixes that hit all of your KPIs. You should also look to communicate more thoroughly and effectively with dev teams and customer support teams because good relationships there will prove very fruitful in the long run.
Survival 101: Rules for Making it as a Brand’s Only Digital Marketer — Rachel Finch
This talk was quite in-depth, and I took some worthwhile suggestions from Finch’s points: find the budget for proper tools instead of relying on cheap or free equivalents; know the limits of your role so you can say “no” when a request exceeds them; and teach something you’re not all that confident about to understand it better.
You can also make a list of positive feedback to pick you up when you have a bad day and feel worried about your value, and I might just do that. I feel terrified every day that I’ll be discovered for the hack fraud I suspect myself to be!
Command Line Hacks For SEO — Tom Pool
I like the idea of using command line hacks to speed up the SEO process, and Pool went through a comprehensive assortment of basic but extremely powerful functions that can be used not only through OS X, but also, as noted in the Q&A afterwards, through Windows and Linux systems through appropriate free software packages.
I wrote down that I could look into using macros to automate keyword research operations. In hindsight, that might have been somewhat ambitious, but it remains a possibility.
Cut the Crap: Next Level Content Audits With Crawlers — Sam Marsden
Hosted by Sam Marsden of DeepCrawl, one of the main sponsors for the event, this talk was all about saving time for businesses with massive websites that want to audit their content comprehensively but can’t possibly find the resources to pick through however many thousands of pages they have indexed.
When a site gets that big, it’s incredibly easy to look track of what content is there, let alone what needs updating, combining or removing altogether. Automation is the obvious answer, and a tool like DeepCrawl might be just the ticket.
There you have it: my breakdown of all the talks I attended at BrightonSEO 2018. Did you attend any of these? If so, what are your personal highlights? Let me know, should you feel so inclined, and thanks for reading.