Not all customers are created equal. The best subscription businesses know the power of customer lifetime value (CLV). They optimize around understanding who their best subscribers are, and then deepening the relationship with those best subscribers over time. Neil Hoyne is an expert at this. As the Chief Measurement Strategist at Google, as well as a Senior Fellow at The Wharton School, Neil helps people use data to win their customerâs hearts. Heâs written a new book Converted: The Data Driven Way to Win Customersâ Hearts on this topic, which I read on the beach while on vacation. Itâs a book about data thatâs entertaining and engaging enough to read on holiday, if you can believe it. We recently spoke about how to measure the full value of each relationship, engage in an ongoing conversation with your best subscribers, and perhaps most importantly, how to find and win new subscribers just like the ones you find most valuable.
—
Listen to the podcast here
How to Find Your Best Subscribers and Develop Relationships That Last with Neil Hoyne, Google Chief Measurement Strategist
Not all customers are created equal. The best subscription businesses understand the power of customer lifetime value. They optimize around understanding who their best subscribers are and deepening the relationship with those best subscribers over time. Neil Hoyne is an expert at this. As the Chief Measurement Strategist at Google as well as a senior fellow at the Wharton School, Neil helps people use data to win their customerâs hearts.
Heâs written a new book, Converted, on this very topic, which I read on the beach while on vacation. Itâs a book about data thatâs entertaining and engaging enough to read on holiday if you can believe it. We spoke about how to measure the full value of each relationship engaged in an ongoing conversation with the best subscribers, and perhaps most importantly, how to find and win new subscribers like the ones you find most valuable.
—
Welcome to the show, Neil.
Thank you so much for having me.
I want to talk about your excellent new book, Converted, but first, I wanted to start by asking you what it means to be the Chief Measurement Strategist at Google.
The easiest way to understand it is itâs less looking inside at our own data and more are looking externally to customers that are using data and saying (We work with 16,000 customers who would be considered our large advertisers.) How do they take the same sets of data, sometimes competing for the same customers and make different decisions? Out of that segment, when you look across companies, across verticals, across different countries, what are the patterns that emerge in terms of best in class and worst in class? The ultimate question we try to answer is, âHow do we scale and repeat those for any other business?â
These are advertisers advertising on Google. Youâre trying to help them, not just to use Google products but to be more thoughtful about how they connect with those customers.
It would be nice if they spend more money with Google but that is a disingenuous goal when it comes to data because people are always wondering if you have your hand on the scale. In the end, we hope that they find growth using data. If our products are built correctly, they should align to help accelerate that growth but they are not necessarily part of the pitch. If we find growth through email and we donât have an email product, we will still pursue that if it helps the company.
âThe biggest challenge when building fluency around data is the reluctance of people to admit that they have no idea what you're saying.â - @neilhoyne Share on XVery few of the CEOs and CMOs that we talk to say, âWe donât have enough data.â They usually say, âWe have a lot of data but weâre not using it. We donât know if we have the right data. We donât know how to best harness this data. Itâs overwhelming for us.â What do you do with those kinds of comments that you might hear from them?
You hear data is the new oil with the idea that itâs a commodity that you can collect. Eventually, somebody finds the data in it. All engineers and analysts cringe. I had one CMO and I said, âWhat are you doing with all that data?â She was puzzled when she looked at me. Sheâs like, âWe are going to hire data scientists to unlock the value of the data.â Most companies lack that middle point between the data and the actual money. Whatâs a strategy that you can use to grow the business? Dashboards and reports help to consolidate all the data but they donât tell a good story about what youâre supposed to do next with it.
How do you get that story? Where does the story come from, and whoâs responsible for finding it?
Itâs two groups or functions. One is there needs to be some type of strategy of the business going forward that the data can reinforce or guide. Thatâs a lot of where our work sits. It is to say, âHereâs a data set about how to look at a customer or how to look at a particular moment in that business. What opportunities come to jump in front of competitors and grow share?â
On the other hand, you also need people inside the business that can translate that idea, that strategy, to something that every function can understand and participate in. You may work with a frontline salesperson where as soon as you mention a confidence interval, they check out and say, âThis conversation is not for me,â which means theyâre not going to participate in the strategy no matter how great and understand the data no matter how wonderful and brilliant it may be.
