These Customers Are Costing You Money (And You Don't Even Know It)

Sean Weisbrot: Welcome back to another episode of the We Live to Build podcast. Customers are the lifeline of all businesses, and the customer is always right. But sometimes customers aren't great at knowing how to tell us really important things like how they found us, why they like us, what they want to see us develop next, and what we can do to make them stick around long term. That's why the marketing department of any company should be getting very serious about how to use data to learn what they can about their users, so those users don't have to fumble over their words telling us about their habits, behaviors, likes, dislikes, among other very important things.

Sean Weisbrot: Today's guest is Allison Hartsoe, the CEO and founder of Ambition Data, which helps companies use customer data to make better decisions about how they attract and retain customers. In this awesome episode, we talk about what is customer lifetime value. How can you increase a customer's lifetime value? The best way to segment customers? One huge reason low-value customers don't move up the value chain. How to develop an amazing referral strategy? The difference between loyalty and retention? How to employ different strategies based on whether you want to retain a user or deal with loyal users, what is the future of customers and data, and so much more. So, let's learn about customers data and how to make more money now.

Sean Weisbrot: What it is you're doing right now that makes you the right person to talk about customers and data?

Allison Hartsoe: Well, the piece that we do that's a little bit different as we triangulate what an entrepreneur might want from the data with what is actually in the data. And then we put a business eye on it. So, it's the combination of understanding technology, the data analytics and the business all coming together around what is the most valuable information that you should know about your customer base? Frequently that goes to customer lifetime value.

Sean Weisbrot: Let's assume for the audience's sake they don't know what lifetime value means. Can you explain that real fast?

Allison Hartsoe: Every company has a slightly different definition of customer lifetime value, but you can assume that it's basically projecting forward with the value of the base is. And if you really wanted to dig into the formula, then you could understand it through the measure of time. How much time do your customers spend purchasing from your company? The cadence between purchases and what is the margin on your products? What is the value of the dollar, and how does the value of the dollar change over time? It's a little bit of a complicated formula, but by the time you pull in all the features that matter, you end up with an outcome that basically says, am I pulling in customers that are increasing in value, or am I pulling in customers that are not increasing in value? That's a natural way of thinking about retention.

Allison Hartsoe: If I'm bringing customers back again and again and they are automatically increasing their purchase cycle, increasing the amount of money that they spend with. Then every time I increase that cadence, I'm increasing my customer lifetime value. When you grab all that data and you project it forward, you come up with a number called Customer equity that says, here's how much the value of your customer base is worth. If you take that number and you apply it to how much you spend in custome acquisition, then you can start to understand things like Kak to CLV ratio, which is, was it worth my time to spend money in this channel when I'm acquiring customers that don't eventually pay off, which is a classic question.

Sean Weisbrot: That's a fantastic explanation. Thank you. So, you take this information, you say, okay, I have a customer. I understand how much their value is to me over the lifetime of our relationship. How can you increase that value?

Allison Hartsoe: Yeah, that's an excellent question. So, we oftentimes stratify it into four different quadrants a customer is a high-value customer. Sometimes we call them platinum's people who are very profitable. They come back again and again. On the other end of the spectrum, you have low-value customers or sometimes we call them lead customers, who could be costing the company money every time you deal with them. But generally, there are a large group of customers that maybe bought once and you haven't seen them again.

Allison Hartsoe: In between we have these two areas of medium-low and medium-high, and the common misconception is to look at the quantity of customers and say, oh, I've got a high volume of low-value customers, I've got smaller numbers of middle tier and higher customers. It actually comes out almost like the 80/20 rule, where 20% of the customers are generating 80% of the business. But the quantity problem is, people commonly assume that they can just lift a small number of those lead or low-value customers and then generate a whole lot in revenue, or the reverse of that. They can sell even more to their high-value customers, who are already buying at a pretty frantic pace.

Allison Hartsoe: Now, in some cases, you can sell more to high value customers, but the sweet spot is the middle. The medium highs, which we sometimes call gold, and the medium lows, which we sometimes call iron. Those two tiers are where people have come in and purchased and maybe purchased again, but they haven't yet established a recency or frequency cadence that allows them to become a high-value customer. Or perhaps they were a higher value customer, but they have slipped down. So Gold and iron is the middle tiers are exactly the right place to push when you want to increase the value of your customer base. But if you never stratify the customer base, you never see these people.

