AMA on Identifying and Multiplying Your Best Customers with Customer Segmentation
EVENT RECAP
Using segmentation to focus on best-fit customers is a powerful lever to help a company drive down cost per acquisition, drive down customer churn, and increase ARPU or customer lifetime value. In this session, Tamara Grominsky walks through how to take on a segmentation project—from analysis to product packaging.
Ideal for Founders and Product Marketing Leaders
Join to discuss:
- Variables to use in segmentation (e.g., firmographics, psychographics, and behavioral attributes)
- Research and analysis to identify segments that are high volume, high performance, and high potential
- How to roll out segmentation to inform pricing, packaging, sales, and marketing strategy
Video
All right. We are officially live. I am excited to be on with Tamara. Funny story. We first had Tamara for a conference that we helped organize.
It had formerly been in person. We did a virtual version in 2021. And she was the highest rated presenter at that conference. And so every opportunity I get to bring back in her content, I’m going to take advantage of that.
The cool thing is that at the time, Tamara, you were a full- time product marketing and strategy leader.
Now you’ve moved into this PMM camp world. And I’d love actually for you to introduce yourself a little bit, your history and your present.
Sure. Yeah. So I’m Tamara and I’ve spent the last 15 years, most of my career as a product marketing leader. So I was a two- time VP of product marketing at Unbounce and at Kajabi. And I also was the chief strategy officer at Unbounce. So definitely subscribed to the more strategic side of product marketing. A lot of that, like, who are your best customers? How do you understand the value that they’re looking for? And how do you price package and position that value, you know, to drive sustainable business growth?
About six months ago, I actually branched off on my own. So now I’m the founder of a company called PMM camp, and really we’re all things product marketing as it sounds. So I do, I have a private network of product marketers, which is amazing. So I create content and education for those product marketers, really kind of raising up the next generation of product marketing leaders, and then do go to market consulting.
So of all the topics I just talked about, kind of my sweet spot being segmentation, pricing and packaging and product launch actually, which is a topic I’ve been super passionate about lately.
Cool. So we’re going to get into, I mean, segmentation touches on a bunch of different things. So we’ll start there and the conversation can go into other things, product marketing. So we pulled a couple of questions that we saw particular interest in, in registrations. The first is, how do you know that a company needs to invest in segmentation and defining segments?
Yeah, usually the very first sign to me is that everyone at the company thinks that you’re going after a different customer. So that’s why I built this line. And it’s like a bit of an extreme version, but sadly, actually happens within many, many businesses. And it’s like, sure, maybe your lower level employees, they’re all going to say something slightly different. But when all of your leaders are prioritizing completely different things, this is like the number one sign that you’re never going to be able to get momentum or align your factors.
That’s one. The other one that I’ve noticed more and more is it’s really like if people come to me, founders come to me and they say like, our growth is just plateaued. That is actually usually a really good sign that they just don’t understand what market they’re trying to capture. And that they’re probably trying to go after too many markets and they’re taking too broad of an approach.
So that would be a time when I would speak with them about like, okay, let’s either go more niche, or at least just like really clearly define who your best customer is so that we can kind of build a strong strategy for them.
The other one might be if maybe for slightly later stage businesses, or even like scale ups, if the customers that used to be very reliable to you, like you could always count on a certain conversion rate, adoption rate, retention rate, if any of those core lifecycle metrics start to change, whether they go up or down, that is usually another indication to me that something has shifted in the market. And we need to go back to like the foundations to understand is this still a customer segment we want to prioritize? Or is something changed about them?
So if the metrics are getting better, maybe we want to put more fuel on the fire. And we want to know more about them. But if the metrics are starting to decrease, particularly this will happen top or bottom of the funnel, then that’s usually a sign that we’re not as close to the customers we used to be, or that maybe we’re not resonating the same way we used to.
Awesome. Okay, we’ll get into we’ll start to get into some tactics. So at a base level, how do you identify your best customers?
Yeah, so I love to take a data backed approach, because this is what often happens. The founder is in a room and they’re like, well, I think we should go after this customer. They were part of my whole pitch to raise money. I’ve been talking about them and thinking about them for years. And usually one of two scenarios takes place either that is like actually quite a generic or broad segment. And it really isn’t a segment. It’s like a market, in which case we need to actually drill down. Or maybe their initial hypothesis was wrong.