They take a step back from the table and say, âIâm going to go back to doing my thing.â That particular group doesnât transform. You need that first layer, which is to say, âWhatâs the strategy? What can we do with this data and direct an organization?â That second layer to say, âHow do we get everybody onboard to be able to take their knowledge and their understanding of their business and apply that strategy to their day-to-day work?â
Bringing this into the world of subscriptions, you have all this data around how you acquire customers, how long they stay, when they leave, and what they do where theyâre with you to some extent. How do you make that into a good story? How do you explain to your colleagues across the organization what youâre trying to do with the data or even harness that data to make it useful to your colleagues?
I would say there is a number of different techniques. I donât think I have figured out a perfect model. I wish I did. I would probably make lots of money. Iâd have a better book. I love the book that I have but if you can solve the translation question, that is a monumental opportunity. Let me tell you what I have seen work.
The very first thing that we see companies struggle with is to get buy-in from other teams about how they can transform or use that data for their business. With most transformations, when you see something in data, the most opportunistic ones will have winners and losers. Youâre moving resources across functions, which means that some teamâs priorities will look better. Other teamsâ priorities will look worse.
All of a sudden, those people where theyâre looking at losing count or budgets become detractors for whatever you found in that data. Itâs not uncommon. I have been on projects where you see great insights and say, âOur subscribers are doing this, which means we shouldnât be doing that anymore.â The team at the losing side of that argument says, âWeâre going to run our own analysis. Weâre going to bring our own data.â You end up with a poor decision-maker who says, âI have one team here that says thereâs more money here. Another team that says thereâs less money here.â The average always comes out to be, âItâs inconclusive. Letâs keep studying.â
The first thing is to find a way to bring everybody onboard. We can talk about the specifics of that. That is the first strategy. The second part of it is to make sure that you avoid common mistakes. One thing that I see oftentimes is that we lose connection to the customers at the end. We talk in a language that is unnatural. Oddly enough, a lot of value was unlocked on our team, where we stopped going to executive coaches for presentations. Itâs not that they werenât talented. They werenât great at what they did but they spoke our language.
When we mentioned something like, âWe should use remarketing. Letâs take a look at churn,â everybody knew exactly what everybody else in the room was talking about. All those people were already onboard with the strategy. It was the people outside we needed. The gentleman I work with, this guy, Mark, out of Chicago, does Shakespeare as a career, which is strange. He can make Hamlet sound interesting. He can also make my presentation sound a lot more interesting.
When weâre talking to him, and we mention that, âThe statistical significance of this number or we should build this campaign,â heâs like, âStop. Youâve lost me. I have no idea what youâre talking about at this point because this isnât the language I use.â Sometimes itâs simply challenging teams to go beyond the status quo to say, âItâs not what you see in your data but how do other teams see it? How do other teams describe that relationship?â
There are two important points that youâre bringing up. One of them is that whatever story youâre trying to tell, there are going to be winners and losers inside the organization. Culturally, you need to pay attention to that. The other one is you need to make sure they understand and care, which is important for data people to connect with their colleagues across the organization. Itâs a challenging thing. Thereâs the left brain side where youâre trying to get to the right answer or be able to explain what your goals are. The other piece is communicating it in a way that people both understand and donât feel threatened by.
Iâll add on another piece to it as well. The biggest challenge for some companies, when they try to build this fluency around data, is people are reluctant to raise their hands and admit that they have no idea what youâre saying, especially with basic statistics and math. Theyâre like, âI should know this as a VP of the organization but I donât want to tell my team I have no idea whatâs going on.â
They sit silently and nod. Occasionally, they jump in. I had a story. I was working with a CMO, and we were presenting an analysis. I brought in a data scientist, a relatively new one who started presenting a whole bunch of things in technical terms. The CMO raises his hand. Heâs sitting there quietly. He looks and heâs like, âDid you use Base Z in this analysis?â
I sat there and like, âI donât necessarily understand what heâs saying and the data science.â That sounds like a statistical term but I donât know what that means. He cracks a smile. Heâs like, âI have no idea what youâre talking about but thatâs how I sound smart in meetings.â I looked it up. âDoes anyone understand Base Z? It sounds like itâs related to statistics. Everyone thinks Iâm smarter than everyone in the room because they donât know how to answer my question.â
Thatâs terrible.
Thatâs his approach. Iâm still friends with him. Heâs like, âI donât know what you all do. Itâs either you can explain it to me or I need to trust you. Iâd prefer to understand what it is youâre doing so that way I can participate.
A key thing that I hope people take from our conversation so far is if you are in a meeting and you donât understand something, ask. Most of the time, not only donât you look dumb but everybody else in the room is saying, âThank goodness somebody asked. That person is pretty smart because they caught me. I was getting off into a realm that wasnât relevant for the conversation.â Ask questions when you donât understand. Raise your hand. Youâre doing yourself and everybody else a service. Youâre keeping the speaker honest.