Sean Weisbrot: You're basically saying, if you've got someone at the top and they're paying you, well, just kind of don't worry about them too much. And if you have people on the bottom, don't really worry about them too much. What I've learned is if you have someone who's bought from you, it's easy to get them to buy from you again. So, does it make sense to not try to sell to them more?

Allison Hartsoe: When someone first comes in and makes that first purchase, they usually end up somewhere in the middle of the pyramid. And by the sheer fact of making a purchase within a recent amount of time, we usually see them land somewhere in the gold or iron section. Over time, the propensity that they will purchase again becomes lower and lower and lower. So, someone who purchased from you last holiday season but hasn't purchased from you again in the meantime has a much lower likelihood of purchasing than someone who purchased every three months from you. Now, a lot of that does depend on what kind of product you have, but in general, you don't want to stop trying to get sales from people who have recently purchased from you. Nor do you want to ignore the lead customers or the platinum customers. It's just a way of stratifying the right kind of marketing technique for each.

Allison Hartsoe: So, with the lead customers, you want to be very cost effective with this group. So, if it doesn't cost you a whole lot to send emails, by all means go after that group with email. But don't try to target that group with paid search. On the other hand, for the platinum customers, that is a really good group to look at. Emotional poles, special features, special offers, special things that make them feel like, wow, I'm so happy I bought from this firm or encourage them to tell their friends. So, with the platinum customer group, you really want to try to push word of mouth, tell other people and create a great way for them to feel good about working with your brand, beyond just the fact that, hey, we have another product to sell you.

Sean Weisbrot: You were mentioning that with the people who are the lead strata, that you should be as cost-effective as possible. What I've noticed in the SAS industry, which is the industry that I'm in, and my co-wrote a white paper about what we found mostly related to wasted spending. We noticed that when you are of a free or low-tier plan on a SaaS product, you typically get horrible customer service because it's not cost-effective to invest your employees' time in communicating with these people, because they're not bringing back the financial value that you need to justify communicating with them. But that is one of the main reasons we found as to why people don't join a higher paid tier because they're not getting good service.

Allison Hartsoe: Yeah, that has to deal with payback. At the time. Let's say I take a group of low value customers in your SaaS model, and I try to give them some kid glove service. Then the analysis that would make sense is to say what percentage of that group eventually became more valuable. They upgraded or they felt special. They got this reason why they should stick with the company. They're not just tire kicking and doing transactional things, and not 100% of that group will ever convert higher. But if you test that, then what you might see is that perhaps one month of kid glove service, and that takes maybe five hours of a customer service reps.

Allison Hartsoe: Time will give you a certain amount of payback calculation. In other words, that person then has to stay with the service for a six-month period in order for you to recoup the amount of money that you spent on them. And a lot of companies will give you this kind of service because they've already done this calculation. They know that if you stay a certain amount of time, that they start making their money back, and then every month beyond that is just dropping to the bottom line. So that's the calculation that I would think about when encouraging a large group of customers that are not paying a lot to move up and then to gradually take, okay, what were those tactics? Was it five hours? Was it three hours? Was it certain things that they really needed? Could that be a manual. How does that then translate into, again, cost effective measures to help them move up a video series is another good example.

Sean Weisbrot: Yeah, I love the idea of providing content. It seems to me like the trend. So, there's several trends, one being CEOs being more publicly available, and the other is positioning your company so that you are educating people about your industry on top of educating about your product and your service, so that they think you are the most trustworthy in the market. This is kind of a method to escape the rat race of functions and feature differentiation. So, what would be the most cost-effective way to create content? In order to help convince these people to move up the value chain?

Allison Hartsoe: So, I'm not an expert in content creation, but in the examples that I've seen, I see a lot of video, and the reason I like video or even audio is the fact that you can play it at 2x speed and you could get through it a whole lot faster than a one-hour meeting. And a lot of times these questions are repetitive. So, I always think of that as really good content for just blocking and tackling. The other part of your question really gets to trust and generating trust, and that is not always something that comes across in a video. Sometimes it does, but I think that element is something that should not be underestimated by conveying the fact that there's a guarantee or even the story behind why you're doing something the way you're doing it. You know, you felt a certain pain, and therefore you're solving a problem and the benefit can be enjoyed by them as well. There are certain elements of storytelling that also help convey trust. And I think in the content creation side, which again, I'm not an expert in, I would tend to focus on the elements that are easily digestible by the audience, but also give a sense of who the brand is, almost like I'm relating to a person of which the CEO is oftentimes the chief brand conveyor. So, if that means that your CEO is recording the videos instead of your customer service rep, then perhaps for a smaller company, that's a good way to go.

Sean Weisbrot: You also had mentioned referrals as a way to help people get up the chain. So, do you have any specific strategies for referrals?

Allison Hartsoe: Referrals are really interesting. When you've identified the right kind of customer, then. Then the question becomes, where should I fish to get more of them? There's two different schools of thought. One says, I'm going to cast a broad net and I'll get some good fish, and I'll get some fish that aren't so good. Well, the reason that's a problem for entrepreneurs is investors are commonly looking at how many people that came through the door stayed with your brand. A company that is hot that people really love has a very high retention rate, you know, sometimes 90% or more. If you are casting a wide net, whether it's through referrals or through other marketing channels, then you may not be picking up the best fish. And that will hurt your valuation in the long run because it looks like you can't hold a customer.

Allison Hartsoe: Instead, a referral should be a very targeted strategy where you're looking at your gold or platinum customers, the ones who have already demonstrated that you are a fit. There's product market fit. They love your product. They're advocating for it. Things like that are worthy referrals, and you're much more likely to get power out of that group than you are out of just trying to get referrals from anybody who happened to, you know, click try. So, I think it's very important to be strategic with the customer groups that you're looking at, how you think about referrals and whether you are using a cast, a net strategy, or whether you are using a targeted fishing strategy.

Sean Weisbrot: So, you're saying don't allow every user to become a referrers, only go for the best paying customers to be referrers.

Allison Hartsoe: It depends on volume. So, if I had enough volume, if I had an inbound volume that was satisfactory and I felt like that was really good, then I would look at that inbound volume and say, okay, of that group, which ones are really enjoying the tool? Which ones are active in the tool? Which ones are? Are coming through now. Again, I'm using a SaaS model. In your case, that is a much more interesting group to get referrals from. Than somebody who, you know, tried it once and didn't do anything else. If you reach out to that person and ask for more referrals, well, they might give you referrals, but they might give you referrals to people who are equally unqualified, which then again takes your retention number, which costs you in the long run.

Sean Weisbrot: The way that you're describing it, it makes it seem like you would manually talk to them. But I think a lot of platforms are using a referral system where any user that creates an account has a means to automatically refer other people, and so all they have to do is enter in an email and the referral link gets sent. So are you saying to identify the user ID and then cut off the referral code so they can't see that there's a way to refer people? Is that what you're talking about?

Allison Hartsoe: I would look at the customer lifetime value of everybody who did that referral. I would hope that people who are actually executing a referral are people who are essentially promoters, people who already love the brand. But it depending on what kind of offer I put with the referral, maybe people are just sending it around and basically adding noise to the system, adding junk referrals just so they can get the benefit kind of depends on the strategy you put around referrals, doesn't it?

Sean Weisbrot: Definitely. So, what kind of a strategy should you put around referrals in order to maximize the gold people, you want customers that are higher value to give you more referrals, and you don't want customers who are one and dones or don't have any real brand affinity, brand loyalty for you to be pushing them to give referrals unless there's some angle that you feel is a really strong angle. And the angle here that I'm thinking of could be around technology. So, let's say that your tool is really good on the Microsoft stack, but it's not so good on the Google stack. And somebody comes in and they say, oh, well, we decided to leave a Microsoft stack and go to the Google stack. So, we’re no longer a good fit. Well, in that case, maybe they were a high-value customer, or they were a customer that was highly engaged. If you don’t know the CLV, then sometimes you can use an engagement as a proxy for how much somebody is interested in embedded in your system that has pros and cons to it.

