How much can personalisation impact your marketing strategy? I reveal how you can use data to create large-scale, personalised creative campaigns.
Hi everyone. My name is Ian Flynn. I am the head of creative solutions here at RocketMill. I’ve been at RocketMill for about seven years now. Started off just as a web designer. I started off before that just doing print, and I’ve managed to kind of get to a point now where using the skills that I’ve done with design and web design, and I’m lucky enough to manage a team of seven talented people.
So that is me. As you could have seen there and seen by me here, is, I am actually ginger. And I know what you guys are thinking: What sort of ginger is he? Because I would be thinking the same thing. So what I’ve done for you, I’ve done like, a little ginger spectrum, so here it is. So on the far left, we’ve got soulless ginger, and in the far right, we’ve got smart ginger. And I’m kind of sitting about there, just so you know. So, you can kind of gauge who I am. Right next to Ed Sheeran over here, and Beaker from the Muppets over here. So that is it.
Good, so creative solutions is kind of an ambiguous term, so I’ll just kind of explain what that means within RocketMill. So, we first and foremost, do a lot of creative strategy and facilitation. The thing I’m gonna be talking about today is personalization, and I think the creative strategy is a huge part of that. Because we need to make sure it’s all anchored into one brand idea and one vision.
We do UX and UI, a very big buzzword, UX, but it’s crucial for things like app development and so forth. We’re still very proud to be graphic designers. The old disciplines of typography and hierarchy, etc. still apply within web design. We don’t want that to change. We do converter optimization, so a little bit more commercially led. An intuitive way of designing involves a lot of Java scripts and so forth. Which leads me nicely into the next bit, which is HTML 5 CSS and Java scripts. All of us guys, we can design, but we can also code as well. Because we’re very passionate about making sure that we can implement the things that we want to design.
Lastly, because of the rise of 5G, we do a lot of video concepting. Because we kind of see it as a new way of designing. Motion is going to be a much more regular thing. So, we need to be aware of that. So we do a lot of concepting and editing as well.
So today I’m going to talk to you about personalisation. Again, a very hot topic within marketing. Ultimately, making sure that our creative and our targeting is as targeted as possible. And using data to do that. I think within the modern day, we need this sort of functionality. We all have very different needs and wants, commercially, and just within our lives. So it makes sense to use digital to hone in on that, and to give us nice, personal experiences.
But the question I like to ask with this is, it’s all well and good to have lots of sioled experiences that are kind of tailored to our needs, but what about the big brand ideas? Like the Cadbury’s Gorilla, as we all sure are aware of. “This girl can” campaign which, obviously, is revered within the marketing world. I think these are kind of things that we’ve all bound together and to support, and to follow. Can you do that sort of thing in a very personalised way of design?
So we’re going to explore that as we go forward. So, just to be completely clear what I’m going to talk to you about in this talk today, we kind of separate, especially from a media perspective, controlling two things within digital.
We target the right people, and we share the right creative. They’re kind of two levers that we can pull. The media guys will very much be talking about targeting, but for me today, because I’m part of creative solutions, we’re going to talk about how we show the right creative at the right points to the right people.
It covers all digital touch points, so it’s not just advertising, it’s all parts to the final. So all the way from advertising to app design, and so forth. So it covers everything.
So, I’m going to start off by talking to you about the current ways to personalise digital creative and kind of give you some parallels from traditional media, as well. And I’m going to put it into the following construct, so this is very simplified: Attract, convert, retain, there are lots of grey areas within the middle of that, but it’s a good way for us marketers to hone in our thoughts and show how creative can differ from each part of the marketing funnel.
So start with attract. Personalization in attract is, to say, is not a new thing. We’ve been doing it for years, so I think we’ve all been familiar with direct mail campaigns such as this, where they use the data they have on you such as your name and address to make it as personalised as possible. So the concept isn’t new, I think it’s just been souped up a little bit for the digital spectrum. This is an example of it with Renault so they’ve used vehicle recognition to personalise it based on the car coming into Westfield, here. And it says, “Hello, you in the silver hatchback. I spy something stylish beginning with mmm” which is quite a clever way of attracting someone’s attention.
There’s not just vehicle recognition, which is a way of doing this. Apparently, the Xbox One has a motion recognition, so you can actually target people based on that, as well as how people are feeling. Which is kind of freaky, in a weird way.
This is really cool and I’m sure we’ll be seeing much more of this as data progress, and technology progresses. But unless you’re Mr Mayweather here, who has lots of money just to blow on budgets, I think it’s probably not the top of your priority list when it comes to your marketing strategy. But there is ways of doing it a lot more affordable, so we’ll introduce you, which I’m sure a lot of people already know, something called dynamic creative. And I’m going to let DoubleClick, who is our Google partners, to explain more.