You wrote this excellent book, Converted, which I read on vacation. Itâs an amazing thing to have a business book that you can enjoy on the beach. It was fun, lighthearted, easy to follow, and full of practical tips and tricks, and almost no data speak. Congratulations on writing a book that was so easy to read and yet so full of valuable insights. What prompted you to write it and to write it in this very approachable style as opposed to writing it for your peers?
The style was front and center. For a lot of the sections in the book, in some cases, the introduction took almost six weeks on its own. If weâre talking printed pages, maybe eight pages. The reason is that you simply accept the reality. People donât read data books, so the very first thing was to say, âI need to write a data book thatâs accessible.â The subtitle of the book, The Data-Driven Way to Win Customersâ Hearts, was a long debate with the publisher to say, âShould we include data even in the title?â Do people see that and say, âHere it is, another data book, not for me,â and write it off?
We also wanted to add to say behind this conversational tone and engaging approach to measurement and analytics, there is a lot of math behind it. Thereâs also a time and place for it. Everything youâre reading about in the book is backed by research and data. Itâs not what we need to talk about now. We need to talk about the strategies that any manager can embrace to say, âThis is me. This helps me lead my organization and my team. I donât need to get bogged down in the weeds.â I emerge on the other side with the confidence to say, âI know how to lead these teams. Now, I know to participate in the conversation and how to demand more than simple dashboards. That should hopefully be better for everyone.â
The first section is about asking questions. I love that you start with that. Thatâs not even most people wouldnât even think of that as data. Thatâs asking good questions. You give the example of getting married. Can you explain what you mean by thinking about the questions you ask a prospect in the same way that you think about getting married?
I look at all of these engagements as customers. I put them in the lens of human relationships. I do that because thatâs approachable. Everybody has their own stories and interactions, something that they can provide and ground their relationships with. If you start with data, youâre starting to say, âLetâs talk about unique visitors, time on site, and bounce rates.â
All of a sudden, youâre stuck in this language of data. What we have here going on with our customers is the equivalent of courtship. They want us to trust them. How do we build that trust? When you start looking at things in that lens, other behaviors make sense. For instance, I was always told when I was dating that you shouldnât call a woman right away. You need to wait a couple of days. Give her space.
You say, âWhy do digital marketers do that? As soon as I leave your website, you start sending ads my way. Are you that desperate for my attention?â The data says people are less likely to subscribe and be your customer if you approach them in that way.
Instead, if you give them space, you wait to 48 to 72 hours, they are more likely to reengage with you. These stories and these connection points allow us to ground how we understand customers and to realize thereâs a very human relationship on the other end. We donât see those people. We translate them through numbers and spreadsheets.
What I loved in that analogy is we all know that you donât walk into a bar, find somebody that seems attractive to you and say, âI like you. Why donât we get married?â You donât go out with them the first time and say, âWould you like to get married?â They say, âNo.â You say, âLetâs go out again,â and then on the second date, you say, âDo you want to get married now?â You get to know them. You ask them other questions. You engage them in other ways. Maybe you go to a movie, meet the parents or talk about your life goals. There are a lot of different ways of asking questions getting to know the other person.
Letâs even unpack that a little bit. What are we talking about? The first thing weâre talking about is beyond anything else, having the confidence that you can make relationships work. Where a lot of companies go astray is they lack that confidence. We donât use these terms in marketing. We use accountability. Itâs to say, âIf I spend time meeting somebody, are they going to marry me or not? I need to know that right away, so Iâm going to ask.â
I would say, âNo, give it a little bit of space.â Organizations have loved that immediate gratification. Now I know somebody clicked on my ad and they bought but thatâs only harming you. Have a little confidence that these techniques will work. Thatâs number one. Number two is to participate in the conversation. Hereâs what happens when we talk about asking questions, a lot of companies immediately go towards, âYouâre saying I need to collect more data from the customers.â Yes, but in a different way. Whatâs the traditional model?
The traditional model is you donât ask the person youâre sitting across from you. Youâre not asking the customer for more information. Youâre going to third parties. Itâs like youâre sitting across, and youâre having a great conversation with the customer. Instead of asking them what direction things are going or what theyâre looking for, you say, âIâm going to pay my friend to tell me. They really like me. Do you think theyâre going to buy it? Who else are they talking to?â
Thatâs also not natural. The alternative is simply where I want to encourage companies to say you have to be more flexible and engaging with customers to ask them where they are in the process, âWhat you can do to improve, personalize that interaction, and be confident that youâll get that information back?â I want you to think the next time youâre interacting yourselves on these different websites and when youâre buying from companies, how few questions they ask.