Allison Hartsoe: So, for example, if I were heavily spending a lot of time in your systems, that doesn't necessarily mean that I'm satisfied with the systems. It could mean that I'm confused. So that takes a little more analysis. But in general, someone who is more engaged but the system no longer fits them could still be a good referral. Again, you're getting the marker of engagement to tell you at least to raise a signal that whether it's transactional in customer lifetime value or whether it's engagement, I've got one metric or the other to tell me this person understands me, this person engaged with me. This is somebody who maybe was a good customer. At one point, but now they're not. And that can still be an opportunity.

Allison Hartsoe: The thing I want to stress here is anytime we work with customers as a business is growing, there are a lot of one-size-fits-all strategies that are fine to use as you try to figure out who are the best customers and how should I be relating to them. But as a business gains a foothold and starts to get traction, you should very quickly move into a clear quantitative strategy that tells you where the long-term value is. It's critical because if you don't know where your long-term value is, then you stand in a place of having to acquire and acquire and acquire and pay more and more and more for people who don't stick around. And that's what happened to Blue Apron.

Sean Weisbrot: Yeah, I was reading about this. This happens to a lot of SaaS companies where they tend to overhype what they can do. And then the reality is, once you get on their platform, they can't do all of the things they said they could do. And then people start to drop off and they start to spend more to acquire more customers, and they end up running out of money, you know, because they can't get people to stick around. And so, a lot of SaaS companies go bust because of this.

Allison Hartsoe: I love what you said there because I personally experience that. And this is why I am so passionate about this particular analysis angle. I ran a startup in the dotcom era, and I remember our venture capitalists just pushing us to acquire eyeballs was the term at the time. And you know, everything was about how many eyeballs do you have? Well, that's just a vanity metric. It doesn't have anything to do with the long-term health of the business, but it works for investors because they often are calculating the worth of your company by the top-line revenue.

Allison Hartsoe: So, let's say that you acquire a whole bunch of customers just before you're ready to go public. And, you know, the numbers are up and up and up. Well, that generates some nice payouts for your investors. But guess who gets stuck holding the bag? You know, the founders of the company, the owners of the company and everybody else who is there after the company goes public, when the reckoning happens and suddenly it looks like your customers are defecting in mass numbers because you did a whole bunch of cheap acquisition. So, this always irritates me that investors are not always aligned with the best interests of their companies. And smart CEOs, I think, are getting the message now and are a little bit better about pushing back with how they know their businesses should grow.

Sean Weisbrot: I totally agree with that. I've heard plenty of horror stories about investors, but this isn't about investors. Obviously, you're talking a lot about customer data, even though we haven't really mentioned the term data. The only way to be able to understand the acquisition cost or the lifetime value or what where they sit in the value chain in terms of different user segments. So how can you use data to identify the right channel? To be promoting in order to get those highest value customers.

Allison Hartsoe: There are a couple ways to think about the collection of data in order to make your systems sing. So let me start with customer identification, and then I'll give you a little bit of an insider tip about something else that really adds a lot of power to your data systems. In customer identification, the easiest and most direct is when somebody actually converts and they fill out all their transactional information, their name, address and phone number. That is your clearest signal of who the customer is. You can use back matching techniques to take all of their digital trails, everything else that they did on your own sites beforehand, and match that login ID with previous visitor ID in order to understand how did they move in different ways before they converted. So, you can enhance some of your customer identification with back matching.

Allison Hartsoe: Another way you can enhance customer identification is through the email or a campaign code. It's common on email. Like let's say I come in and I sign up for a newsletter. Every email system actually has a customer ID system inside that email address, and sometimes you can surface that along with any other link data that you might put in the email. That again gives you the customer ID and the traction of what's happening with that ID over time. But ultimately, all you need to calculate customer lifetime value is who is the customer and ID. Sometimes people are using email addresses. I don't recommend it, but sometimes that's all you have the product that was purchased, the value of the purchase and the date. So those four pieces of data will allow you to run customer lifetime value models. Now when you try to get into, well, where should I fish? That's where you need the campaign codes to come in that I alluded to just a second ago about the email IDs.

Allison Hartsoe: So, when you when you run a campaign, whether it's email or paid search or an affiliate campaign or a social media campaign, you're able to take any link that you're sending out and put in a series of it's called query string parameters behind that. That will tell your tracking systems where that person came from and any other bits of information that you want to convey. So, one way to get a lot more power out of your customer data is to tell the system, what are you doing with that campaign? Was it a retention campaign? Was it a new customer campaign? Was it a referral campaign? Because a lot of times when you don't put that. And all you have is the referral source. That is a whole herd of problems when it's just the referral source, which I won't go into here, but it's much, much better to use a campaign code. It's stickier, it's more powerful, and you can add in layers of information that eventually give you more insights about what kind of tactics are working for you, and what might not be working as well.

Sean Weisbrot: Yeah, my COO, he was trying to figure out how he could put together all of our systems so that he could have as much information about a user before they even get on to our system. So, this campaign code, I'm not sure if he thought about that. He probably has. I'll mention it to him for sure. So, I appreciate the tip.

Allison Hartsoe: Not just are we setting campaign codes, but are we putting in valuable information we would use for analysis.

Sean Weisbrot: So, this campaign code is like embedded inside of MailChimp or something like how do you create a code?

Allison Hartsoe: It's an analytics tracking technique. So, most companies who are in the growth phase would be using, say, Google Analytics for tracking. And in Google Analytics you can set up a campaign code structure. You don't set it up inside Google Analytics, but you say, okay, every time we send an email and we use specific links that we set, we're going to code the UTM source medium in certain ways and the campaign name in certain ways that allow us to say these things always grouped together. So maybe I want to separate my paid social from my organic social. Um, that could be an easier thing to do through your campaign codes coming through. And then you can split all the information again going backwards through that back. Matching it basically gives you all these little tendrils that you can pull on for analysis. Does that make sense?

Sean Weisbrot: Yeah. So I have seen the words UTM in a URL before when I shared an article from Business Insider.com to one of my friends. So, you saying that the UTM is where this code exists?

Allison Hartsoe: So you put it in the link manually. So, let's say my link is dub dub dub. We live to build comm. If I'm going to put a query string parameter behind it, then I'll add a question mark which tells the servers that I'm picking up a piece of information, but it is not the URL. It's not a destination. For the purposes of loading a page, the information is passed into the tracking system after the question mark, but everything before the question mark is the address you would be sending someone to. So, if you were sending them to a specific landing page, you might say dub, dub, dub. We live to build .com slash landing page and then slash question mark and then UTM equals source and then the source value UTM equals medium and the medium value UTM equals campaign name and then the campaign value.

Allison Hartsoe: So, these are the different things that you can add. To your link whenever you're creating a link. Now this becomes a real bear. If you're trying to do paid search with that, nobody in their right mind wants to label paid search. So, this can to Google and the other tools have auto tracking and they will automatically encode different pieces of your campaign name and your keywords. So, I don't recommend it for paid search or anything that's high volume. I do recommend it for your strategic campaigns like an influencer campaign or anything that has to do with something you're trying out and you want to see how it works.

Sean Weisbrot: That's fantastic. Can you really quickly explain what UTM means?

Allison Hartsoe: It actually comes from urchin so many, many, many years ago before Google Analytics. What is this? 2004? Google acquired a company called Urchin Technologies and Urchin became Google Analytics, and UTM simply refers to the urgent technology code that they were using prior. Now, as Google changes and goes forward, they're changing into Google Analytics four, which is a different system. It's designed to pick up more data, more information, respect PII. It's a whole different system, which I would be hard-pressed to go into exhaustive details on. The old system for campaign tracking is still available. It's still there and still avidly used by many people who know how to push value into your data systems, and it is a common way to glue information across multiple sources.

Sean Weisbrot: I just went to Business Insider now and tried to create a share link. At the end of the URL there is a question mark UTM source equals copy link and UTM medium equals referral and UTM content equals top bar. I've seen it before, but I. I either ignored it or I would delete it from the link before I sent it to people because I, I just, I wanted the link to be cleaner.

Allison Hartsoe: Let's use the example that you just called up. The information there is about where somebody clicked to do the share. That is not necessarily the kind of information that you might want to understand about your customer base. You know, the fact that you clicked a top nav is a UX question. The fact that I was trying to get you to invite new users or to offer a 10% off or something like that, that never appears in the analytics, but that might be a better use of that space. And you can you can dogpile things in there. So, in the campaign name, you can put dashes in between, don't use a question mark and don't use an ampersand because those are reserved for other actions. But you can use just a dash in between to help you again feed more information into that element.

Sean Weisbrot: So, what is the future of all of this look like?

Allison Hartsoe: You think privacy is a big deal and it is very much on the mind of every company. Bunny that is trying to use customer analytics data to better understand who their customers are, what a good customer looks like, what they buy and where they go. And I think the I think the way we end up managing privacy will be influential long term, three, four years out. We may have wholly different systems for how we think about privacy, how people give permission, but we don't have that today. So that's something we're definitely watching a lot in the future. Is how is privacy affecting the rights, the regulations for how data is used?

Allison Hartsoe: The rule I commonly encourage the companies that we work with to subscribe to is what's called the Sunshine rule, which is if the way that you're using customer data were to suddenly become public, would you be embarrassed? Would this cause you concern? Because if the answer to that is yes, then you probably shouldn't be doing it. But if the answer is no, you're probably safe, because I think there's going to be more and more customers that have rights to their information, but no one has rights to understand how your business operates, at least not at that level of depth. So, they'll always be a little bit of back and forth. The other thing I think about a lot in terms of long-term value is today, when companies use AI or they use marketing automation, they're typically doing it in broad clusters. So, they're saying everybody who looks like this, you know, give them that view or give them that answer, we've all experienced that. If you've ever, you know, tried to apply for a loan at a bank and you didn't have exactly the right information that they wanted in exactly the way they wanted it, um, you know, maybe you get shut down or, you know, credit cards oftentimes will flag fraud, um, using the same kind of big, you know, rule-based systems where I think we eventually go is a layer of hyper-personalization.

Allison Hartsoe: And that reflects each person is their own model. It's their own trend line, their own set of activities. And what is normal for them may or may not fall into a broader cluster. But this age of clustering, people buy, you know, what product did they buy or their age or their sex or their race? These are throwbacks to the old demographic marketing systems, which is what we used. That was all we had. So, today's more modern methodologies, we'll get more and more granular until eventually we treat and understand customers like the individuals they are.

Sean Weisbrot: Well said. Thank you. So, what's something I haven't asked you yet that you wish I would ask?

Allison Hartsoe: Ask me about the difference between retention and loyalty. That's something you haven't asked.

Sean Weisbrot: Okay, go for it. What's the difference between retention and loyalty?

Allison Hartsoe: People use the term loyalty all the time as one generic term, and a lot of times it brings to mind the idea of a punch card. Or, you know, how can I get you to come back or join our VIP loyalty program? But loyalty should really be segmented from retention. And this is why when somebody comes in and they buy the first time, that is a wholly different path as a new customer, then somebody who's buying the second time, but when they buy the second and the third time, they still have not necessarily decided that you are the company for them in that particular category. So, it's not until they make additional transactions, three plus, that you start getting into the realm of loyalty and by breaking out retention activities, which is encouraging someone to make that second and third purchase from loyalty, which we then can see a little bit of what strategies we might use for our loyal customers that could be wholly different from people who still have one eye on our competition.

Sean Weisbrot: What I got from that was, what you need to do to get someone to look at you is different from what you need to do. When someone has already proven themselves to like what you're doing, what would the difference be? What should the difference be between retaining someone and making them loyal?

Allison Hartsoe: Well, retention is a transactional activity. What I want to happen when the problem is retention is I want to encourage you to buy a slightly different product of the product line. So maybe you came in and you bought a shirt, and now I want you to buy a pair of pants. That is transactional retention. But when we get to loyalty, it's an emotional pull. We're pushing in the brand element, the brand love and at loyalty. The element that you're playing with is here's something that's just for you. I think it's Nike and New York City has a special floor for a certain type of high-value customer.

Allison Hartsoe: Starbucks also does this really well when they have the Stars program, and there are different levels that you earn in the Stars program, and they entitle you to different privileges. That is starting to be less of a transactional, and they're moving into more of the emotional hook. Hotels are really good at this. Airlines to a lesser degree, because unless you fly all the time and maybe you do. Sean, I like to think more about the hotels and all the amenities that they entitle you to when you are in that higher level. That's a love and loyalty move. That's not a transactional move.