Why you shouldn't dismiss the Google Display Network

If you would like copy of the presentation please contact us on hello@jellyfishconnect.com

Video Transcript
Thanks very much. So yes, today I'm going talk to you guys about the Google Display Network and why you shouldn't dismiss it. First thing to point out is I do not work for Google, I get no backhands for this if I tell you and convince you to use the GDN it is of no benefit to me. And actually all of my talk is to tell you how to save money and how to make it work better and not waste money on the Google Display Network. So, I've been working in PPC for about seven years now and I love the Google Display Network. I'm kind of a geek but I do really love it. I've seen it do some amazing things.

I've seen it get excellent results across insurance health and beauty, e-commerce and publishing. And we're not just talking about awareness and things like that but talking about a direct response. I've seen it do all that. But, for every good Google Display Network campaign I've seen, I've probably seen 10, 15, 20 awful ones. And a lot of the accounts I order when we get potential new clients in have Google Display Network campaigns in them that have run, maybe spent, 500, a thousand, tens of thousands of pounds and then been turned off because they're seen as not working, not delivering any results and just left. And that's kid of what I wanted talk about today, is how for publishers how we can actually get the Google Display Network to work and why we shouldn't just dismiss it.

So, before I go into why we shouldn't dismiss it, I just do wanted do a little bit on what is the Google Display Network, just for anyone who's not quite sure exactly what I'm talking about. And then also talk a little bit about the difference between the Google Display Network and search. So, Google Display Network, I thought it'd be quite hard to put back-up on last night. So Google Display Network is basically websites where we can advertise on them alongside content. You can see two adverts on here, one of the top and one on the right hand side, and usually text or image ads. And these come alongside the content like that simple ads.

There are over two million publisher websites in the Google Display Network. And if you think all of these websites have hundreds, if not thousands if not tens of thousands of pages, and all of those pages have ad slots on, it is a lot of infantry. There's a lot of space out there and a lot of reach. And so, because of that, the Google Display Network actually reaches 90% of internet users worldwide. You've probably seen hundreds of these ads, probably see them every day. So, the way it's used is in two ways, for prospecting and re-marketing.

From the re-marketing side, this is where you all have a habit, you go look at a pair of shoes or share a dress and then it follows you around for the next year until you eventually give in and buy it and that works really well. But I'm not talking about re-marketing today, I want to talk about prospecting. And that's where we use the Google Display Network to actually find brand new customers, which seems really hard but is definitely possible. And in terms of where this is managed, it's managed within Google AdWords, alongside our regular search campaigns. And again, I think this is where one of the first mistakes comes in. Because it's managed directly in AdWords alongside our search campaigns, the first tendency can be seen to treat them the same, which isn't necessarily the right thing to do.

So that is the Google Display Network. Now, as much as I love the Google Display Network and it does work really well and I think we should all be using it, you can't deny the power of search. It was already alluded to by Mark earlier, about how good search is. And for me, hand on heart, search advertising is the best form of advertising in the world. And the reason for that is the way it works. With search advertising, someone goes to Google, they have a search, they search for it. Then they see relevant listings to that search. They asked a question or they want a service or a product and then they land on a relevant page where they can get more information about that product. It's brilliant.

A user wants a pair of shoes and so the advertiser can say, "Here are those shoes" directly to them. They're looking for an advert. Essentially, anyone on Google is saying, "Please show me an advert". Whereas with display, the journey might start the same, and in fact, it probably does for most people. With display advertising, the user still goes to Google, they search, they see a landing page and they found their end. Then what the display network's trying to do is pull them away from that end. They've already done the first part, they've already gone through the search process then they've landed on a page and now we're trying to say, "Here's our advert. Look at us".

So essentially, it's a difference between this and this. On the one hand, we have a search for life insurance. Now anyone searching for life insurance isn't just going to Google to read that text. No one goes to the page of Google just to look at it, it's pretty boring page, all things considered. You're going to Google to find the information. Whereas on this side, with display, you've already found your information. You've gone to Google and you've search for maybe, Jeremy Corbyn and you found this article. And now we're trying to show you a life insurance ad and pull you away.

So, even though the person searching and the person seeing this could be the exact same person, and could be just as likely to want to buy life insurance, the intent of what's happening here is completely different. And that's what we've got to remember, is that they work in completely different ways. Not to make it too simple, but the way I always think about it is search is very much like being in the supermarket. You know, you don't go to the supermarket just to have a browse around, maybe some people do, it's a bit sad, but most of the time you go into the supermarket to buy what you need. You need milk, you walk into a supermarket you buy milk, it's as simple as that. Whereas display advertising is more like traditional TV advertising.