Theyâll ask, âHereâs what I need to sign up for your subscription. Hereâs what I need to get you through the checkout funnel. Iâm scared to ask anything else because you may leave me.â What the data shows is that after somebody subscribes and they purchase from you, thatâs the highest their trust is going to be at that moment. They gave you money and a phone number. Things are going great. Itâs yours to lose. The Thank You page is what? âHereâs your confirmation. Here you go on your way.â Thatâs the time youâre supposed to respond.
What should you have on your Thank You page? What would be an example of something that you should have? I signed up. I subscribed. I bought my first product. What happens then?
Generally, Iâm a big fan about flexible questions, which is empowering your organization for anybody to ask the questions and try to capture data that can encourage and build new hypotheses for your business, not tying everything up in a single annual survey but to say, âLetâs have a thing.â Weâre going to try different questions. Some may work well, and some may not.
The first thing is the Thank You page is going to become our playground to ask. âWhat do I like to ask? I like to ask about the share of wallet.â Southwest Airlines does a fantastic example of this. Every year or so, Iâll get a survey that says, âHow much are you spending with other airlines?â If youâre looking at the value of a subscriber, you want to know how much are they spending in a category. If theyâre already giving you all their money, that relationship is in a good place.
If theyâre not only giving you a portion of the time and itâs going to someone else, I want to be curious as to why. What can I do better to service your needs? Why are you not having your needs met by another party? What can I do to be better? Another question I like to ask comes as part of surveys, and this is one thatâs mentioned in the book, which I love because it becomes so intuitive and powerful.
On surveys, people generally ask, âWhere can we improve? What can we do better?â What did some professors think? They said, âHuman nature is when you ask somebody, âWhat can I do better?â they start thinking about all the ways you let them down.â If I go to my wife and be like, âWhat can I do to be better?â Sheâs going to be, âYou should have picked up the kids. You should have cooked a better dinner.â She doesnât complain at all. When we ask people negatively, they think about the negative.
They would say, âWhat if we turn that around? What if we ask people about the positives? What did we do right? What do you like about our offering? What do you like about our company?â As it turns out, when they ask people that question, they were able to measure effects for more than a year and a half later that those customers ended up being happier, having higher lifetime values, and subscribing to more products and services. Thatâs from a single question that carries almost no risk in something that all businesses do but havenât asked.
Where do you ask that? Is that once a year question? Is that a Thank You page question?
The effect was measured on surveys. They took standard ten-question surveys, and they put that question on top. You still had the same 10 but now you had 11. Thatâs how they did their test in control groups. In other cases, I would have no issue putting it on the Thank You page because the psychological effect should be equivalent. Thereâs nothing special about it being in front of those other ten questions. You need to ask, âWhat do you like?â It is a nice word to say, âGive me a compliment. Tell me what you love about me,â but in a more balanced way.
Those are three great tips. Use the Thank You page, ask about the share of wallet, and ask what weâre doing well instead of focusing them on what weâre doing wrong.
âWhy am I a bad match for you? Tell me.â Weâre on the path of self-improvement but it seems an odd question to ask somebody individually.
âAfter somebody subscribes, after they purchase from you that's the highest their trust is going to be.â - @neilhoyne Share on XYou talked a lot at the beginning of the book about how to ask good questions. You talked about asking good questions as if youâre talking to a real person, thinking of this as a relationship where you want to learn about them over time or itâs not a once-a-year thing. Itâs not a one-undone, which very much aligns with how I think about subscriptions. How do you do that at scale? How do you do that when youâre trying to reach a million people and make each one feel like you are carefully and thoughtfully developing a personal relationship with them?
I was talking to a very successful restaurateur in New York City. Until I thought about it, he was challenging the notion of what Silicon Valley calls “scale”. He said, âIn Silicon Valley, the idea of “scale” is that youâre going to build a product that can deliver services to millions of people. Thatâs a scale they look for.â
He challenges that the real definition of scale is being able to deliver the same type of service, experience, and relationship to 1 person and delivering it to 2, then 10, then a million. It challenges us to say if we had one person come to our website, that was our entire day, and we can have that relationship with them. How do we then embrace the two tools and techniques that can offer that same service and level of conversation to future visitors?
The definition grounds us to say we shouldnât be making compromises based on technology but letâs talk practical tools. What are companies doing? The first thing is the language of the company has to change where theyâre talking about people, things that are valuable, and understanding the relationships instead of calling people unique visitors or weâve received.
