CREATING THE PERFECT CUSTOMER PROFILE

Key Takeaways:

  1. Just because you have access to real-time data on things like clicks and opens that does not mean that you can neglect your customer profile.
  2. To create a customer profile you need 3 different dimensions of customer data Behavioral, Demographic, and Attitudinal.
  3. Customer profiles will change over time, this change can be driven by some kind of internal factor like changing your price point or it can change because of some kind of external factor.

Below is a lightly edited transcript of Episode 29 of the Inevitable Success Podcast with Damian and special guest Stephen Yu. (Listen Here)

Transcript:

Damian: So here we are back with Stephen Yu at BuyerGenomics Headquarters in New York City. And actually, we’re going to talk about an interesting topic today. But, we had a new addition to our marketing team this week. We have Luna, which is our head developer’s puppy, running around. So I thought that was pretty cool, never had a –

Stephen: Company dog?

Damian: A company dog.

Stephen: It’s a morale booster for sure.

Damian: Yeah definitely. So anyway we’re going to talk today about probably the most important thing to every business, which is knowing their customer. Because if you know the customer, then you can really be an effective marketer – you know who you want to speak to, how to speak to them. Stephen Yu, what is the best method that you have to know the customer?

Stephen: Well you’ve got to look at the data. So let’s go back a little bit. Looking at the data is what we call analytics right? So you refer to about what, eight episodes ago, remember we talked about different types of analytics?

Damian: Yup.

Stephen: We talked about dashboard BI reporting – that’s about hey, what’s going on, I dropped this thing and what happened to me. It’s all about companies right saying, “How am I doing?” – that type of thing. What was the second one? The second one was descriptive analytics. What is that? That’s about the customers now. The third one being predictive analytics, the fourth one being optimization. Let’s go back to the real basics of descriptive analytics – that is okay, so I know who’s been clicking, who’s buying, and all that – great. And so you see all these reports coming in, what are the clicks we’re able to open, great, you know all these things. But that doesn’t mean that you know who’s behind those clicks. Are they old, are they rich, are they being a good customer, high-value customer, or what are they? So digging deeper into who they are, is the beginning of the customer profile. I’m sure you’ve heard this word many times.

Damian: Customer profile.

Stephen: Exactly, the customer profile is an age-old thing. It’s not anything new. And why is it important now? Because a lot of people are neglecting to do it. We rely so much on the real-time data of, “Oh yeah I dropped this email, I did x, y, z clickthrough rates, I’m doing fine.” That is very important. That does not negate the need to understand how the customer looks like.

Damian: Right. I’m sure you’re going to get to it. But, I’ll ask you directly – what are the different ways to profile a customer, or what would you say encompasses a customer profile?

Stephen: So and also we talked about seven episodes ago, we talked about the dimensions of data that you have to look at and we talked about three different dimensions. Let’s remind ourselves –

Damian: I think what you’re saying is you have to listen to every single episode.

Stephen: No, that’s why I’m repeating myself because I know that nobody is going to listen to every episode. So what are the three dimensions? One is, OK what have you been doing lately? What’ve you been buying, clicking, viewing – whatever it is, all this purchase history, browsing history, all the things that people do?

Damian: It’s behavioral.

Stephen: Behavioral data. The second part will be demographic data. What do they look like, how much money do they make, do they have two cars or one car, or what kind of a house do they have – is that a rental, or owner, what’s the value of the house, do they have kids? All those things. How old are they? By the way.

Damian: Right.

Stephen: So those would be demographic.

Damian: Got it, so demographic profile.

Stephen: Exactly.

Damian: OK.

Stephen: The third will be attitudinal, just because you are rich and do certain things doesn’t mean that you are automatically conservative for example, or more liberal. You don’t know that sometimes. You have to stop and ask sometimes – that’s attitudinal. What do you feel about, your attitude about certain brands, about certain ideas? That is not a direct reflection of what you’ve been doing or what you look like. By the way people have a hard time guessing if I’m conservative or liberal because I live in a very conservative town and I may be fiscally conservative for example, doesn’t mean that I have different ideas about, I don’t know I mean there are so many topics out there these days and really hard to wear your heart on your sleeve these days. But the point is people have a very different attitude about certain things and if you want to find out about, okay so what do people perceive our brand – as what? Is it a luxury brand? Sometimes you have to stop and ask.

