 Rich stats of secret stash media once taught me that a lot of people are intelligent But there's this special next level intellect that few people possess Which basically allows them to beautifully articulate complex ideas to anyone That person is Robin Lord One of my all-time favorite talks of dare I say the last 10 years was by Robin at search love where he so gracefully Showed the power of Python and Panda's data frames in comparison to Excel Robin is a senior consultant at distilled where focusing on unusual questions and tricky answers is Just part of his job description. So let's give it up for Robin Lord New York 1930s Little boy called Joe is living with his parents in Brooklyn during the Depression They're so poor that they have to move at least once a year because back then your landlord would give you your first month of rent Free and that was the only way they could survive and yet at six Joe decides he wants to become a lawyer So he works and he works and when he gets older He takes day jobs to pay for night school until he can apply for Harvard and get in And Harvard he's one of the shining stars of his class part of the Harvard review peers tell stories about everyone else in lectures furiously scribbling notes while Joe sits and listens and Absorbs and comes out with thoughts and arguments that hadn't occurred to anyone else in the room Hiring season rolls around and when the dust settles just two people don't have a job Joe is one The firms were looking for people who are smart and driven and well educated Joe was all of these things But even if you jump through all the hoops they set for you Someone could take you into a room and gently explained that the firm couldn't possibly hire someone who's Jewish Joe ends up working for a much smaller organization that has to take whatever work comes through the door a lot of that Ends up being things like corporate takeovers, which are seen as too uncivilized for the major firms I'm sure this is a surprise to no one but halfway through the 1900s things change Some the corporate takeovers are big business at this point Joe is one of the leading experts in this field He helps his firm grow from having just four lawyers to over a thousand and People will hire the firm even if they're not going to use anyone on the staff Just to be certain there's no risk of being in an important meeting and finding themselves arguing against Joe Joe's a flon died named partner in the fourth largest law firm in the world And you could be asking yourself Why is he telling us this story? But bear with me my points coming the thing to remember about Joe is yes He was smart. He was driven, but most importantly He was building up a set of skills that nobody knew was going to be very important He was able to surprise people Now there's a lot of parallels with changes that happened in our industry So for example when Joe was at his peaks. I was the concept of brand We had this idea that brand has value even if we can't measure it and some of the most important brand building campaigns came out around this time A little while later a couple of small organizations began work Google began its quest to completely catalog the internet Facebook began its quest to completely catalog my aunt's means and To begin with nobody really thought much of these little companies and the data they were offering But as we know over time those companies grew until they ain't so little anymore and The people who've been playing around in the background who knew how to work with social media communities who knew how to work with search engines They had an advantage like Joe. They were able to surprise people And around that time a lot of us kind of locked away this concept of brand Now why would we try to measure something so big and messy when we know how many people are searching for our product? How many people are going to see the next tweet I sent how many people land on my site and then convert Now things are changing again So those little companies have become some of the biggest monopolies we've ever seen Marketplaces are getting more competitive costs going up to advertise It's not enough to know how to get a site crawled and indexed when we're working against Bayer moth Companies the days of quick wins are waning which means that as SEOs We're more impacted by the performance of the business and the brand that we ever have been before As an example We can all agree that Maz is a very successful company But in 2018 ran fishkin gave a talk in which he said he wishes He'd been able to measure Maz's product brand because he might have seen competitors like Samrush and Ahrefs coming Likewise Tom Kappa has shown that brand correlates more strongly with rankings than links do You might be show that in 2017 2018 and in 2019 Now these talks hit home particularly hard for me because around that time I'd started working with the APAC wing of a global B2B company It's our job to help 10 teams across 20 countries 30 languages to perform in search And we started with a lot of the usual stuff tech audits and keyword research Training and dashboards we kicked off projects across Singapore and Melbourne and Tokyo and every time we did we asked What does marketing look like for you and the things we heard back with things like? blogging and emails and digital ads, but also things like local country partners Small events opening new locations There was a whole host of things that these teams were doing that we weren't contributing to at all and it was impacting our metrics We would see the organic conversions had taken a dip and look into it and realize that they dialed back events in that territory and conversion rate across the board had fallen or We'd see that we'd started ranking for a couple of terms particularly well And they would say yeah, we've been pushing those terms quite a lot to our industry partners We think we're just getting interest and links. We're not seeing We have competitors who are experimenting with driving up specific branded searches to improve their rankings 35% of our traffic is branded So even if everything else is looking good it can still take a dive if brand falls and a lot of the stuff We're doing does contribute to brand, you know, none of us thinks that when we write a blog post That's going to take someone from I've never heard of them to yeah, I'll spend ten thousand dollars a month Please this kind of activity has value because it contributes to brand And initially we thought this is just a funnel problem, you know blog posts and events They're just very top of funnel will improve our attribution and account for them that way but attribution is very hard to do it doesn't account for things like offline events or People taking a long time to decide about something and it relies on things like cookies to track users over time And we know the browsers like Safari and Firefox are targeting cookies so much. They'd be safer on Sesame Street So our funnel is starting to erode We're as SEO is more and more impacted by brand. We were bad at measuring it to begin with and we're getting worse So at this point, I'm starting to feel a bit hot under the collar, right? But we look into it and we realize that yeah, we're facing some real challenges as an industry But without knowing it we're ready for them We've been building up a set of skills and like Joe like the upstarts of the 90s We're ready to surprise some people in this talk I'm going to take you through my journey to measuring brand the way I do a broad way of getting brand insight and I'm all granular approach to get tactical brand info Now I knew this was information we needed so I started by finding any guide I could but they resulted in frustratingly vague conclusions like you have bad brand awareness in Japan spend more money on brand advertising master all relevant sales channels I Can't say this to my client. They would say I Asked you to help improve the performance of my business and you're telling me to get the attention of the entirety of Japan You realize someone constructed a full-scale Godzilla in Tokyo and still couldn't get anyone's attention So we took a step back and we thought okay. What do we want here? We want brand interest data And we could use surveys, but the problem with surveys is they're expensive They take time and we can't ask everyone so we're missing out on a bunch of data What's more academics have said for years that we can't always trust what people write in surveys So even if it's part of the picture our data can't just be self-reported What's more we need to be able to do a competitor comparison. So it's not enough to be able to just say our brand is strong We have to be able to say is stronger than these people or weaker than these people Now as SEOs for a while. We've had a way of knowing how many people are interested in something. We asked Google We know for example that Kanye West is a very popular man, and we know that Taylor Swift is even more popular still Normally we can ask tools like stat and Mars and SEMrush But they tend to report at country level and we know that brand data brand interest in New York is not the same as brand interest in Seattle So we need a way to break that country up and they also tend to report in blocks of numbers So if we get results back saying yet you and all your competitors get a thousand searches a month There's not much to work with So we need data that's not just self-reported that lets us do a competitor comparison So we can't rely on things like search console and gives us the granular data to break up a country and see the difference between brands Now this is a pretty tall order and as you can probably guess I exhausted a lot of my options before I turn to the free tool I'm going to talk to you about now As a warning when I put this on screen, you're probably going to think I'm a bit of a moron that I've been wasting your time But honestly by the time we finish this section. I think you'll agree that This tool is not being used to its full potential and a lot of people are missing out on valuable brand insight So when I was putting what I was trying to make head or tail of my client in a pack we put their brand into Google Trends This line shows the searches for my my client in Google Trends in India over the last 10 years. I Didn't know anything about their industry in India. Certainly nothing about it 10 years ago But this tells me a really clear story We can also add in competitors and see the red competitor started in the lead and has lost that over time Their industry is getting more competitive. The yellow competitor was behind but over the last two years They've taken the lead This also lets us zoom in on specific activity. So for example, this spike here corresponds with a large influencer led conference they ran This got a lot of press attention really caught the eye of our team on the ground But as we can see after that spike things kind of drop off And it's not until later on when we see their graphs steadily rising That's when they're running a series of smaller ongoing Events scattered across the country Now we already had a big conference planned We didn't cancel that course, but we made sure that from the off we planned in a series of follow-up events Smaller targeting the people we really want to talk to and before the COVID virus hit We were starting to see some really nice changes in this graph and a boost in organic search volume Now one problem we run into here is it's not always clear exactly what the competitors done So we've added in this green competitor now. We can't pinpoint a specific activity What's more the lines are far enough apart that we run into that same Godzilla problem How do we make an entire country pay attention to something else? Well, we don't we have to be more tactical on that and that's where the maps in Google trends come into play So this is exactly that same data searches for these brands, but broken up at territory level We can see the green competitor is dominating and the deeper the green the stronger a hold they have We can also see percentage breakdowns by territory So in Maharashtra, for example out of all the searches for these competitors 75 percent are for the one in green Now one problem we run into here is not every territory is worth the same amount You know 50 percent of searches in one territory can be worth the same as a hundred percent in another So we needed a way to know where's valuable This is where we turn to census data So every country seems to make census data freely available and tell you number of businesses based on Location and industry and size or number of people based on age and demographics So if you have an idea of your target market You can combine census data with proportions of search to work out which territory is worth how much to your you and your competitors As an example Bangalore looks like a good target for us, you know growing tech hub But we could see that the green competitor have been strong there for a while It's actually their global headquarters and according to census data. There actually weren't that many relevant businesses So even if we're talking purely about branded organic traffic We're not getting the return on investment Mumbai on the other hand well the green competitor was investing but they haven't always been strong The census data showed there's lots of valuable businesses there and lots of small to medium businesses Which we know the green competitor doesn't cater particularly well to so off the bat We know that we can push out lots of messaging about small to medium businesses We can look at what people are saying on social media and in press around Mumbai to find out what matters to them Use that to generate content now We start ranking for terms that people are searching which is generating organic traffic, but also Strategically relevant organic traffic. We can take that content push it back out to press They're going to cover it because it's targeting the kind of things that audience are interested in So now we're getting links and press attention and organic traffic and cut through everything feeds into each other So Google Trends oft overlooked is a really powerful source of brand insight It will give us an idea of the history of a territory It'll tell us whether big events are working for us or our competitors and it'll help us find valuable Territories and narrow them down with census data But it's not perfect. You know, it doesn't give us actual numbers We have to rely on proportions out of a hundred and using census data It's not terribly fine grained. You know, we tend to work with relatively large territories and we only get brand level data And we know that a brand isn't just one thing So for example, say we're looking at Sony you could know Sony for PlayStation Headphones TVs batteries cameras DSLRs phones watches movies semiconductors or multi-application cell solters So if we're thinking about what data we want here, maybe it's not just brand level data Maybe it's product specific brand cut through This is where my more granular approach comes in and I was facing this problem when I happened to talk to my colleague Rich cotton who's a senior member of our paid advertising team and he suggested getting this information out of Google ads Now we know when we're working with Google ads We get lots of lovely data like impression numbers. How many times people have seen our ads and impression share Now Google calculates impression share by looking at our targeting criteria Working out the total number of times we could have been seen and then telling us the proportion of times We were actually seen So say for example, we're bidding on mattresses in Seattle and Google tells us we got a thousand impressions and 50% impression share From that we can estimate that we got 2,000 searches for mattresses in that time This is how we're going to get that more granular data Now if you know Google ads you could be thinking can we really believe impression share because Google's really tweaked this formula Say we're bidding on mattresses, but we're not bidding high enough They'll remove a bunch of searches from those impression share calculations. So it completely throws off our numbers We end up having to bid really high on really specific mattress terms and it's not worth the while But in this case, we're looking for brand insight, which means two things First we're willing to spend a little bit more, you know brand investigations can cost tens of thousands of dollars And this is way cheaper Second because we're able to be very specific with the searches we're focusing on we can drive those costs down until they don't really matter Now as you can guess Product level brand data is impossible to anonymize. So I've gone out and I've gathered data on fashion brands across West Coast America Now you could be wondering does this guy know anything about fashion in West Coast America? No I don't I don't work with any of these brands, but when you're doing this kind of analysis, you will have that context So you get an idea of how I approach this I bid on phrase match terms So the exact phrase is H&M dresses, J crew dresses, Zara dresses, Zara maternity, Uniqlo bikini and a hoax of other besides. I had separate campaigns running in LA, Vancouver, Seattle and Portland So I could get city-level data and you could be wondering how much did all of this cost because taking other people's branded traffic gets expensive But the thing to bear in mind here is we don't want their traffic. We want their data So yeah, it'd be really expensive for me Robin law to bid on Zara sale and take their traffic Which is why when I was bidding on Zara sale, I drove all of that traffic straight to Zara Most of the time these brands weren't bidding on the terms themselves. I didn't want that traffic I didn't want to take anything away from them. All I wanted to know is how many people are searching Zara sale Which meant that it cost me just three cents per keyword per day So we don't have enough time to go through all of the examples here But we'll go through three and start with a very simple graph So this is the total number of searches for Zara dresses across LA Vancouver Seattle and Portland during February this year This is the total number of searches for Zara dresses Zara jeans Zara pants in those territories during February this year This is Zara compared to competitors. Now one thing we can see from this graph is Zara is smashing it They're getting way more product specific searches than everyone else, but if we turn to Google Trends, we can see that Zara is doing well, but their brand search actually isn't that much higher than everyone else So what could be causing people to search for Zara plus products so much more than everyone else? One of the things that came up when I was looking into this brand is the site can be quite hard to use So to break down what could be happening here say someone's on the Zara site and they're looking for dresses But they find it quite difficult to use so instead they just jump to Google and search for Zara dresses instead Get the page they want the problem here is they're passing through Google Which is a place where many people advertise dresses and they could go to a competitor, which is lost sales in and of itself What's more Google is seeing someone land on the Zara site not get what they want come back to search results quickly and go somewhere else Now it's controversial as to whether Google uses this data directly in search rankings But we do know that Google's feeding more and more UX signals into the search results that are rolling out changes at the moment So even before we look at competitor versus competitor comparisons This gives Zara some data to work with they if they see that their Generic branded search is roughly in line with competitors, but their product specific search is much higher That gives them the sign that maybe something's broken in their UX They could run surveys on people leaving the site to see what they're struggling with or use Twitter advanced search To reach out to social media users find out what they're struggling with when they say the site is hard to use They can even offer these people incentives to get involved These are people who are already tweeting about the brand who care enough to care that the site is hard to use So they're likely to get their followers involved as well, which is free promotion We can also treat this as kind of similar to a tech order. We're rooting out systematic issues that could be impacting performance So maybe this helps improve our rankings over time Even if it doesn't it helps improve our conversion rate, which makes our organic numbers look better anyway But we didn't come here for brand insight. We didn't come here for UX insight. We came here for brand data So let's look at two of the biggest brands. We have H&M and Zara at a product level kind of clear that Zara is winning across the board except for in maternity There are way more searches for H&M maternity than Zara maternity, which is interesting given how much better they're doing everywhere else Now I asked around a bit about this and overwhelmingly the response I got was Yeah, I'm not really surprised Zara doesn't really do maternity So that's kind of a boring thing for our data to point out I suppose, you know people don't search for something that someone doesn't do So if we look at H&M's maternity category, we see that they've got over 400 maternity products They've invested heavily in it But if we look at Zara's maternity category, which they do have we can see they've got over 200 products So even if it's clear to us as consumers that people wouldn't search for Zara maternity, someone at Zara disagrees Now if we just say Zara is doing worse for maternity, that gives them something to work on But they're still trying to change the opinions of the entire West Coast about a topic, which is a tall order So let's break that data down a little bit more These are numbers of searches for Zara maternity and H&M maternity across LA, Vancouver, Seattle, and Portland Now one problem we run into here is even though we can see Zara and H&M are doing best in LA and Vancouver We're looking at cities of very different sizes. So we end up basically just tracking city size with search volume Unless we bring in our old friend census data So if we do something as simple as divide these search volumes by the number of people in a city We get kind of per capita search number Proportional search. Now we can see that they're both doing best in Vancouver So maybe Zara starts thinking about birth rates in Vancouver and whether it's an area to target H&M is doing second best in Portland. Now Zara is doing worst in Portland. So maybe they decide this is an area we need to focus on. This is where we're losing out the most So in terms of outreach, they could focus on Vancouver and Portland targeting mum bloggers and press to drive up interest That should increase interest in things like their category pages Maybe get them some links which should help them improve their rankings They could also do things like overbidding on maternity terms with PPC Because the value that we're driving with these organic sessions with these paid sessions. Yes short term We're getting sessions in sales But we're also appearing in search results and letting people know this is something that we do. We are an option So it's going to have short-term impact on organic sessions But also much longer term impact on branded organic sessions continuing to drive value As our final example, let's look at anthropology and J crew They have roughly similar overall branded search And it's kind of a mixed bag when we look at product until we focus on plus size There are far more searches for anthropology plus size than J crew plus size and to be fair anthropology really puts in the effort You know, they're number one for plus size They have a number of category pages targeting plus size searches which adults a drive organic sessions now and also change minds They show a variety of body types on their social media They sponsor events like curvy con which let them have models on the runway and pop-up shops and got them featured in blogs They have in from Instagram influencers with over 260,000 followers talking about anthropology plus size and it got them featured on good morning America Which has a morning viewership of over five million? J crew on the other hand We can look at by doing something as simple as searching J crew plus size in Google news That'll tell us how press are talking about J crew plus size and we can use the tools Function to filter to a certain time period to say how that changed over time So for example in 2014 We can see that J crew Aren't doing very well for plus size searches. They're kind of getting slammed for the focus on the lower end of the scale Now if we jump back to 2020 things have kind of changed J crew are doing considerably better The brands the publications that were previously slamming them are now talking about J crew as a brand that is changing modern fashion and Lisa Greenwald their chief merchandising officer has promised shoppers can expect to see their favorite stars in a wider range of sizes And that is what we see on the product page You know, there's a variety of sizes on show the problem is the product page itself doesn't make that very clear What's more J crew have taken the decision that If someone's searching for dresses, they shouldn't have to search in a separate plus size dresses page If you want to dress go to our dresses page, which I think is fantastic But it means they haven't got any plus size category pages So if we search J crew plus size as we see some people are We get an ad for the home page resellers and then a four or four deals page. So we're probably losing traffic here What's more if we click on that ad for the home page the models there Don't really show a variety of body shapes if we look at things like the Twitter or Facebook ads library We can also see what kind of advertising J crew is pushing out at the moment and again There's a variety of styles, but not necessarily a variety of body types So what could they do to address this? Well, they could start by looking at that territory level breakdown and see that Anthropology is doing best in Seattle and LA Well, that's a really clear target for them You know, they can focus on press and outreach around Seattle and LA to make sure that they're really getting this cut through They could also consider doing things like sponsoring events there and definitely consider sponsoring curvy con Even though we don't have data for West Coast America. That seems like a fantastic opportunity They could also specifically target those territories with their social media ads So that lets them get a lot more cut through and lets them run a series of tests on what's working Because they're focusing on smaller populations. Everything gets cheaper. They can do more they can also again overbid on plus-sized terms in PPC and They can create a plus-size category page I think it's great that they don't want to segregate their audience But they can have a plus-size category page alongside the dresses page That at the very least is going to let them rank for J crew plus size, which is free traffic It's going to drive down costs for their PPC and social advertising it's an excuse to link to that page from the home page so now they can show different model types on the home page and Now they've got a targeted well-linked to page targeting plus-size terms So they can start ranking organically for those plus-size terms They could even decide to put a banner right at the top of the page saying why are you here? You don't have to shop in a separate category shop in our main dresses page Everything is there. That's turning a category page into a brand statement Which can get press interest and links which will help them rank even better, which is getting them more traffic Which is changing minds about what I'm through what J crew does which drives them branded traffic further down the line So everything feeds into itself So Google ads Will give us city level product level data It will give us insight into how people are interacting with the search engine around our site to see if our UX is broken It will it is paid I mean it costs us some money not loads But some which means that we can only really get city level data Otherwise it gets very expensive and we don't get any historic data like we do with Google trends Now I've spoken to a lot about the kind of things we can do with this data But haven't told you really much how to set up the campaigns. I'm put out a bunch of Fairly dry bullet points that you can go through here that will tell you how to get the results you want But by means of wrap-up Whether we're using Google trends or Google ads because they both have their uses we can use this data to tell our Colleagues our bosses our companies what to put on billboards and where to put them when to judge ads by completely different metrics What newspapers to target social media to invest in influences to reach out to emails to send when to judge organic sessions by Completely different metrics everything up to informing our five-year plan Which will help us give our customers what they want rather than just what they know us for This is all ways to give your company direction if you're ahead of marketing I've given you data your competitors don't have if you're ahead of SEO I've given you a way to start expanding your role owning these strategic conversations They're impacting your team already as it is. This is just a way for you to get credit for all that hard work But the most important thing here isn't here are some tricks in Google ads here are some tricks in Google trends The most important thing here is none of this is new to us This is looking at what people are searching Figuring out why we're not appearing and doing better That's our day-to-day. That is our skill set We are facing some real challenges as an industry, but we're not just ready for them We're made for them. So let's go surprise some people. Thank you