The language matters?
The language matters because thatâs how we condition and treat it. When we talk about things like clicks and conversions, what are we doing wrong? Weâre almost commoditizing it. We can get more people anytime we go out into the market, and we increase our budgets. There are only so many people out there, and youâre turning away great people, so use that very human language.
The second part is to look for those types of tools that allow this scale to happen in a more natural way. This is where machine learning fits in, artificial intelligence. Whatâs it supposed to do? Itâs supposed to take all the patterns, all the conversations youâve had in the past, and tell you what to focus on. What are the moments that truly matter in the relationship? When youâre looking at your website behavior, you may see someone bounce in between 30 different pages.
Machine learning is going to have the best avenue to say, âWhich of those moments do we think something changes? What should you pay attention to?â When you see a lot of research that comes out, theyâre using those techniques. Iâll give you an example. A lot of companies focus on adding something to a shopping cart or starting the checkout process to say, âTheyâre getting serious.â Thatâs the next step in the flow. Itâs to say, âWe had a date. Theyâre giving me their phone number. Weâre having a second date. Things are progressing,â which may be true.
At the same time, they turn away other softer signals they should pay attention to. At least in the retail setting, people that remove things from their shopping carts, thatâs a sign of curation. We may think, âTheyâre saying no. Theyâre pushing back.â No. Theyâre being more selective with the products they want to buy. Itâs nothing to walk away from a website. Itâs another thing to come back.
That is an interesting point. One of the things that I encourage companies to do when theyâre thinking about their product offering for a subscription is to start by doing it in a very manual way. When you think about the example you gave, if Iâm in a store, walking around, I got my shopping cart full and start to take 1 or 2 things, if Iâm done with the store, and I donât find anything I want, Iâll abandon the shopping cart and walk out.
If Iâm saying, âIâm going to take the red blouse out and keep the green one,â that means that Iâm taking it more seriously. If you think about that logic of what itâs like to be in a store and say, âIâm shopping online. Does it mean the same thing when I take something out of my shopping cart?â You start to have those a-ha moments.
It totally speaks in terms of treating the user as a person and not thinking about clicks but thinking about actions and trying to understand whatâs going on in their mind or journey to accomplish something. What any purchase decision is, âI have a goal. Iâm trying to accomplish something. I want to look beautiful, so Iâm going to buy some new clothes. I want to get fit, so Iâm going to join a fitness program. How are you going to help me get there?â
âWhen somebody asks what they can do better, they start thinking about how you let them down. What if we ask people about the positives?â - @neilhoyne Share on XI hope, for all the readers out there, that the takeaway isnât in our customer funnel. We should have removed things from the shopping cart. Donât make this a linear process. The larger takeaway is to encourage and establish that curiosity. What youâre doing is youâre setting the stage. This is who you are as a company. This is what youâre offering to consumers. You can guide them. You can have your expectations but consumers are going to do things that surprise you. If your path is to say, âI want to force them to think how I think they should act,â you must go through this page next.
Thatâs what I want people to get away from. Where the real value comes in is saying, âWeâre going to set our stage here and let our customers interact the way they want to interact. Weâre going to be curious about why and what we can do better. Weâre going to open our minds. People are going to use products in ways we never anticipated. Theyâre going to use our website in ways that we didnât plan. Weâre not going to try to corral them and say do what we want.â Weâre instead going to understand why and how we have better relationships by adapting to what our customers are saying they need.
That goes right back to your original point around asking a lot of questions and being curious. If somebody is taking stuff out of their shopping cart, maybe ask them. I donât know if this is a good or a bad idea. You can tell me but if youâre at that Thank You page, I had a green blouse and a red blouse. I dumped the red blouse, and I checked out, âThanks for shopping with us, Robbie. Why did you dump the green blouse?â
What happened?
Was it, A) Because you were getting ready to leave us and somehow we won you back or was it, B) Because you were optimizing which one am I going to wear tonight?
Look at the opportunities and draw a list. We didnât cover all of them in the book but there are some retail sites that say, âWeâre going to mix up our categories for return visitors. Give them a new way of experiencing us, not here is menâs, womenâs, kidsâ but as you brought up by occasion and what are we going to get? More curiosity from the consumer, which is what they find. They increase engagement with products but also signal to say, âWhat are you looking for?â âIâm looking for work clothes.â We didnât get that when you went into womenâs and separates. Now we have that understanding.
