This article is based on a presentation delivered during Neoteric AI Breakfast London.
Personalized customer journeys are impactful. Whenever you go online, use an app, or click an ad online, new touchpoints are created. With so much data available, companies can interact with customers, and not just any customers but the right ones, in a variety of different ways.
So why should the customer journey be personalized? After all, it’s still the same process, isn’t it? You make a customer aware of your brand, you convince them that your brand is right for them, and after conversion, you want to retain the customer. Sounds simple, but it’s not easy.
Is your marketing right?
Just think how many times you received an ad of an item you’ve already bought. Or a new customer campaign from a company you’ve been a customer of for years now. Isn’t that frustrating? How come that having all this data about me as a customer a company still can’t figure out at which stage of my customer journey I am? They should be able to have it figured out and provide customers with a more personalized approach. Fragmented views of customers are bad for business. It’s easy to irritate people with targeted ads. If you look at a pair of shoes online (or even buy them), the ads for the same pair will be following you online for the next 6 weeks or so. No point in that. Why not, instead, target this person with followup advertising saying “Hey, you’ve got these shoes, how about you get these jeans – that’s a great match!” or a simple “How do you like your purchase?”.
Using identity, you can retain your customers to make those most valuable even more valuable. Let’s say they’ve been your customers for a while now, it’s good to have loyal customers. It’s a “sure” revenue (though nothing ever is a 100% certain), but you can increase their lifetime value by making them aware of your other offerings. Let’s take the example of Apple. You may have bought the latest iPhone, but hey, a new version comes out. Wanna have it? Or you might need a MacBook, too. That’s when awareness and consideration come back in and the lifecycle starts all over again.
However, sometimes knowing your customers across different channels is not that easy. Actually, it’s really hard to know that the customer who visits your website is the same one who walks into your store. There is no uniform connection across all the different platforms: CRMs, point of sales, call centers, loyalty systems… So how do you connect these data points to have a fuller picture of your customer? When you have a customer ID and monitor the activities across different platforms, you see anonymized data – so it’s customer 1234, not Jane – but you still know that there’s a customer who’s considering your brand and you can address them.
The solution to that is to build an identity infrastructure. Using identity resolution, we can seamlessly connect all the information: that a given customer walked into your store, and then also browsed some particular items on your website.
What data makes up the persona?
Look at the example above: we’ve got Liz or Elizabeth. She used to be Elizabeth Jones and after she got married, she changed her last name to Wilson. Her friends call her Liz. You can see that we have many different touchpoints in the ecosystem. She has a personal email address and a work email address, she has a cell phone, she watches a connected TV, browses websites. How do we know that it’s the same Liz interacting with us across all the different channels? How do we know that Elizabeth Wilson is the same person as Liz Jones? Identity resolution helps overcome this challenge. By building identity resolution and building a graph of the client interactions, we’re able to tie a single identifier to this individual, so no matter where Liz interacts with us across the channels, we always know it’s Liz.
Building this graph, we start with the known touchpoints. If you have a CRM, you know that you have an Elizabeth Jones who lives in Texas. You also have a different Elizabeth Jones or Elizabeth Wilson with a phone number, then maybe she has a Texas area code but you’re not sure that it’s the same one. So you use an identity resolution provider to take all those different touchpoints and tie them to Liz. These are known identifiers – the ones which we can ask people about – if you ask someone “Hey what’s your email address?”, or their phone, or home address, they naturally know the answer. So we take that graph and we de-identify the data to protect consumer privacy, and we create an anonymized graph while maintaining all the key data. All of these things we know about the customer will influence we market to them online. Although the data is anonymized, the data is still tied to a particular key. This allows us to publish content targeted to her persona, not her person.
Then, there are also anonymous identifiers – the thing with those is that if you ask someone about them, they just won’t know. Try asking someone what their cookie ID is. Any success? So things like that are generally classified as anonymous identifiers and many people don’t even know they exist.
Identity resolution helps to resolve different identities across different touchpoints and solves a lot of problems in the ecosystem: validate records to look for your active customers, narrow down the individuals to target with your marketing budget.
But then, if Liz used to be Elizabeth Jones and now she’s Elizabeth Wilson, are these two different people? No, we don’t want to create two personas, we want to have the most accurate and recent representation. A similar thing happens in over-consolidation: John Junior and John Senior may be one person in your base. However, John Jr and John Sr are probably totally different people. One is a 50-year-old executive, the other one is a 24-year-old college student. You don’t want to target them with the same ads. With the right identifiers, you appropriately target or retarget people with the right ads and messages – like “welcome back” or “we missed you”.
Now that you’ve connected the identities and you know that this is Liz, what do you do with all of those connections? By providing or attaching an identifier to all those different connection points and identifying a single individual, you can now interact with different data providers and data vendors. You can append data: What car do they drive? What political affiliation do they have? What’s their demographic data? It’s the most common to append attributes like general income or age. And knowing these things can really help people understand who their customers are and to get to know their lifestyle. You can also source data like geolocation, online behavioral data, weather information where they live, etc. This combination of various data points helps you understand your customers better.
By taking your data and appending complementary second-party or third-party data, you can see where data overlaps. Let’s say that Tesco wants to see how they overlap with Coca Cola. Even without sharing who their clients are, they’re able to compare their clientele because they have a key that speaks the same language. This gives them the insight to say: We share a lot of clients, maybe we should do a joint marketing effort. Or the contrary: We don’t share a lot of clients, but should we? Maybe we could do something to drive more customers into our store?
Read also: How to increase revenue from your customers using predictive analytics
Looking for your perfect customer
Let’s say you’re a company that has a small percentage of very high-value customers. You can append this data and analyze it to say what these clients look like. Maybe when you append data to your identifiers, you’ll be able to say that a huge percentage of the base of your clients are 30 to 35 years old females who like yoga. You can use this information for additional customer acquisition, without spending a lot of your marketing budget to market to males who are over 50. It helps target lookalike audiences that are more likely to interact with your brand. Find more of your best customers.
Another important element is to measure channel effectiveness. You can have clients contact you in a number of ways. They can interact with your website or app, they can email you, they can call your customer support center. How do you track which channels are the most effective? By putting identity resolution across all this data, you’re able to see which channels perform better and interact with your customers in a safe, GDPR compliant way. So without seeing the personal information, you still get to see the demographic data and attributes that make up this persona.
By Erin Boelkens – VP of Identity Engineering at LiveRamp; with over 10 years of identity resolution experience, she leads a team of creative and innovative engineers working to provide industry leading identity products that span both the online and offline channels.
Learn how to use predictive analytics to drive business results and revenue