 Gwbod hi heddiw. Rwy'n syr wych i'r bobl wrth fy nghymiadau ac rwy'n gweithio'n hoffi'r ffordd o'r ysgol. Rwy'n gweithio. Yn 100, rydyn ni'n ddechrau i chi sy'n oed, Gregg yn y ffren, felly rydyn ni'n gweithi'n hoffi. Rydyn ni'n gweithio i'r hoffi i'r hoffi. Ac rydyn ni'n gweithio'n hoffi ar gyfer ei wneud, ym mhwng i'n gweithio. Chwych. Fy llydym yn ei tyfyd arweithio ac i'r hoffi, gallwch chi nid yw'r yw,? Ond rwy'n buswch ydych chi'n gwaith y Cymru yn ei sio gyda o'r rwy'n cwlad yma? Fawr, dim ond, y gallwn ei wneud, ond mae hynny'n cael ei bod chi'r gweithio ar y mod i'r ffont, yn pethau gyda'r unig, yw'r fwy o'r gwrwstio, ac rwy'n dweud rwy'n cael eu bod chi'n gweithio. Dyma'n mynd i gwybod, y bydd y Cymru yn bidau. Felly i gwasanaeth yn bosb gael, architecture available on the market. So, this is a very simplified version. If you think about it, when you first have a baby, there's quite a few things they need. You can feed them, you look after them, play with them a bit, and generally they're all right. When they become an adult think about a teenager. They want a lot more, they discover what wine is. They discover that they can date, so then think about SEO. It's the same thing applies. So we've started off by having links and keywords and that's all we need it. And now we've progressed and we think more about architecture, we think about duplicate content, we think about time on page engagement. All of these extra factors that become very important in being an SEO. So what's going on? The fact is that Google is only 19 years old. And it wasn't until like Amazon was 19 that it started offering streaming services. When I think about when I was 19 I was drinking lambrini in a park. Like that's not what I do now. Much. So then when you think about what Apple was doing when they were 19. This was what Apple were making and this was their website. I mean it's hardly what we'd expect of that sleek modern brand we see today. This isn't rocket science. The fact is that things change a lot as we age. We develop things become more complex. And that's effectively what Google's been doing. When you think about what their core mission is, ultimately you can simplify it down to being about you and your. So it focuses on whatever your need is. It gives you back the kind of information that you're looking for. So who's this you that Google keep talking about? Because surely that's our job as SEOs. And the answer ultimately is it's the consumer. Google is changing pretty much everything for the consumer. It's focused on adapting. So as SEOs that's what we need to do. We need to focus on adapting too. So I'm going to cover my top five ways for how we survive the changing fickle natures of Google and what we do with all of these changes going on. So my first piece of advice is to stop measuring only traffic and rankings. When we look at things such as these ranking charts we can see. We're comparing high traffic to top page rankings. There's an obvious correlation. Like you don't need to be at a conference to know if you rank number one you're more likely to get more traffic. But that's not the only case. Those of you that were here for Sarah's opening talk, she talked about searches where there were no clicks. And terms such as this and this comes from an AHF study shows that when there's a simple answer or a direct kind of yes or no, quite often there's no clicks. The total traffic that goes to those websites is much smaller and it's pretty obvious that's because of ranking boxes, things like that, answer boxes, all of these different forms that Google are giving us. So what we don't know from that is that high rankings don't equal high traffic. And more importantly, high traffic doesn't equal high revenue. If you rank for how to say December in Spanish, do you think you're actually going to make much money converting on that? It's a very niche product if you do. So how do we measure this? For us, we chart our rankings against the average traffic that we get for those positions. We track answer boxes and new setup displays. And more importantly, we group our content into informational, transactional pages and we track across those as groupings. So we say our informational pages actually, we're not too bothered about the traffic on them because we know that they're just definitions that are coming up. And so then that leads us on to our next point, which is to not look at singular keywords. And why is that? There's been a 61% increase in last year in who, what, where, why, how searches, like those question-based things. And then there's been, come on, there's been a 70% increase in searches that are performed in natural language, like questions. So that's consumers' behaviour that's changed. And what does that mean for us? Well, the first thing is, it's not 1999 anymore. Now, this is a lot of fun on one page, but basically, Google actually sorted their synonym recognition out. They knew what they were doing from the early 2000s. They had a plan. So we no longer need to look at single keywords if you read this because according to Google, we already know what synonyms are. And I think we can all agree their recognition for those is fairly good. The other thing to note is that voice search uses a lot more words. So we can speak 150 words a minute, whereas we type 40. Look at people like Jono, who spoke earlier. He can speak a lot more than 150 words