Now that’s very expensive. In the old days, you’d have to put together a focus group or do the primary research and all that and they still do that if you have deep pockets you can do that. But thanks to social media, there are so many ways to listen to these people and say, “Oh yeah the perception of Sprint is going negative,” or – you call that a sentiment analysis and whatnot. So yeah, that will be the third element which is the attitudinal data profiling. So going back to how do we do this? Well, it depends on what kind of data you want to do and what I recommend is start with the data that are consistently available and what are those? You have behavioral data sitting in your transaction history. You can break it down. When you break those down, you’ll find out that, “Whoa, I thought that I deal with a homogeneous group of people buying expensive things from me,” and you realize that “Wait a second, we’re luxury brand but a lot of people come in and buy some souvenir or some small item and maybe it’s a gift, maybe they are travelers, who know at this point we don’t know. But at least behaviorally we have a very dichotomous group of people and I see this all the time in the behavioral data profiling, that just based on their average order size – days between transactions. How long ago did they buy, are they a one time buyer or multi-buyer, or you know sometimes you look like a multi-buyer but if you look at just one year, I don’t come back that often, then you have to look at days between transactions. If you look at all these things you realize that, you know what, I’m not dealing with one type of customer. Sometimes you have three or four or five different pockets of customers.

Now there are ways to do it with a bunch of reports, by looking at all these parameters and say you know let’s make an assumption that maybe you’ll just create what we call newly acquired customers, or high-value customers and all these guys, and start looking at those things and to find that you know what, we have whales who show up once in a while and drop a lot of money – I’m just making it up but these are all true in a lot of cases – and then I have a lot of habitual buyers who visit a lot but they don’t buy or they buy a lot of small items. So in this example alone we have three different, you know, very distinct groups of people by doing this profiling using behavioral data. The only hard part is that you have to really build what we call the 360-degree view of the customer using the behavioral data and we talked about this in an episode as well. I remember I talked about it two episodes ago, that you have to convert this behavioral data into the descriptors of customers. Now if you did that then by looking at those numbers you can determine that, oh yeah I have different pockets of customers. So that’s one way to do it.

So what’s the second way to do it? This is a more stable way to do it because now we are talking about very – how do I put it – a standardized set of data called demographic data. We have a lot of reputable third-party data vendors who sell these things. If you have some set of customers you can send this file out, if you have some kind of PII they will append this information to your file – that’s step number one. And they append like, I don’t know, 2 to 300 variables easily: income, age, number of children, number of rooms in the house, owner, renter, value, stock portfolio value, all these things. The number of children being a really popular one by the way – educational level – all those things are available. What kind of occupation – are they blue collar, white collar or deeper than that. Now you have an issue with too many variables. So what you do now is, you run a series of reports to find out that, ok so my behavioral profile says I may have three or four different pockets of customers, high-value, high-frequency, recent, newly acquired, dormant, inactive whatever it is right? We have those things in our toolset off the bat anyway.

Now you compare those segments against all the other things that I just talked about: income, age, number of children, and housing value and all those things, then you look for differences. So are high-value customers really richer than the other guys? Or is it just something else? Is it about the region? Is it about the number of children? Is it about single parents? Or is it about not having a child? You know that type of thing. Until you look at this data you just don’t know. So what is a key takeaway? This customer profiling is not a one-dimensional thing at all. You have to do it in multiple steps. You’ve got to define the behavioral segment first, imaginary or real. You’ve got to start with the recent guys, high-value guys, and all those things should be done first. Then you look at them against all the standardized set of data and say that you know what, if I really want to go after high-value customers for the acquisition or for reactivation for whatever reason, then you know that you know, I have a higher probability of going after the homeowners or women over a certain age, or you have a much better idea of who they are. That’s the basis of profiling.

Damian: So you said something in there that was interesting – imaginary or real – what do you mean by that?

Stephen: Because you’ve got to make assumptions. The scientific approach starts with the hypothesis that’s imaginary, then you prove the point.

Damian: Got it.

Stephen: So in other words, in the beginning, you imagine things – we’re all marketers right, so let’s imagine like a marketer. It would be great if you know that, oh yeah I want everybody to be a high-value customer, that is imagination.

But when you look at the data you go, like what is high-value? Is that about high-frequency, or buying multiple categories, or spending a lot of money? But if you spend a lot of money in one transaction, is that high value? Those are not proving your point by looking at the data. And then you find out, oh well you know there are multiple types of high-value customers, do we all deal with them the same way? Then you sit down and say you know what, let’s compare these groups using the set of variables that are available for everybody, which is demographic data, to see if they are really different. That would be the, beyond the hypothesis sizing the whole situation and you’ve proven the point and now you can act on them. Now I will be like, oh yeah yeah yeah, so if I have an acquisition plan to find more of these high-value customers, this is where we find these people, this is where they frequent, this is what they look like.

Damian: This is actually something you and I spoke, you know, kind of offline about recently, which is you run a customer profile.

Stephen: Yeah.

Damian: It’s almost like a snapshot, but what do you do over time?

Stephen: Ah that’s a very good question because the profiles do change, and the one thing about change is something that brand managers really should remember because brand managers are very ambitious people. I’m selling well but I want to go into the high-end market. By the way, we do have a customer who really drastically changed the pricing point of their own product and went really high-end and they successfully did so. Now, that’s a very risky thing to do, isn’t it? And so your customer profile changes sometimes because of what you did, and we talked about the chicken or egg discussion just now too.