And there actually is a better market for the business.
So what I love to do is I love to look at data, whether you have 50 customers or 5, 000 or 50, 000 customers, you can still look at the data available to you. And I developed something called the map model. So it works its way up from the bottom.
We start with this concept of like measuring volume.
So looking at the data you have today, who are you attracting and who do you have a lot of? And if you’re super early stage and you’re like, I have five customers, I’ve actually worked with folks with this model where it’s like, well, it’s actually resonating to the conversations that you’re having. Where are you getting those initial kind of good signals from? That can also be used as a data point.
When I worked in a lot of PLG models, we would even use like trialers as part of this data set because we’re like, well, they haven’t paid us yet, but they are interested.
And so we’ll use that data.
And what I like to do is I look at that data set and I start to form clusters. And so I wanna look at like, where are there groups of people who have something in common? I think we’re gonna talk about that a little bit later.
Types of shared attributes we might look at.
Then once we have these clusters, I usually aim to have maybe three to five. And I think about these as hypotheses. I wanna either prove or disprove my hypothesis right or wrong, right? So I’m looking to do that by actually saying, well, I can’t just have a lot of them or I can’t just have interest from them. I need to make sure they’re good for my business. And so that’s where I look to things like, is their conversion rate better than the average customer? Is their lifetime value better than the average customer?
Is their retention better than the average customer? And you don’t need all of the life cycle metrics to be better, but you wanna make sure that there’s a few that are, and the ones that aren’t better, you wanna feel pretty confident that you can improve that. So you might say, well, maybe they’re not adopting as much as I would like, but like, I know that if I ran adoption programs, I could get that over the hump. So that’s one that I look at.
And really you’re looking with this concept of like, if I were to get more and more of my best customers that increase the ratio of them in my customer base would naturally the performance of my business rise without me even having to improve the performance metrics of those segments. So that’s kind of like the second stage. And usually by this point, you’re starting to filter out. So let’s say we started like five, we might be down to like two or three now. And then I wanna look outside the business.
So the first two steps are looking inside the business or at the data you have. And now I wanna say, well, if I were to prioritize this customer segment, could that market support my own growth ambitions?
So who else is going after those customers?
How large is this market?
So I’m looking at some like TAM numbers here.
How fast is this market growing? I remember one time at Unbounce, we had a perspective segment that looked large, but actually when we dug into the data, we saw that that market was declining every single year. We’re like, yeah, we’re like that’s, it could be interesting to us for the next 24 months, but we’re not gonna kind of invest all of this technology and roadmap resources when we know that this market could be half of the size three or four years from now, right? And so sometimes a smaller market that’s really growing fast.
So at the time for us, that was e- com. And we’re like, wow, e- commerce, smaller market today, massive market tomorrow. Let’s kind of hitch our wagon to that. And we’ll actually just like grow with the size of the market, right? So, and then the final one being like, okay, great. The market is large. We feel like it’s growing in the right direction.
Do we have an understanding of how we would actually go out and reach them?
Because sometimes you’re like that market’s awesome, but like there’s just no path for you to get that customer. And so I like to think about if ideally you have already acquired some of them today, what are the channels they come in from?
What is the customer acquisition costs?
Are those scalable? And if you don’t necessarily have customers in this segment at least understanding that there is a path to acquiring them at a reasonable rate.
Nice.
Not this model.
So one question that I think relates to kind of line two is how do you use customer segmentation to identify customers who may churn or need re- engagement? I’m actually wondering if we have to think first about whether you want those customers, like maybe they’re not ideal if they are churning. So I guess in line two in LTV, how does churn factor in either, do you want the customers or not? Or how do you treat them differently maybe to prevent that?
Totally. So I would say if there’s a large group of customers who are churning, most likely I don’t want to prioritize that group.
I have, though, at times said, well, what would need to be true for us to prioritize them?
And maybe like all of the other things are really interesting, like maybe you have a massive amount of them, maybe you even have a bunch that have churned, the market is large.
And then I’ve worked with my product team to say, well, if we had this value, this value and this value, we would be able to retain them. Or maybe it’s like, well, actually, for that group, it’s a smaller price point and package, right?
So maybe they don’t fit within our current business model.
But if we had a freemium plan, maybe they would, or if we had half the price.