Look at companies that optimize menus. Starbucks does a great job dropping the currency sign from their menus to make people less anchored by price. Thatâs something we can try on our website. Letâs not remind people of price. Even the way people interact with products by zooming in and looking at the texture, size, and color of products generally is an indicator to say, âAre they going to be happy with their purchase? Are they going to call customer support right away and incur more costs for your business?â
When you talk about, âAre they going to be happy with their purchase?â If theyâre not happy with their purchase, the chance of them coming back goes way down. The lifetime value declines. Youâre a big believer in optimizing for lifetime value. Can you talk a little bit about why that is, what that means, and why itâs important? In the world of subscriptions, itâs something we talk about but itâs true in almost any business.
I look at it from a very simple human lens. I meet a lot of people in my life. Some people are very close to me, my wife, my family, and close business associates that provide so much value to my life that I simply couldnât do without them. They change my life and how I respond. I certainly want to find and build relationships with people like them.
On the other hand, there are people that youâll meet only once or twice in your life. It may be a flight attendant or the Uber driver that took you to the airport. Youâll never see them again. Theyâre great people, and they had that moment. First of all, imagine how different your life is if you donât understand and identify those relationships and you donât spend a little bit more time, if you havenât seen a close friend or family member in a while, if you donât reach out to them or even worse, if you think about the other side. What if you treat everybody equally?
What if you go out and youâre like, âI have to make an important career decision. Do I ask my family or do I ask the Uber driver?â I then average the two answers together. The reason why I say this is because practically speaking, when we look at how companies build out their subscription sites, what do they do? Theyâre going to run an AB test. Theyâre going to say, âDo people like the red button or the blue button? This creative or that creative?â They then treat everybody equally.
They say, âThe family member is worth as much as the Uber driver.â Even though the family member is going to come back and be with us forever, weâre never going to see the Uber driver again, and weâll average their results together. When people start looking at their business from a lifetime value lens, it allows them to prioritize opinions, data, and questions a little bit more. We want to listen to these people and build for them. We donât want to turn these people away but weâre not going to put as much emphasis on their opinions even if we ask.
Listening to the customer doesnât necessarily mean listening to every customer equally. It means understanding who your best customers are and optimizing for more of them for them to be happier. This is very much, in the role of a past guest that we had on the show, Peter Fader, a friend of yours and mine, who talks a lot about this concept of customer-centricity.
You talk about this in your book. You have a whole section dedicated to a very practical way of incorporating customer lifetime value into your work by segmenting your audience. Can you talk a little bit about how to do that and where to get started if itâs something that youâve never thought about before as an organization?
The first practical step is certainly to calculate it. One of the things we talk about in the book is best-in-class practices. If you search how to calculate lifetime value on Google, youâll end up with 3 to 4 pages of poor results. The first thing was to clear the air to say, âNot only hereâs how you calculate it but to give everyone confidence that they can calculate it without it being a month-long effort, that it is relatively straightforward and proven in these models.â
The second thing is, once you understand who your great and customers are, understand why. âWhat makes them different?â Was it where they came from, the time of day, a promotion, a coupon code that you had, the products or services they were interested in early on, âWho introduced you, or maybe the agent they spoke to over the phone that won them over?â Thatâs the next question. That helps you to build that profile.
Behaviorally, what makes people different? Afterwards, you have three choices. You certainly can go out and acquire more customers. You can develop the customers that you have a little bit further or you can retain the high-value customers that you have now. Thatâs everything past that point. Itâs walking people through to say, âHereâs what options are available to you to acquire more of these people. Hereâs what we can do to develop them. Hereâs what we could do to retain them,â all with a little bit of segments to say, âHereâs how far you can push these rules before things start to break.â
For instance, you want to acquire better customers. You donât necessarily want to say, âWeâre going to go after the very best customers, no matter how attractive,â instead, you simply want to meet better people now than you did yesterday. Those rules are put forward, so people know, âCan I save a little bit of time by making sure I learn from the best and worst practices in the industry?â
When I think about the companies that I work with, I felt like this was one of the most important sections of the book, this idea that itâs not that hard to segment your customers by value. You break it into four quadrants. You can use a few data points as a starting point and use some hypotheses to start to figure out whatâs different in quadrant 1 than quadrant 4.
Start to tinker on the margins so that youâre spending a little more in the channels that attracted more of the quadrant 1 people using those messages a little more and a little bit less in quadrant 4. One of the things that I hear companies talk about is they say, âIt seems very risky to segment and go after quadrant one.â Thatâs not what youâre saying. Youâre very explicit about that.