You're watching a TV show, so you're watching "The X Factor" on a Saturday night and the ads are an interruption into what you want to do. You turn the TV on to watch the TV show, not the ads. The ads just get in the way. And, in fact, we'll do anything we can to get out of watching the ads, might boil the kettle, play on our phone, go to the toilet, spend some time with a loved one, you know, three minutes of ads get a lot done in that time.

So that's the difference. And the reason I'm probably on about this so much is because this is where the mistakes first come in. With search and display, we have to remember the mind-set of the user is completely different even if it's the exact same person. The mind-set when seeing those ads is completely different. Not only that, the environment they're looking at is completely different, one is a lovely search landscape with the information on and another is a, you know, an image ad alongside the content or the stuff. So the user environment is completely different. And so because of these first two points, we cannot go into this expecting to get the same results out of display as we get of search. If we go in with that mind-set of, our display campaign should perform exactly the same as our search campaigns, it won't work because they're completely different. And if we do want to get results out of display, we cannot treat them the same. Again, because of the first two points we can't treat them in the same way, we have to look at them completely differently.

So, by this point you might be thinking, well, why bother with this display, it seems like a lot of effort and it's not as good as search. But there is potential to use it and it does work really well. Now, the things you've probably heard in the past, if you've been talking to someone like me, an ad agency, or other ad agencies or Google, the kind of things you may have heard about display is it's really good for brand awareness, you know. You can get your image out to millions of people, you can get your brand in front of a lot of people really quickly. Attribution, you might have heard, and Google are going to talk about this after me, you might have heard, while you're looking at last click attribution and that's waiting it all to search, what you should be looking at is first click or linear and seeing what's happening there. Or you're looking at the user journey in the wrong way and actually your display campaign is playing a part in that user journey that you're not taking into account.

So these are the traditional reasons you might have heard. And don't get me wrong, these reasons are brilliant and I believe in all of this. And I do believe display generates excellent brand awareness and that you should look at attribution in different ways and not just that last click. And also that you should take the whole user journey into account but having worked with publishers for a long time, I also know that that's what they might say but what you actually want is sales and leads. That's what you want your display activity to do, you want to look into your account and see a campaign that is delivering sales and leads and a direct response from the advertising. That would be ideal.

Now, it's not so we can't take the other stuff into account but is this possible? And that becomes a new question, is that possible? Is it possible to get that direct result? I think it is and I've seen that it is and I have lots of campaigns where it has been possible but it's only possible...Oh yes, yes, it is possible. Sorry, forgot I had that. Yes. So it is possible but only if we do two things correctly, if we use the right targeting and the right ad. So if we get those two things...and I love this app by the way, that's an all you can eat rest stop, by the way brilliant.

So, if we get those two things right, the right targeting and the right ad, then display can deliver a direct response and can work really well. And this is what I want to talk about, is getting these two things right with our display campaign. So let's start with the right targeting. And, I don't know if any of you will have seen this before but I find this really fascinating. If you go to that URL when you're logged into a Google property, what it will show you is who Google thinks you are and it kind of explains to you why you're seeing the certain display ads that you are seeing as you browse the internet. And that's my profile, that's me when I'm logged into my Jellyfish email address. And it's got it pretty spot-on. It knows that I'm male, it knows I'm between 25 and 34, much on the lower end obviously. And it knows what I'm interested in, it's got different sports, soccer, American football, different types of music that I like. And even has, funny enough, banking and some investing ones, which is probably based off the clients I work with, its picked up from my browser history from that.

But the point is, it knows me. And this information feeds back into the Google Display Network. So, as an advertiser, we can prey on this information. We can use this to advertise to men between a certain age who are interested in X Y & Z. So what I would expect to see, on the Google Display Network, when I'm browsing the internet is ads for finance, for sports, maybe sports apparel and stuff like that. But instead, what I always get is stuff like this, Gucci. Now, I have never in my life been on the Gucci website. I've never bought anything from Gucci. I've never even thought about buying anything from Gucci as a gift. I'm not that generous. And I've never even bought anything similar. And I wouldn't fit into that category anyway. More to the point, I'm on the Telegraph looking at the politics section. So, at the moment, I'd say wrong person, wrong time, wrong place, everything about that is wrong.

Yet, I'm still getting served this ad. I didn't have to look hard for this ad. I thought about putting this together and this was the very first example I wanted. When I went to the Telegraph and this was the first ad that I was served. This happens all the time. I'm sure you've all had the same thing, where you see ads and you think, "I just don't get why this is happening". Now, the way that targeting works on the Google Display Network...there we go. Here are base different targeting methods. I'm not going to go through them all in mass amounts of detail but essentially what we've created here is a kind of diagram to show that we're going from broadest to most granular.

So, at the very top, we have keywords. And this is where I'd say maybe Gucci have fallen down. They might have gone into AdWords, they've got their search campaigns and started creating display campaigns with the same kinds of keywords. So they might put in keywords for fashion and handbags and stuff like that and then it might have merged to the Telegraph because the Telegraph has a section on fashion. But then what's ended up happening is it shown up on their politics page to me, who is completely irrelevant. Now on the starting side, it might seem right you've got into Google, you've put those keywords in but then the way it's actually been picked up because that website does have a section on fashion, its then mismatched.

Now, with the Google Display Network, as I said earlier, millions of websites on there and billions of pages, you could put a handful of keywords, and I've seen this done many times, and just two or three keywords into your campaign and then instantly starts matching to say five thousand websites. Now, let's say it matched the five thousand websites and each one of those websites delivered us one click. And each one of those clicks cost us ten pence. Now, we spent £500 and we have no real data to make any decisions.

We don't know if any of those websites work or not because they've all only had one click but we might be £500 down without having generated any leads or sales and then we're thinking, "Crap, does this work? Let's turn it off". But really, can you turn it off? You don't know anything about it, whether it works or not and it's a bit of a wasted venture. And that's a lot of what I see, kind of campaigns turned on with keywords, it matches a lot, spends money, turn it off, no real earnings from it.

After that we have topics. And topics is where it groups websites together into different topics. Not going talk about that too much but the next two, affinity and end by market. These two use that information I just showed up before from Google. These two pick up people based on their interests and their likes and their age and gender and we can use all of that information. And then the most granular one is placements where we can literally hand pick one website to show our Google Display Network ad on.

So, from thousands of website to one website, it's very different. And if we apply the same mass, if we were to use just one single placement and we would spend that £500 on 10k clicks, then we'd end up with 5,000 clicks from just one website. And you definitely know after that whether that website works for you in terms of, should I advertised on it. In fact, I'd probably argue that you'd know about spending £50 after having a hundred clicks rather than even having to spend the £500 by picking out one.

So, the way I think about this, again I love my analogies, is keywords are very much like fishing in the ocean. Now there is a lot of opportunity in the ocean, there's thousands of fish. There's fish everywhere, miles and miles and miles of sea and you can be there all day fishing and get a lot out of it. But it's hard and it's going to take a lot of money and you might find fish today in one spot and not find them there tomorrow and you might travel miles and miles without ever finding a fish. Whereas, an individual placement targeting is more like that.

Now, I don't know whether there's a fish in there or not but if there is he's probably going to find it and he's probably going to find it pretty quickly. And that's the point. Yes, there's less opportunity, you're not going to reach millions and millions of people by picking one site but if there is any opportunity off that one site you're going to know about it pretty quickly without spending loads of money. So, I had to draw that down for my own good and I just like putting up funny pictures. So here's an example of picking one individual website.

So this is a campaign we did a few years ago for Money Week and this is a website nationaldebtclock.co.uk. It's a very scary and awful website that basically just shows the UK debt going up by £5,170 every second and I kid you not, the site just has this number constantly scrolling around all day and all night, awful. But what we tie this up with was an ad here for Money Week, saying Money Week predicts a looming financial crisis. What we also did was then use that information we know about people on Google to target people who were interested in finance.

So, now you have a website about debt only showing the ad to people interested in finance and you're showing them an ad about debt. Now, compare that to the Gucci ad sharing to me, it's completely different. And this is where the GDN can be very powerful, this is where we can go very granular and just target the absolute right people at the right time. Now, I can't share the results of that example specifically but here are a couple that were in the same vein. Now, the first one is a similar financial title and you can see here we did keyword targeting versus individual placements. Now, again, keyword has the reach.

You're talking nearly 10 million impressions there in the same time that an individual placement didn't even...well, just over a quarter of one million. And, from a clicks point of view, obviously that's 76,000 clicks versus just 2,000 but when we get down to conversions and again, this is the whole thing with the ocean versus that little puddle is that yeah, there is more opportunity in the ocean, there is more opportunity with the keywords. We got 94 conversions but each one costs £112 whereas with the puddle, with just the placement, 26 conversions, okay it's not the same volume but just £22 each.

So if we keep running it at that cost we're going to save ourselves a lot of money. It might take longer to get the results but we'll spend a hell of a lot less to do it. Some more here, this is from a consumer title, and again, you can see from the keyword, 1.1 million impressions versus 100,000 for the individual placement, 43 conversions verses 32 and again, a massive drop-off in that cost per conversion, £60 down to 11. And again, the interesting thing is, and this is real data, and that placement is just one individual website that are I'd showed on whereas that keyword generated impressions on over 12,000 different websites. So even though I do have data here, it's a lot harder to get anything out of it and understand how I can make it better than by picking out the one website and just running on there.

So picking the placements - and again I don't want to go into this too much with picking placement of where we want our ad to show. A lot of the times it might just be the websites you're already showing ads on. You might already have agreements with certain websites to show ads on there and it might be that you want to work more with them but Google do give us this tool which is basically the display planning tool within AdWords, free to use even if you don't have an AdWords account you can use it. And what you can do is put your keyword in at the top.

So in this instance I've put magazines and then you'll see all the placements that Google feel are relevant to that are part of the Google Display Network. So in this instance it shows us what kind of ad format is allowed on the site, in other words which sizes of ads they'll have and then how much...they record how much it's going to cost for each click. Take that with a pinch of salt. And then that even the number of weekly impressions available, so you can see there 500,000 to a million.

So with this, what I would say is, rather than putting that keyword into a display campaign and hoping that we appear on some of those sites, which we will but we'll also appear on a lot of relevant ones. Put the keyword into the planner, see what sites are out there and handpick them and then just test running ads on them. And like I said, you'll know pretty soon because it's just one website. You might spend fifty quid, a hundred quid, but you'll know from that, does that website work or not.

That's the targeting side but targeting's only half the story. If we don't do the right ad then all of our great targeting has gone to waste. And I want to go back to that Money Week example. Now, this did work really well but it worked well after we tested a lot of things. The targeting was great but, you know, we had the right website and we picked the right users in terms of people who are interested in investing, but then we had to get the right ad and user journey right.

Now here, we're talking about a looming financial crisis which plays into what the user's looking at. We tested it against ads like this and this is probably more of what you're used to seeing on the internet around the GDN, offer based ads. Now, the difference between these two is this one here is saying you get four free issues of Money Week. It's a great offer. You're getting them for free, no obligation. This one here mentions nothing about anything free, doesn't even mention issues, doesn't even mention that it's a magazine. If you didn't know who Money Week were, if you didn't know the brand, you'd have no idea what this is.

But this worked times better than that did, even though this is offering something for free, because this is relevant to where that user's mind is going. Right back to what I was saying at the beginning, if someone's searching for Money Week fine, given the four free issues, give them what they want. But if someone's reading content they're not in that mind-set of, "I want to buy" they're in that mind-set of "I'm reading content" give them content, which actually Mark alluded to just before. And simply, it's not just the ad, it's the whole journey.

So here are a couple examples I found while browsing around and this is an article that refers to a hundred and it's talking about the impact post Brexit therefore I'm Brexit in my talk, I'm the most current guy here. And so here is an ad for Footsie500, a trading company. Now, I would say from the outset it looks pretty good. They've got the targeting pretty spot-on. They've targeted an article about the Footsie500 and they're trading company, well done. And this is showing up to me and I am in that group of avid investors because of my browsing history.

But then they sent me here, "Start trading" "Spend 50 quid you'll get bonus straightaway" "Start trading". Well, hang on a minute, I'm interested to in how post Brexit's going to affect my finances. I don't just want to start trading straightaway, that's not where I was that. I was reading about it. You've now just thrown me off and I'm probably going to bounce back. Whereas what we want is something more like this, from legal in general. Again, similar situations this is an article on mortgage lending and how more people are taking money out their mortgages, more people are borrowing against them and what's that going to mean for everyone.

So they've targeted that and again, I'd say they've targeted that well, they picked an article that's similar to what they're doing and they're targeting it with lifetime mortgages rather than just sending through to here's how you get one. They send it through to an article on what is equity release. And equity release is referenced in this article and it's talked about and then they send it to you saying, "You might not be sure what it is. Here's what it is". They still have their calls to action, they have get-it-guide and get called back at the top but they're sending it through to something that the user is already looking at and this is the same thing we try and do.

So, some results off of that and again, these ones are anonymized but it's another finance title. And what you can see here, we tried this so we use the exact same placement, exact same targeting methods and we have an offer ad versus a content ad. So similar levels of serving over this time and what we found, quite interestingly, is that the offer ad got a lot more clicks. It got higher click-through rate. It was bright and shiny and it actually offered a product away for very cheap and so a lot of people clicked through.

But when it actually came to conversions, the content ad, the ad that you're sending through to another piece of content and didn't even reference an offer or saving any money, got a lot more conversions and again, £40 versus £274. It's a lot better so again, think about that mind-set. So, if you ignored everything else I said, three key things to take away from this is, first of all, don't approach them the same. So don't approach the Google Display Network with the same mind-set as search. It won't work. They're completely different. They're apples and oranges, direct-mail versus TV advertising, it's completely different. So make sure you're going into it not expecting the same results.

And then get the targeting right. If you are after brand awareness, if you're after getting your ad out to loads of people, then great use keyword targeting and do a big brand campaign. If you're after just a direct response and you don't have the money to risk, then go granular and highly targeted. And then once you've done that, use what you know about users, use the right ad and use the right landing page. You know a lot about these people, you know what pages they're on, you know what they're interested in, you know what gender they are and what age they are. Use that to your advantage, create ads and landing page journeys that reference that. If you do all that, hopefully should work. And that's me.