Damian: Yeah.

Stephen: Or maybe over time they changed, that hey I don’t know, I mean maybe you were selling a rock n’ roll memorabilia. Now, where is rock going these days? They’re going the way of jazz. And I think that rock will disappear, but I’m sad to hear this because I’m a rock player myself. But you know, things change and all these rock stars are dying and nobody’s replacing them, let’s just face it. So, if you’re in a rock memorabilia business where is your future? Your customer base is dying off or stop buying. So, the changing world changed the customer profile. So, it’s a force that is beyond us sometimes. But what’s important is, is this initiated by you – that you started selling more expensive items and you are going after more higher-tier customers and that’s something that you did? In that case, you’ve got to draw a big thick line in the timeline and say, you know what let’s keep old all customer profiles that we ran before that thick line, let’s do it again after that, and let’s watch it over time. So now you get into the discipline of having to watch this over time. A lot of people don’t do this and they make excuses – oh yeah if only if I had a freeze file. You know what, this is the age of big data, when are you going to freeze the whole thing and just wait for you to study later. Please don’t even plan for those things. All I’m asking is have a very methodical way to profile your customers the way we described here. Keep it, keep it as a report, not as a whole data set.

Damian: Right.

Stephen: Keep it that way and then keep doing it in a similar or the same way and then start comparing to the older report.

Damian: Now what would you say your snap, you know getting a profile of customers that you acquired, you know, let’s say this year and you run that report once a year, once a quarter, what do you think?

Stephen: Once a year would be good unless you did something drastically different yourself.

Damian: Right that makes sense.

Stephen: This is a B2B example, but let me bring it back from my recent memory before I joined this company I was in a consulting company, I have plenty of examples. There is a software company who did all this interesting profiling work over time and they watched the behavioral data, firmographic, demographic data that all the time. You know what this company did – they changed the business model itself. Like, oh you know what I’m going to go into the licensing model instead of selling software. Guess what that does to a whole base – it’s like a fundamental change – it’s almost like a revolution in terms of customer profiling. So not only does it change the whole study period of certain things to follow, you really have to keep in mind that this is the change that happened that month, that year. Just like if you look at any kind of a financial analyst, they know the black day that everybody went bankrupt and all that that, some days in October in 2008 right? If I talk to, by the way, I talked to some marketers of the department stores in New York, they know that date because every profile that they did, all the behavioral profiles that they ran changed almost overnight because of a seismic shift that happened when the market crashed. It’s not the worst, people used to buy two things, three things – they didn’t stop coming in, but either they just browse or buy only one thing. Those things are not your fault, it’s not something you did, but you cannot rely on the old profile anymore.

Damian: That makes total sense.

Stephen: You just have to do it.

Damian: Yeah. Well, it’s interesting because when we speak to, when you speak to enough brands there’s always some percentage of them that are in the process of evolving their brand. And it’s a surprisingly high number right?

Stephen: That’s right.

Damian: It’s, I would say maybe a quarter of the brands that, you know, you end up speaking to are in some form of delta or change in how they think about their brand. You know, and actually, I know there’s a couple of things – you used one where they went into a higher price market and we know there’s plenty of examples where brands have done the opposite.

Stephen: Absolutely right. In fact when we see some explosive growth in new customer acquisition in our report – and by the way, what is our job here at BuyerGenomics – it is to consolidate data so easy so that you don’t get lazy about doing these things on a regular basis, you know these changes. Okay, so example: we know that oh, whenever somebody has explosive growth in very short time we know that it’s because of something that they did. So as a, you know, sane or disciplined analysts what do we do? We stop and ask what did you do? Did you just change the way you started giving discounts away? And they go, “Oh yeah we started using them, you know, 20 percent discount on it.” First of all, how did you decide on the 20 percent you should have talked to us okay? Because we could tell you what banding is most effective, okay? Instead of like changing that to 20 percent overnight and start attracting barnacles. That changes everything about your customer base so you grew, but what is your future potential here? You were just attracted a lot of bargain seekers. Now, this kind of conversation is a continuous discussion, it’s not some, oh I ran a profile years ago and I know my customer base. No, you don’t. In other words, whether you’ve started to change or the world change upon us, doesn’t matter. You’ve got to keep watching.

Damian: I love it. Well, we’ll end it there – ‘til next time.

Stephen: ‘Til next time. Thanks so much.

Damian: Bye. If you enjoy today’s episode we ask that you please leave a rating and write a review. Or better yet share it with another marketer. Be sure to subscribe to the podcast for new episodes. If you’d like to speak to someone about any topics covered in today’s episode please click here and start a chat with the BG team today.

2018-11-02T20:57:49+00:00