Iâm clear about it because was that dramatic? Imagine youâve built your business on whatever strategy youâve had for years, and then all of a sudden, the CMO comes and says, âEverybody stop. Weâre going to go after these people.â Youâre like, âWeâve never reached out to these people. We donât know how to message him. We havenât learned this conversation with them. We happen to gather them. How hard could this be?â
Instead, you say, âLetâs take everything you have. Letâs not try to reinvent your business. Letâs try to nudge you in a better direction so that you make more money. If that will work, we can do it again.â We find teams generally are more accepting of that change and something overly dramatic to say, âNo more doing this, only doing that.â Thatâs far too binary. It doesnât lead to a lot of success.
At the same time, not all customers are created equally. If one customer is worth $40 and one is worth $200, donât say the average customer is worth in between. Understand the difference between the $200 our customer and the $40 customer but be gentle, curious, and experimental as you continue to optimize around those higher-value customers.
You donât need to abandon the lower value customers and say, âGet away from my business.â Maybe you donât want to spend as much marketing investment on them. During that time, you want to prove to yourself to say, âIf we believe these customers arenât going to come back, letâs take a couple of hundreds of them. Letâs stop marketing to one group. Letâs continue marketing to the other and look at the results.â
The reason we do it is not because I lack confidence these techniques will work. Sometimes you need to prove it to your organization that they do. That way, you can make iterative changes. You can learn from it. How do these people respond? You can move forward with the next test. That way, youâre keeping yourself lightweight and nimble in your learning as you go.
The systems you need to do this arenât that sophisticated. You can use a pretty straightforward spreadsheet. You donât need to have tons of data on your customers.
A lot of people think, âI need big data systems. I need cloud.â Iâm like, âMost of the work I do is in Excel. Itâs easy to work with but thatâs also why these techniques rise to the top. Do you want to go out and buy a new enterprise system takes eighteen months to install? That would be disappointing. Thank you for reading the book. For the next eighteen months, youâre going to be doing the exact same thing we started talking about not to do, collecting and organizing more data. Letâs use Excel and get into work.
This is also an important takeaway. Start getting curious about your customers. Ask them more questions. Try to understand who the best customers are using the data you have available now. Use your Excel spreadsheet. Donât worry about saying, âWeâre not going to do this for two years while we evaluate, vet, build our new big systems, and hire as the data scientist to give us the answer.â
âData scientist, go do your thing.â Youâll never hire any good data scientists if they find out thatâs the job.
The last thing I wanted to talk about with regard to the book is culture. I love that you dedicated a chunk of the book to talk about how to build relationships. What I see with a lot of companies that move, letâs say, from transactional models to subscription models, which almost by default are more focused on customer lifetime value, long-term relationships, all the things that you talk about.
People within the company donât understand what the subscription people are talking about. When they understand it, theyâre threatened by it. Those are the two parts. There are winners and losers in terms of who gets the budget, who gets the headcount, who gets to be the hero at the exec team meeting. What have you learned in your work at Google and in your work more broadly working with hundreds, if not, thousands of companies about the cultural piece of moving to this model?
âTwo great tips: use the thank you page, and ask about share-of-wallet. - @neilhoyne Share on XIâve learned candidly that even if you have the guaranteed right answer, there is a 50/50 chance that companies will accept it. Thatâs what it is. Iâve run experiments that show, âHereâs a company that can make 5 or 10 times more revenue, proven in a best in class methodology and still come like, âOur data is a little bit different,â and they ignore it.
This is the reality. We tend to trust ourselves and our intuition more than we trust data. We donât admit it. We donât think how data wins. Data rarely wins. Itâs oftentimes the people in the room that will make that decision. What that last chapter is about is to say, âIf you want these ideas to take hold, itâs more than learning the principles. You need to learn whoâs on the other side of the table.â
We touch on the best practices. What are the core areas? One is this idea, this theme we spoke about around incremental improvement for your business as opposed to chasing a perfect answer. We talk about the need for experimentation to collect data, build confidence in what you are doing and the changes youâre making. The models that youâve constructed are, in fact, leading to better profitability.
We talk about how you get other people in the organization on your side. How do you get that agreement to make sure that when youâre running these tests and implementing these models, people arenât going to be threatened in that transformation, and they have the opportunity to grow? The fourth and perhaps most important area is, how do you bring other people in your organization along with you?
How do you make them evangelists of your work and what you know as opposed to simply coming along and saying, âWhat does Neil say is the right answer,â and they do that? If you donât do that, you undo that entire storytelling aspect we started with. Theyâre not able to participate. Theyâre not able to use their experience to help you ask better questions, formulate better insights, run better tests, and being able to educate them on what you see in the direction you should go so that they can participate and add value themselves.