a minute, right? Okay, so think of someone like that doing their search on their mobile on voice. You're going to get a lot more detail. More detail, more words mean a much clearer intent. So if someone searches the phrase SEO, you don't know what they want. They search, how do I do SEO? It's pretty obvious they're looking for advice. So we measured that by splitting it down into the purchase funnel. We use a very simplified version purely because it's client-friendly and it works for every business. And here we look at grouping them by the type they fall into in the funnel rather than looking at site folders. How often does your architecture lead to things being elsewhere that maybe doesn't follow your exact journey of the user? And tying into that, if we're not looking at single keywords, then we need to expand beyond just search volume. So as far as I'm concerned, there is no such thing as an accurate search volume. All of the tools may present you a number. Likelihood is it's not truly what's going on. So this is an example stolen from Verve Search. There's a link at the end if you want to read the rest of their study on it. But what we can see here is that the search volume for all of these terms is no way is it the same, yet every tool will tell you that it is. Quite often people make typos that sort of thing. Google's aggregating it all together. The way we cope with that at the moment is we use Verve's keyword cleaner. That canonicalises down to the simplest version of the keyword. So it would take all of those and work out they're all the same word by doing some nice Excel work. And then once it's worked them all out, it will then split that by how many terms are identical. It gives you kind of a more realistic forecasting figure for what the search volume is in that sector. But more than that, people actually do more than just search on the internet. Like a lot of us think about Google and search as the only action people are taking. They search, they buy, great. But that's not necessarily true. Sometimes they type things. So Q&A forums happen. So if you search how do I do SEO and those 21 essential tips didn't get you to top of Google, what you might do is go on to Cora. And you might decide you're going to read all of these answers and these specific questions. So what is it that make people do that? We use a model called the Valence Arousal Dominance Model. And that states that if you can get someone to feel an out-of-control emotion, they're more likely to type things in and have an active interaction with you. Now, think about out-of-control emotions you might have if you were looking at how to do SEO. You'll be frustrated, angry. You might be so confused you can't cope. There's too much information there. And that's what's triggering these people to type the questions into the forums. Like, yeah, you could Google it and find the answer, but they've gone, no, this is too much for me. And catching that emotion and understanding the feeling of the searcher is going to give you a lot more insight into what content you could be doing and how you could be optimising to go the extra mile to help people through that. The way we do this is dead simple. So we just use some advanced query operators in the search. We put in the domain we're going to look at some of the terms we're thinking about. And then we scrape the search results. We scrape the first 100 and then we search them through an Excel for common words that we see coming up, common themes. We categorise them. That sets a content strategy for us and for our clients. So how do we measure it? The first thing is using the canonical keywords in the forecasts just so you're not telling someone that there's 50,000 searches in their particular niche they're looking at when actually there's 20,000 and it's split over two words. Look for trends in volume changes, not absolutes. So we track over the year, two years, however long we're working with a client rather than just saying this is your search volume today and use a combination of metrics. So we all know that search volume isn't necessarily accurate. So use a couple, put different numbers together and present three sets of the same data point from different tools. You should all know what structured data is by now if you've been listening at all at this conference. So structured data is really going to help you become an entity. Becoming an entity is essential because it gives you the ability to rank in things like this because it's going to show Google who you are. And we're talking about SEO in a world where you maybe aren't going to be just looking at traffic. SEO in a world where rankings aren't the only thing and that means getting that golden position zero and ranking in the answer boxes. So think about the format. They like tables, use tables. They like schema, they like structured data. Use those methods on your site. Zero search results got a lot of critique, all those jobs they've ruined for people that are working on their sites but actually we just proved those sites weren't getting clicks on the traffic necessarily on those particular queries because the number of clicks are decreasing. So either we need to help Google to understand our site or we accept we're not going to get these types of results. When I first got into SEO I can remember being completely confused by someone telling me not to use structured data because you're giving Google your content and your secrets. And I was like, great, but if I don't rank then it doesn't matter what secret information I have because no one's ever going to see it. And that's the point here. We're not in our own ecosystem, we're in Google so we need to cooperate and play ball. So how do you measure? We like to look at the schema and any markup we can that is on the sites of those that are already ranking. We look at really common languages we see coming up again and again and we obviously implement those. We test the performance of pages against each other having different levels of markup on them, different information on them. And the other thing is sometimes we just don't measure it. We simply accept that this is what Google wants, this is what other websites are doing and we get on with it. So the other stage in becoming an entity is building a brand not just a site for SEO. The reason this is so important is because the majority of metrics that actually decide tell us who's going to rank where and how we're going to do really aren't accurate. So they are great at predicting positions one to three. Four to ten they're okay at and 11 plus, they have no idea where you're actually going to rank. Okay? Now that's scary and it means that all the maths and all the ranking signals we're looking at maybe aren't quite right. There's got to be something here that we can't quantify and I believe there's a lot of debate on what SEOs jobs are and I think as SEOs our job is to find out this bit that we can't quantify. Why is it the tools aren't right? Why is it we can't look at a website and go that's going to be position 12 for that term? It's because we're missing something here. We're missing part of the algorithm. So tie that into voice search. If you do a search on voice and you don't have a screen showing you results your Amazon home whatever is going to decide what should rank. So if the products are the same, the price is the same, there's decent delivery. Why would they choose your brand? Why would they read your brand name out to the user? They do that because there's evidence that you're the preferred brand. Whether that's they always buy from brands similar to you or whether that's actually they've searched for you before they've talked about you. How do you get that brand recognition where you deserve to be the one they choose? We believe it's all about making people think, talk, search, click on your brand. And actually as bad as it is as an SEO sometimes this means going offline. Sometimes it means, yeah, put a bloody billboard up. Like, do offline advertising to impact your search, impact your SEO? The way we measure it so we can quantify. We think about brand plus product as a search volume. We know the search forms aren't accurate so we look for a trend line in those. If we've got the budget for the client we do big brand recognition surveys. Do people expect to see you ranking? And then the cheapest, easiest, dirtiest way to measure is to do your CTR versus your average position. You can do this in search analytics. What you want to do is see your average position remain static if it's roughly flat and your CTR is going up and up and up. It means people are recognizing your brand and even though you're not ranking any higher they're choosing you. And it means you are going to become that preferred brand. So let's tie this all together. We know that keywords don't equal rankings or traffic or conversions. And that's because keywords can't equal rankings. We've got conversational search. We've got algorithm changes. There's no magic keyword that's going to get us the traffic. From rankings into traffic we've got more time and surf, fewer clicks, knowledge graph, voice searches. And then even if we do get that traffic people still have loyalty to brands like Amazon where they know they can buy everything they want and it's a click to buy and there's no difficult conversion process. They care about delivery. They care about price comparison. So you won't always win out. So we know that rankings and traffic are no longer the single sign of success. In fact, we'd love it if our clients would stop tracking rankings on individual terms and just track keyword buckets. But I think we're a little way off that for now. Changes in search mean we know so much more about intent. And we can match that intent with the structured data we're using to become an entity. And we can look at more than search volumes so that we can meet that intent. We be the entity. We meet the intent and we build a brand that's going to solve problems and that brand is going to be what consumers want not just what you want. So my final three points. The first one is do not rely on individual metrics in isolation. It's not going to work. You risk losing that data one day and you risk presenting an inaccurate story to the clients, the brands you work with. The second is use the data that can be compared competitively and spot trends. The more and more stuff that's available to you the more trend lines you can look at the better your data is going to be. And the final is prioritise people and user experience and the consumers over your pure SEO. We know that Google is optimising for consumers not for just us mere mortals running websites. So think about running your own site in that way. And that's everything. Thank you so much for listening.