And so then it becomes a bit of a business investment decision, which is more of a future strategy.
So putting my CSO hat on, these would be the kinds of chats I would have around, okay, we actually want to go after this customer, it’s probably gonna take us a year and a half before we actually can, but let’s dedicate some resources to do that.
But then if I put my PMM hat back on, I’m focused on like, what are we doing today to grow the business?
And that’s where I would say, well, they’re not going to be a target segment today. That said, I have had segments where for many reasons, they are very lucrative and appealing to us.
But their churn is the worst thing about them.
So again, I’ll use an example from Unbounce, where one of the groups was agencies were like agencies have always long time been a great customer at Unbounce. And metrics were great. We had a lot of them.
They were really engaged. We had amazing relationships with them, so we could acquire them very scalable. And but what we saw was we weren’t, the churn was a bit higher than we would have liked. And so what we actually did, because we were really interested in this group, we did some customer interviews with agencies, both those who were customers and non customers, and we understood that there wasn’t actually a lot of product development we needed to retain them.
It was more about how we let them engage with their customers. And so we actually were able to just allocate one of the existing headcounts in our business to be a bit of like a agency partner manager.
And we were able to really prioritize this customer segment with no additional product needed.
And so in that case, we basically put like a tiger team together to say, okay, we’re going to prioritize this customer segment, but we’re going to start an initiative to kind of decrease that churn rate immediately. So there are ways that you can work around it.
And it’s what I love about this model is because you’re actually breaking your segments, your customer base into different segments, you can clearly see where along the lifecycle you need to prioritize. And so you might have two ideal customer segments, and maybe one of them you need to focus like top of funnel and one of them’s bottom of funnel. And then it makes it super clear for your teams, what initiatives you need to plan against it. Awesome.
Okay.
So let’s keep cruising into what data goes into segmentation. And we’ll have a couple of questions kind of to follow up on this one.
Totally. So I get this question a lot. And the answer is in a nutshell, whatever data is available to you, because I know like sometimes, especially the company’s really early stage and they’re very enterprise, they’re like, I don’t even have that much data versus, you know, maybe someone’s a bit more mid- stage and they’re PLG and they’re just like sitting on hordes of data. But sometimes that can be a curse too, because you actually can’t like access it or merge it or join it.
And so I say at the simplest level, whatever data you can access and attribute back to an individual customer is the best place to start.
When I think about buckets of data, though, I think about these five that we have on screen. So I’m going to skip the first one, I come back to it, but we have demographics. So these are pretty obvious. I think most people would consider demographic and geographic to be like the most basic forms of data.
So if we’re talking about like a B2B world here, it’s like, what’s the size of the business? What vertical are they in our industry? Is the buyer a specific job title?
Is the user different than the buyer? I’m kind of looking at all of that. Are they located anywhere particularly interesting?
And then I love to look at behavioral and psychographic.
So behavioral, if you have a lot of product analytics, this is usually where you can tap into those. We can start to look at, you know, maybe someone is coming to you in droves, but they’re actually not even using your product. We would consider them to be dormant. That might not be good for the business because A, they’re not really extracting value and B, if they’re not extracting value, they’re probably going to churn at some point, even if it’s not now. So I love to go into what products are they using? Is there a depth of usage?
And so maybe they’re using a lot of your product, but they’re using kind of like an older part of your product, not like the newer direction that you want to go in. And so that might not be a customer you want to focus on for the future.
So there’s tons of nuance you can get here. Psychographic is this idea of like, how do they buy? What’s important to them?
It’s tapping into the more like emotional, personal side. Usually this is available through things like customer interviews, so not always as scalable.
But I have come across some industries like healthcare, for example, or a business like Headspace or the Calm app.
They’re probably actually going to have some like interesting psychographic data because we give that to them when we sign up for those apps.
It’s like, I’m a high anxiety individual, I don’t sleep well, right?
So don’t feel bad if you don’t have this.
And then the fifth category is what I call account data.
And people often miss this in the data set.
And account data is basically linking that individual customer profile back to their impact on your business. So I want to know, well, when did they sign up for the subscription? Or if it’s a physical product that maybe they’re purchasing, it’s like, when was the first date of purchase?
How many products have they purchased?