Thatâs the secret sauce. There are lots of smart people out there who can tell you the right answer. The hard part is bringing together a team to support implementing that knowledge.
Weâre only human after all.
I want to close out with a speed round. Are you up for that?
Letâs go.
What is the first subscription you ever had?
It would have to have been a newspaper at that point. Iâd probably say the New York Times. If not that, Iâd say maybe Nintendo Power when I was a kid. They had that magazine. It was a big thing when I was five. Iâve got the inauguration but that wasnât my subscription. That was my parentsâs.
What is the subscription you get the most enjoyment out of now?
Iâd have to say Netflix. Itâs not any good movies. Itâs just because of my kids. I have a 2 and a 5-year-old. They love Netflix. Nothing makes them happier than watching cartoons with me.
What is your best advice for a reader spending a significant budget with Google on advertising?
Trust anything your account team says. The best advice I have with anybody spending time with Google if you are working directly with Google, you should be. Let the people on the other side know what success looks like for you. Oftentimes, especially with advertising platforms, they send over too much data. âHereâs more data and case studies.â They donât necessarily know the burden of proof in the organization to say, âHow do we build that trust? What do you need to see?â That way, the information and guidance theyâre providing are speaking the language of your business.
Thatâs helpful. Whatâs your favorite question to ask a new customer in general?
My favorite question to ask is probably the one that I already brought up, which is, âWhat do you like about us?â I gain a lot of insight from that question. The worst question I ask people is, âHow did you hear about us?â As you put in dummy answers, people will select channels that youâre not even marketing on. Sometimes, it can be difficult to get honest answers in those categories.
My final question is, what is your favorite underused metric?
Weâve spoken so much about lifetime value. I have to throw it in the mix because people donât look at it often enough. Theyâll spend more time on new customer rates, which is our garbage bounce rate. I donât know what you learned time on site. Itâs like, âWhy arenât you looking at lifetime value customers?â âWeâre too busy looking at CPA or ROI.â
It says, âLook at what your customerâs worth even if you donât use it or youâre not going to bid on it. Know itâs there.â If you have a conversation where everyone is saying, âLook at the CPA of our campaigns. Look how much it costs for each new subscriber.â Somebody puts up a column next, and youâd say, âI know weâre not bidding on it but how come these subscribers are going to stay longer than those subscribers?â I love those debates within organizations. Thereâs no pressure because you didnât tell anybody to use it but I want people to start asking why.
That is fantastic advice. Itâs been a pleasure to talk to you, Neil, as always. Thanks so much for spending time with me, and congratulations on a terrific new book.
Itâs my pleasure. Thank you so much for having me and for the wonderful questions as well.
—
That was Neil Hoyne, Chief Measurement Strategist at Google, as well as a senior fellow at the Wharton School. Neilâs new book is Converted: The Data Driven Way to Win Customers Heart. For more about Neil and his book, go to ConvertedBook.com. For more about Subscription Stories, go to RobbieKellmanBaxter.com/podcast. If you like what you heard, please go over to Apple Podcasts or Apple iTunes and leave a review. Mention Neil and this episode if you especially enjoyed it. We read all the reviews because we want your feedback. Thanks for your support and for reading
Important Links
- Neil Hoyne, Chief Measurement Strategist at Google, Senior Fellow at Wharton, Author of Converted: The Data-Driven Way to Win Customers’ Hearts
- Converted Book
- Uber
- Subscription Stories Episode with Peter Fader
- New York Times
- Nintendo Power
- Netflix
- Machine Learning
- Customer Lifetime Value
- Data Science
- CPA
- ROI
- RobbieKellmanBaxter.com/podcast
- Subscription Stories: True Tales from the Trenches on Apple Podcast
About Neil Hoyne
Neil has served as an analyst, researcher, inventor, lecturer and, in his words, the father of many forgettable slides of glossy funnels and Venn diagrams. A witness to and participant in billion-dollar successes, and instructive failures, all in the pursuit of building indestructible customer relationships through digital media. A key player in the executive rallying cry to be more âdata driven.â
As Googleâs Chief Measurement Strategist, Neil has had the privilege to lead more than 2,500 engagements with the worldâs biggest advertisers. His efforts have helped these companies acquire millions of customers, improve conversion rates by more than 400 percent and generate billions in incremental revenue. Immensely proud of the degrees heâs earned from Purdue University and UCLA, Neil returned to academia in 2018 as a Senior Fellow at the Wharton School of the University of Pennsylvania.
Love the show? Subscribe, rate, review, and share!
Join the Subscription Stories Community today: