Episode 13

Expert-led podcast discussing trending articles, and news in the AI and attention spaces.

Intro_square (5)


Webcasts led by experts

Driving Personalization with Data

When we talk about using data in effective ways, one of the key use cases is personalization. The impact is undeniable: 

  • 89% of marketers see a positive ROI when they use personalization in their campaigns 
  • 60% of consumers say they’ll become repeat customers after a personalized shopping experience 
  • 80% of consumers will share personal data in exchange for deals or offers 


Who better to discuss this exciting topic than Head of Media and Digital at Arla Foods, Rob Edwards, and VP Omnichannel Commerce at The Mars Agency, Ross Halliwell. 

Q: How do personalized recommendations and targeted advertising impact consumer behavior and brand loyalty in the ecommerce landscape? 

Rob: There are two simple answers to this, either incredibly well or incredibly poorly. Depending on how you go to market, the data and the way you approach the concept of engaging with consumers, it can be tremendously powerful as an accelerator to loyalty from a consumer. If you get at the opposite end of the spectrum, it can turn people off. It's not authentic. 

It's a really important piece of focus for us within our business. We don't have first party data, so we have to rely on third party support and using the time and energy to get into that granular data around persona behavior for our categories, for our products is really important.  

So really two ends of the spectrum. It can be a superpower or it can be an absolute turn off.  

Ross: It's the classic right message to the right shopper at the right time. If you miss any one of those, those three rights, that's where it kind of falls off in that sense. Clearly the opportunity is, is huge in that respect in terms of, we see better dwell time, reduced acquisition costs, improved conversion, average order value and volume going up. It all drives up the flywheel of loyalty in terms of recommendation and repeat purchase.  

Rob: When we look at media metrics, we look at effectiveness. We can drive and be successful in determining KPIs and success in KPIs. But those KPIs also have to connect to a consumer app or a conversion. If it's all about reach, if it's all about getting eyeballs then you can win, but are you winning on the bottom line? For a CPG or FMCG business, that connectivity of KPIs along with your measurement and approach and tactics, really important. I've worked in businesses where it's not been the same, and KPIs drive people apart and departments apart. There's a real premise around how answering the question you've asked, but also internally as a business, it's really important that KPIs align through that process of sales and marketing together, working in harmony. Because if you're telling the right message, you're telling the right people, great. But if the sales approach and the impact in store is not there, then there's a disparity there.  

Ross: I certainly see that a lot with the metrics that aren't quite aligned, whether or not its sales driven, and people looking at ROI and ROAS, that that's where it tends to fall down depending on what people are targeting, right?  

Rob: Your point on ROAS is a real one. We were old enough and long enough in the tooth to have been involved in many different businesses with many different clients. You can play the ROAS game. I can play the performance game, but is it driving branded growth? Is it driving sales really outside of a core cohort of consumers? Potentially not. This concept of incremental growth and understanding new data that's coming in is the really important phase when you start to look at again, how do you drive consumer loyalty and connectivity outside of the people that know you. 

Q: Do you think AI driven personalization can lead to unexpected consumer discoveries? Or does it risk narrowing consumer choices to what the algorithm protects that they were like? 

Ross: It's arguably both depending on what the shopper experiences like ultimately. The intelligence that sits behind that from an AI perspective, the machine learning and the data in which the AI is drawing from, because it's only as good as that, that it's got to draw from in. It's also a case of striking the balance between providing the shopper with choice but also showing that you've actually taken on board that personalization element because if you take a shopper who's in the market looking for chocolate or they're searching for chocolate, they're looking for a gift to somebody, what do you choose to present to them? Because if you search on a Tesco or a Sainsbury's or an Asda for that, for argument's sake, you're going to get easily a hundred results. That's too much. But how many is a personalized recommendation? I don't think anyone actually knows what the right amount is, but clearly you want to be able to show probably a bit of a trade -off opportunity. You want to show them some brands are familiar with, but maybe you might want to show them some other bits, but that's a retailer stance on it. From a brand perspective, you're going to want to show the wider breadth and get people more immersed in the world that you've got and the wider portfolio. In Rob's world, you're always going to want to try and go for the cross -sell where you can and where it's appropriate to do so. 

Rob: I think the concept of personalization, just by the very definition, is very personalized and everybody's different. This is the interesting thing around e -commerce in the UK from a retail perspective, from a platform perspective from the ability to understand and learn and then be more creative. I think that's the next avenue for personalization is the creative feel of that experience on an ecomm platform. I still think it's very simple at the moment. I would align with your point Ross is, you know, it's an accelerator and from a brand perspective, the more we can get people to look at not only committing to one brand to maybe into a whole skew of that brand or a cross-category piece is really interesting. You'll probably see more and more of this kind of co-op petition mentality with brands. So the brand partnerships, how does that translate into ecomm and that food inspiration within that moment of conversion? If you can start to look at how you how your joint bid for attention, if you like, at that point of conversion, how your joint bid for inspiration and meal inspiration at that point of conversion. You're competing for eyeballs, you're competing for pounds, but 50 % of something is better than 50 % of nothing. Using that partnership mentality could see another kind of evolution within personalization within ecomm as well.    

Q: What are some innovative examples of personalized advertising that have notably increased engagement and sales for CPG brands? 

Ross: I think it's probably tricky in terms of pulling out specific examples for certain brands, but I'd certainly say in terms of who's leading the charge in this sense, I think it's similar to the retail media topic we were kind of just on would be around Amazon AMC and their custom audiences. For some of our clients, seeing some real success stories here and brands punching above their weight and overturning offline market shares in that sense, it's just all around the unparalleled granularity that you can go to and how far back you can retarget.  

I'm probably coming at this more from a retailer lens, but obviously there's opportunity for brands to get involved. You might have seen earlier on this week around Tesco going down the AI route with their loyalty programs. They tested this about nine months ago or so. I remember receiving it myself and it was kind of spend X amount on cooking sauces in a set period and you unlock certain amount of loyalty points, for example. The fact that they're going big with this later on this month suggests to me that that's worked quite well for them.  

There's certainly more and more of this happening in people getting on board with the data that's available and obviously utilizing presumably some AI behind the scenes to make that happen in a more efficient manner.  

Rob: From a brand perspective, relying on that opportunity to plug in that AI and again, that true understanding of consumer habits and events is really interesting. The ability to then plug that in as a personalization map, because not everybody reacts to true personalization, whether it's for me, Rob, or us as a household or us as a certain group audience persona, then sometimes that's enough personalization. Tech enthusiasts are constantly saying more personalization, I want more connection, more data given for more value return. I hear you, but for some people that's quite uncomfortable.  

When you're working in margins for a product, if the cost of the inventory or the opportunity to connect outweighs the profit you're going to get back long term is that of value. There's a fine line between going, this is brilliant, we want to be inspirational, we want to be innovative with our activations.  

Ross: Just to build on some of that, some of the pieces that we see around like smart passes, you imagine on your Apple wallet or Android, Android equivalent that you've almost got that personal connection with the brand whereby you can communicate through that, that you create an agnostic loyalty card system specifically with the brand. It might lend itself better to a DTC environment, but you could still push them for a QR code to work in a physical store or have a code that worked in that side of that environment. We'll certainly see more from a loyalty perspective of driving, driving that personalization throughout them. 

Q: What are the next big trends in AI and data analytics that could redefine personalization in e -comm for consumer goods? 

Rob: Everybody talks about connected home I think that's obvious you see that with the apps available to turn your heating on to set up any recordings through devices etc. you know the connotation of the connected home is obvious. The connection with retailers and the ability to understand the connected home from either arguably the big shop or the convenience shop is really important as well. If you then throw in wearables to that, so the connected environment at home, the connected individual and how you then sink back to a retailer experience. So, am I leaving home? Am I traveling to work? How am I being communicated with if I need food inspiration? for that evening or for breakfast or for whatever it may be, any whatever it may be. I might be going to the gym at lunchtime. I think this concept of connectivity and taking the data from home to store to wearables is a really interesting dynamic that could offer a whole new world for brands, for people, for retailers and media.  

Ross: My take on it was probably more around just how quickly and dynamic, I guess the shopping journey can be in that sense, whether or not it's truly much more personalization of what landing page that you arrive on, as opposed to having a handful of products that are picked for you, that the whole page is more bespoke to what it is that you buy all the other factors in the data from demographic perspective, that you would have a look at, you'd be considered for in that sense.  

It wouldn't surprise me if we see more dynamic pricing. Amazon's normally the one that's heralded for having prices that changes very dynamically in that sense, but I think we'll probably see more of that coming through and sure of not logging in people will be unaware about that.  

We've touched on it with kind of partnerships in the past was probably surrounded almost like bundling to that sense whereby you might cluster products together that may not necessarily always be seen together on that kind of algorithm basis and machine learning of what do people frequently buy that I think there's opportunity to increase basket sizes by utilizing the technology behind the scenes there.  

Rob: I think there's an opportunity to combine both of those. The concept for us as well is around the food inspiration piece and the impact on food usage and reducing food waste. The convenience of food mapping, food planning for busy households, consumers with less time. 

That spider web of making a shopping experience fun but easy and how you connect that through multiple layers is the next evolution. Everyone's trying to map that out and make it simple for everybody else, simple for consumers, inspire consumers to spend more, spend it on brands rather than on label, for example.  



What an incredible set of experience reflecting on personalization driven by data. Thank you to Rob and Ross for joining us in this episode.  

This is only the highlights! You can watch the full episode above or listen on Spotify here.

Don't forget to subscribe!