DoubleClick Rich Media Dynamic Creative lets you set up campaigns with flexible content that you can change anytime without ever having to read traffic creatives. Use Dynamic Creative to optimise your campaign, re-target engaged audiences, and fine-tune messages on the fly. Dynamic Creative makes it easy to target by an audience, publishers, time of day, season, location, and more.
So I guess the summary of this is using technology to give you different ads on the fly, based on certain data that is available for us to use. So with DoubleClick, we have something called geo-targe ting, so you can put different ads based on where someone is based, so it’s quite good for language and translation, etc.
You can do it based on the website the ad appears on, so contextual. If an ad appears in a bike website, you can make sure that message is tailored to that audience. Audience targeting in itself so DNP such as Choozle, or we can almost buy audiences based on demographics, or likes, or interests and making sure that our creative is changed because of that and scheduled update so maybe time of year, or season, etc.
We also use a partner called flash talking, as well, who give us a lot more than just these four. And go into a lot of granularity, so it’s super powerful stuff.
To give you an idea of how it works, just broadly speaking, it’s a feed. So this is one for geo-targeting, so we got, obviously, a different locations there in the blue column. And then you can see how the creative change is based on those locations so headlines, call to actions, etc.
Then all of that stuff gets dynamically populated within HTML 5, through headlines and images, and so forth. And it means that you can create one template and it basically edits based on those variables that we’ve set.
From that, as well, there’s a discipline which is coming out of the ether dynamic creative optimization, or DCO, which is the automatic optimization of fewer dynamic ads based on how they’re performing.
Again, really super powerful stuff. There’s a couple of apps out there that do it. But I wouldn’t be surprised if this becomes more of a regular discussion in the next two to five years, or whatever.
Technology is powerful but I like to look at it from the other point of view, as well, so, it can’t be too good to be true. So what’s the the kind of ‘watch-outs’ we have here?
How many of you have heard of Gestalt theory before? Yeah? Oh, loads of you. Good, so the people that don’t know what Gestalt theory is, ultimately, a philosophical theory, I think, which is “the whole is greater than the sum of its parts.” So I’ve explained this in previous forefronts before, and the best way I can describe it is if you think about a recipe like carbonara or something, as one thing, and then the ingredients within that are eggs, cheese, ham, whatever. Put those together and it becomes carbonara, so it’s the individual things become something new. That’s the best way I can describe it.
This applies to a lot of things, like recipes. But it applies to layouts, as well, and graphic design. So first thing I’ll show you here is an abstract symbol, we don’t quite know what it is. Looks like a little squiggle with a plant at the top, then it says, “Plants make us happy.” I’ve got a little bit of more context now, I can see that their silhouette is kissing. Then it says, “They make us want to smooch, neck, and kiss. They also make our bottles.” Now I can see the reasoning behind this poster, there’s actually something that I need to know about. Then I’ve got a tiny little logo down there, which probably doesn’t mean much. And then I’ve got Coca-Cola, so now I know the brand, and actually, now it all starts to make sense, and it’s quite a clever ad for it. But one of those things on its own, doesn’t really mean anything, but once they put it together, it becomes a bigger thing. So that’s Gustalt in action.
I guess the point I’m trying to make with this, in regards to dynamic creative, it’s very easy to discredit dog’s dinner. So if you’ve got lots of different things like images, and text, and so forth, and that don’t necessarily work together, then it’s not really going to work.
This is why I really endorse the idea of a creative strategy. You need to make sure that all of your iterations within dynamic creative work together, and they’re all bound into something bigger than just personalization or dynamic creative. So that’s attract.
Going into convert now. Again, given some comparisons to maybe things beyond digital or outside of digital, a lot of people will know this campaign: Coca-Cola personalising their bottles. Obviously this is much more in the convert side of things, because you’re about to convert when you’ve got a Pepsi and Coke and whatever, this is the personalization in order to help you try and convert more. Again, nothing new. I’m sure it really worked, as well, for Coke.
The best way that I can draw a parallel with that with digital is our conversion rate optimization, offering within RocketMill and digital marketing, generally, is the ability for us to test our websites, to make sure that we are converting as best as we can. There’s lots of tools that we can do this, we’ve got Visual Website Optimizer, which is kind of A/B testing. Google optimizer is the same thing, and as is Optimizely. They all have different functionalities here and there, but they are all CRO platforms.
Personalization platforms are kind of the step up from that. So there’s things like Dynamic Yield, Optimizley X, take it a step further, so it’s not just a stranded A/B test, it’s kind of A/B test to different segments and different audiences. I’m going to let Optimizley X who have a nice walk-through of personalization, to show you how this works.
You see on Optimizley X here, you click on the scarf and you go back to the home page, it says, “your favourite scarf”. This is a perfect example of personalization on the fly, based on your behaviour on site. Which is really cool and powerful.
I spoke to Rhys, our head of data about this, and he kind of words this in a different way which I quite like… I also like his image, because it’s really tasty. Just by the by. It’s hyper segmentation, it’s nothing special. It’s just segmenting people into different audience gaps and refining it and refining it as much as possible.
The benefits of using personalization platforms such as dynamic yield, you can learn about the user as they browse. You can prioritise information based on behaviour. Which, again, I showed you that with the scarf. Some of these platforms have the ability to algorithmically update experiments in real time. It’s a bit like the dynamic creative optimization I mentioned earlier.
They have algorithms which will just show segments of the things that work better and kind of write it automatically. Which, again, is super powerful. The idea with this, hopefully, is to improve your products user experience faster.
The reason why people convert more is that it’s easier for them to do it, or they’ve been given the right information at the right time. By doing this, you can do it very quickly on the fly.
But looking at the flip side of it, again, is a success still relies on an accurate hypothesis. So this is part of the machine learning stuff that I spoke to you about here, but you still need a human to create the experiment and to understand, “Alright, if we do this, it will have this effect.”
It still needs an accurate hypothesis, and by doing that you need to make sure you do the right research and get the right context within the data that you have. The analogy I’ll give here is if going to back to the scarf example earlier, a sound hypothesis might be that you say that people that look for a scarf might want gloves, so you then show them gloves at a certain part of the journey. But if the scarf, for example, was bought because it looked very similar to a Harry Potter scarf and people were buying it as a souvenir, then gloves really don’t apply to that. And you’re showing them something that actually doesn’t mean anything to that user, so you really have to do some digging and make sure the hypothesis is as accurate as possible. Otherwise, you’re using that platform action, maybe, to have a negative effect on their experience rather than a positive one.
Something that I really like to put across is, actually, you might not need some super fancy, expensive platforms to do this stuff for you. A good user experience is now coming to a point where they’ll happily personalise for you. So you don’t necessarily have to try and guess what they want, they’ll just tell you what they want and you just give them the right thing at the right time.
74% of customers feel frustrated when a website content is not personalised, so customers want it, customers want to be able to personalise the content based on what they want, so let them do it. If you give them a user experience that allows them to do it, then both parties win.
This is an example from Bulb. The first part of personalization here is to put a postcode in, nothing new, I then get shown a lovely form, and I’m personalising on the fly here. I’m saying it’s a house, I’ve got three bedrooms, and all of a sudden, I’ve got a personalised quote of how much it’s going to cost me. I can then further personalise it to see what my current gas and electricity supplies are and, just some really nice things like my installation is amazing, average, not great, I cook occasionally. So there are things I don’t necessarily get a whole bill out and do, but it still allows me to personalise in a more refined way.
Again, the cost has gone up. So again, that actually tells me they’re being honest, they’re being transparent here. But for me, just so people don’t know what Bulb is, it’s a green energy supplier. For me, it was really interesting to do this, because of all the questions I had, it was answered for me without necessarily having to talk to anyone or whatever. And this is an example of personalization on the fly. Yeah, you could save yourself a lot of time and money, possibly, by doing it this way, rather than using fancy platforms.
So, lastly, retain. With retention for me, is all about making sure that you have an affiliation with the product that you’re buying. This is a perfect example with the Mitchell and Web, “I’m a PC and I’m a Mac,” really, obviously, well-known campaign in the UK. But, it’s a perfect example of people identifying themselves with a product, and actually kind of grasping it as part of their identity. This is nothing new, and probably more modern examples include Apple, Google, Spotify, and Amazon Music for maybe music apps. Lots of people would choose their preference based on their loyalty towards that company, and even more specifically, the loyalty to the user experience within the app. If that app is really clunky to use, you’d probably go somewhere else. You may actually pay more money as a result of it.
I, personally, am a Spotify user and it’s because I like the experience. I like the functionality that’s involved. Same applies to banking apps. And actually, it’s probably more crucial for a lot of banking apps because they’re going through a massive transformation in the financial industry, where it’s not necessarily customer-facing as much as it was. It’s much more digital-only proposition, so, doesn’t necessarily want to talk to an accountant anymore. You want just to be able to do your stuff yourself, as easily and quickly as possible. Things like apps are a perfect way of doing that. They’ve had to really up their game – the companies – because it’s such a crucial part of their new business.
52% of users say that bad mobile experience made them less likely to engage with a company. So it shows you that it is serious stuff here, and if they for a minute think that it’s clunky and not providing the functionality that they want, they’re just going to go somewhere else. And that’s why user experience is such a crucial part of retention.
The good thing about user experience is you already have the person’s data because they’ve converted. So you don’t necessarily have to go find it; it’s there. For me, now, it’s making sure that your data is properly positioned and structured, and you have the ability to pull the necessary data at the right time to the right places. Often, we’ve got some clients that have loads of data, and so much they don’t necessarily know what to do with it. It’s making sure from the data that you’ve got, you now understand, “Okay, what benefits do you have to use with the data that they have?” So, the way that I kind of tend to do it, is to say, “What value are you having on peoples lives with apps?”. We will see these apps that don’t necessarily provide value, and it’s a bit of tat, you download it and you think, “What is this? This is you just trying to sell to me.” And then you probably delete it shortly afterwards.
But if you’re actually providing value to people’s lives, and not necessarily with a screen, as such, then that’s where the retention piece comes in, because you’re actually being very empathetic towards their day-to-day. A perfect example of this is actually mentioned by Alex, who’s in our team. The Trainline app is really, really good. Basically, why it’s so good is they really consider these parts of peoples’ days. Her example is, she bought a ticket to the Trainline to Three Bridges, and five minutes before it went, it says, “Your trip is trying to depart from platform six in five minutes.” And it tells the collection reference. Before she’s even looked at the sign above, she already knows it. So you’re actually providing information to her before she even realised she needed it.
Shortly after that, then it says, “Your train to Brighton is delayed, now it’s leaving at 18:37.” Again, you might know if you’ve been told that within the boards or whatever, or you may have missed it, you might have gone to the platform and you think, “Where is it?” And so that information is so, so crucial, and provides value to people, and it’s using the notifications functionality in the right way, rather than selling them stuff, and telling people to sign up and whatever. So, a perfect example of using data in the right time at the right place.
Realise that I talk quite fast, it’s just me having a bit of a pause. I guess, the endpoint that I want to make with all of this is with personalization, and all parts of the funnel for me, getting it right for every touch point is absolutely the essence for successful marketing strategy. I’m going to show you how we tend to structure it, and hopefully, it’ll be a takeaway for you guys.
For me, it always starts with two things. It starts with company vision, and it starts with the audience data. The company vision is, obviously, “Where do you guys want to go? What is the purpose of the business? Where is it going to be in ten years? Or even five years, or even two years? Why do you do what you do?” And it’s such an important question. Audience data is making sure you know what the market wants, so if you’re thinking about a company vision and where you want to get, you want to make sure that the people that you’re selling to are going to be wanting what you want.
That is part of this blend here, is making sure that your company is going in the right direction and the audience is receptive to what you’re trying to offer. From that, then, comes the exciting bit for me, which is the brand idea. So that brand idea is a creative expression of where your company is going, and what you’re offering to the audience. And making sure that that is accurate.
Then you go to attract, convert, retain. And examples of those things from a creative perspective, are advertising, Conversion Rate Optimization, and your digital product, so like your app, or whatever, which we’ve been through. Then comes the personalization, the dynamic creative personalization data-led UX, and then comes the iteration part there.
So the point I’m trying to make with this, is a lot of people get caught in the personalization dynamic creative data-led UX as solutions, and forget all the bit above it, which is probably the most crucial thing. The creative strategy should always tie into the brand idea, and all your messaging throughout attract, all the way down to retain, should have that focus in mind.
Should understand where your company is going, understand what audience want from it. That should then give you a stead to then start thinking about changing the nuances with dynamic creative personalization and testing, and going from there. And then that will have a life cycle, and then you might want to adjust the brand idea based on the market. And that’s kind of how it goes.
So an example of a brand idea, going back to Bulb again. They actually use their brand idea as part of the attract side of things. So, making energy simpler, cheaper, and greener. Very, very simple, but it’s a brand idea that people want. And ultimately, the audience wanted cheap energy and their company vision is to provide that, so it blends really nicely, and they’ve used it as part of their attract campaign. Simple as that.
From that, I think, it’s really important to consider the kind of human psychology side of things. There’s a couple of books I love enough, I’ve mentioned a few times before, Seth Godin: Tribes, and Sapiens, which talks more about the human make up of how we act together as a species.
Both of them have parallels, and those parallels are that we like to follow things. We like to follow ideas. We like to believe in myths that we buy in towards. And that is human nature, and that’s not going to change. That’s what brands are, ultimately, is things like RocketMill and Google, we follow them because we believe in what they do.
And that is really the most, for me, the most effective way of marketing. Making sure you do know the ideas that people believe in, and you believe in it yourself. Your main purpose as a marketer is to spread ideas that people believe in. Personalization is not the end goal, spreading your idea is. Personalization is just a tactic in achieving it. And that’s it. Thank you, guys.