Are they on a monthly plan or an annual plan? Because this is going to help me start to understand tenure, LTV, some of the more business side of things. Because I really want to look at both sides.
Where would business maturity fit in? Somewhere in the demographic category? It’s an interesting one where you can have two businesses, same size, but within a function, they look very different in terms of their sophistication.
Yeah, I think it would probably be in demographic. But probably some nuances would also show up in that psychographic bucket as well in terms of what they’re trying to solve for, the jobs to be done at that particular stage.
So that’s why as much as you can look at data in a data set, I love to talk to customers as well. And it really is the blend of qual and quant.
So oftentimes, though, I will ask, how long have you been in business?
And that will be a clear indication if they’re like, I’ve been my first year in business versus someone, to your point, they’re the same size, but they’ve been in business five years.
They might be solving for completely different things.
Or so I always get, I sometimes get tripped up between ICP and persona. Like persona is the person, but ICP is the company. Here we’re talking about the company, right?
Yeah, so I feel like persona, I’m like allergic to that word a bit. Well, for me, a best customer or an ideal customer profile, as I call it, this is like the segment of the market you should be going after. A segment of a market could have multiple personas, actually. So let’s say, so let’s go back to maybe the agency model, the agency when it unbounds again. So let’s say we said one of our customer segments was agencies. And we actually had more nuance.
We’re like agencies with less than 50 employees or whatever it was.
Yeah.
This much money.
But there’s actually like two personas within that segment.
There is the agency owner.
And then there’s like the agency account executive almost. And the account executive is the one who’s in unbounds all day long.
They’re managing.
They’re building the landing pages for their clients. And they’re going to have different buying criteria.
Of course.
They’re going to have different behavior than the agency owner.
So that would be kind of how the two fit together. And I think sometimes people try to solve for personas first, but they don’t even understand the segment that they’re trying to solve for that persona.
And so within, the agency model is an interesting one. So let’s say some agencies have data scientists and are far more sophisticated when it comes to their data approach and some don’t. Are those two different segments? And how do you suss those apart? It’s probably did they hire data scientists, but how do you think about?
Yeah. So one of the things I might do is even with, so let’s say 50% of the customer base is agencies.
And I’m like, we should focus on agencies.
Then I’m going to try to understand, well, who are the best agencies?
And maybe out of that data, I’ve seen, well, agencies that have data scientists are actually way better for our business than those that don’t.
Now, going back to the map model, I’m weighing off a few things.
I’m going to assume that agencies with data scientists is a much smaller market than agencies without, right?
And so I might decide, actually, we’ll go after any kind of agency as our segment. And then maybe that’s a persona, or maybe that’s just a marketing campaign. Sometimes we go too deep on things, and it’s like, great. There’s probably two different campaigns here.
We could do one campaign that would really resonate with agencies who have a data scientist.
Maybe we’re showing up at a certain conference that they all go to, or whatever it might be, versus we’re going to acquire agencies that don’t a different way.
But it’s not necessarily changing what the segment is, or even necessarily what the persona is.
That’s helpful. One more question on this slide. So how do you think about balancing these different categories of data?
Oh, sorry.
I’m pulling up the wrong question, a different one than I thought. The question, I can’t find it now. But it was about how you balance these different categories. Should you weight demographics the most heavily? Should you weight account data the most heavily?
Yeah, that’s a good question. I would say, I’m pulling up the wrong question.
there was no right answer. And this is where it’s really important that you weigh the account data in. So basically what I’m doing is I’m pulling a master data set that has as much of these data points as possible.
And then I’m just doing some data exploration.
And I’m basically trying to answer the question like, what do our best customers have in common? And so for example, you may see that like, customers with the highest LTV for example, there is no business size in common. So then I’m like, okay, well, I’m not gonna weigh that is very important. So maybe it’s actually that they all live in Texas. I don’t know, like, it really varies. And so going into the exercise, I actually have no idea.
I might have some hypotheses, especially if I’ve worked at the business for a while.
But let’s say I’m going in to do a consulting gig on this.
I’m like going in very unbiased and I’m letting the data speak to me. And it really is different at every business. So it’s hard to say.
Cool, okay. So let’s talk internal documents. What are you documenting? Who are you sharing it with?
Like, yeah, what are these different docs?
Totally, so it’s great that we talked about personas because I feel like most people completely skip all of the middle work. And they go from like having an intense data set that absolutely no one’s gonna read.
Maybe you’re a finance person or an analytics team to going to like a fictional persona.
And it’s like, Bob is an agency owner and he loves to eat spaghetti, you know? And like, do we need to know that? So what I love to create are artifacts or documents that actually educate and enable all of the teams across the business. And so for me, this usually starts with something called a segment profile, which is basically what we have on the screen here. And in this, I’m basically trying to summarize all of the learnings, the business learnings and market learnings that we know about this target segment.
So this I’m gonna be able to do after a few customer interviews.
And so usually when I’ve narrowed it down, let’s say we narrowed it down to two segments, I’ll go and do a minimum of at least, you know, five to 10 customer interviews with each segment to really pull up these insights. And I wanna share here, what’s their like top priority jobs to be done? What’s the pain that they want to be relieved by buying our software or product?
And what’s the gain that they need to have achieved? And I’m not like rambling on like the whole notes for my interviews. I’m looking out to those that people want to go into them, this is like the Coles notes version, right? And then I also wanna share and showcase what are the features in our current product capability set that matter most to this customer segment and how much are willing to pay for that.
And so this requires a bit of pricing work, but it’s actually easier than it sounds.
And what’s really fun and exciting about this type of work is this is where we start to see the nuances. Like if you get three segments and the preferred features and willingness to pay are the same across the board, they actually probably aren’t three unique segments.
You know, you probably should go back to the drawing board. So usually what we see is that the feature preference varies, their willingness to pay varies. And a couple of things you can start to do with this.
One, you can actually start to design packages that resonate. So usually I’d love to map each package to one persona or one segment.
But if you’re like, well, I’m not ready for pricing, I don’t wanna do that.
Even just how you position and market to these customer segments can be completely different, right? So it’s not to say that like you’re gonna hide a feature in app just because someone says it’s not preferred, but you better not put that as like the hero in the ad sending to them, right? Or imagine you’re gonna build, and I think we’re gonna talk about this a little bit, but imagine you’re gonna build some buying paths on the website. You’re gonna talk first about their preferred features.
You’re not gonna talk about the features you know they don’t want. And then I love to just kind of wrap it up, kind of put a bow on this whole thing with like the core value prop that speaks to each of these segments.
So this is like the elevator pitch. So if my CEO is at a conference with this customer segment, what’s the one sentence they should be saying that should be summed up here.
And then this work is kind of the work that like, it’s the blueprint for the business.
And so instead of that first slide where we saw everyone was saying something different, if we can get everyone in the company to say this slide and to know this slide like off the top of their head, then we’re in a good place.
And then there’s going to be teams that need much deeper information than what’s on this side, right? So then I work with the product teams to develop the product roadmap based on the preferred features. I work with the marketing team to develop full messaging frameworks based on the core value prop. But like this slide, is enough for everyone in the org to be aligned.
And so this slide has, here’s an interesting question. This slide has three different segments. Can they all be ICP or can only one of them be ICP? Yeah, it’s true.
It’s like, what is ideal? Yeah.
Okay.
In a perfect world, a business should have one ideal customer segment.
I have very rarely ever seen a business that lives in a perfect world.
I’ve worked with a lot of founders. And so, I mean, I think three segments is too much.
There are multi- product products, like multi- product platforms, where maybe you do have two ICPs. I have found in my experience, most businesses have between one and three.
And when that happens, it’s fine. Like I’m never gonna be able to talk the founder off the cliff of that.
But what I will try to do is prioritize. So I think there was another question I saw come in around prioritization.
And so you cannot do everything for everyone all at the same time. And so I actually don’t think it’s bad to have multiple segments. At Kajabi, actually, we had multiple segments. And we were able to describe them in one larger bucket. Like we kind of said, here’s our segment, here’s our anti- segment. Within our segments, here’s like three micro segments, let’s say. And what we did was every single thing we built on the product roadmap, we tagged to one of those segments.
And we were able to say, well, okay, this quarter we’re building 50% of our roadmap for segment one.
Does that feel right?
Or do we actually think we need to be investing in segment two a little bit more?
A lot of businesses have two segments because a lot of people will have like mid- market and then like an enterprise motion.
Both are naturally two different segments. So I think it’s ideal, but not really reasonable for businesses to have one.
Cool.
All right.
So continuing on, let’s talk websites. And I’ve got a couple of questions here. So you want to give some broad advice as well, and then we can go deeper?
So my main piece of advice is like when any prospect or customer goes to your website, they should know within like three seconds, whether they’re in the right place or not. And this is like a really great check you can do. Like right now, we go to your own website and say, if I did not know what this was, who would I think that this business is for? And more often than not, we wouldn’t be able to answer that question.
Certainly not with the first, you know, half of the website, right? Before the full.
And so what this looks like is completely different in different businesses. It’s different whether you have a horizontal play versus a vertical play, whether you have one segment versus three, whether you have sub- segments, whatever it might be. But I kind of like it needs to answer that question.
So that’s kind of like my main piece of advice that I have.
And then my second piece of advice is for any segment that you have, you need to map the unique buying flow for that segment.
And actually it starts even before the website. So it’s like, how are they getting exposed to you? The ads need to be unique, or it’s maybe a conference or whatever they’re coming from, right? Whatever those acquisition channels are.
And then we’re dropping them on the website.
And does this feel like a continuation of the conversation we’ve already started with them? Very rarely is the website the first place they get exposed to you, right?
So this should be a continuation of the conversation. And then we need to thoughtfully design the path we want them to take.
If you have one primary segment, most likely you want them to go through your homepage as their first path. But this is a great place where you can reroute someone who’s a secondary or tertiary segment to you.
And so I throw up the example of Lessonly. Lessonly actually like rebranded because I got acquired by Seesmic, but I always go back to the Lessonly website as like an amazing example of segmentation because all throughout the website, they clearly call out, we’re for you, we’re for you. If you are this person, go here on the site. And it’s easier than you think, right?
So on the Lessonly website, if you actually went to any of those where it says Lessonly for customer service, sales, talent, whatever it might be, if you went to any of those, it’s the exact same structure of the website, but the value propositions and some of the images are just slightly different.
So it’s not like they needed to code three separate pages. Like there’s a lot of shareability and scalability that you can come across here. It’s just about really understanding what needs to change for what segment.
And a related question here, how do you think about balancing cohesive company messaging with those different segments?
And
Yeah. So it is a balance.
And it’s probably a balance you’re going to have to continuously experiment with. So if you’re a business that does have multiple segments, your first job is to identify what is the common thread across all of those segments. What are they all trying to do? Or what matters most to all of them? And then usually this highest level of value proposition would be what you see in the header here, right?
Or in the hero.
So that’s the first part. And when you go to the home page of someone’s website, you are expecting a slightly higher level message. Here, I would just urge you to be more clear, think clever. We really want clarity, especially in those kind of what I’m going to call generic or general messages.
Then again, as long as you have off ramps, you can keep the more high level accompanying message at that high level and give the more customized messaging elsewhere in your website or in your marketing material. There are ways to tease in. So at Unbounce, we actually added a section in the bottom of our home page.
So it was well below the fold. It was actually below halfway, where we said, if you’re this segment, we spoke to them in one way. If you’re this segment, this way. So if they didn’t happen to go through the nav, like in the lesson example, and they got all the way through our highest level messaging, there still felt like something that was personal for them.
But the website is a unique experience.
Any other marketing material should be customized, I would argue. And so let’s talk about customer marketing. So if you’re launching a product to someone, you should not be writing the same product launch email to all of your customers.
Why would you?
You have so much more information about them. And so I would usually, at a minimum, I will have different segmented messaging. But even within that segment, I might go even deeper to have different segmented messaging, depending on where they are in the lifecycle or certain products they’ve used.
And so I always say, the level of customization you can give is totally dependent on the level of resourcing you have available.
But at a minimum, you should always be experimenting this.
How are you seeing AI impact messaging and customizing messaging?
Yeah, that’s a good one.
I mean, there are some personalization tools where you can do personalization at scale. These have always been more expensive tools than usually I work at, like, Series A, Series B type of businesses where we’re not investing a million dollars in that type of software.
So for me, it hasn’t been a reality.
But definitely, there are tools that can help you do that.
On a smaller level, though, even just AV testing landing pages is a great one.
So not to use Unbounce as an example, but it’s just top of mind. So at Unbounce, we had originally this functionality where you could AV test, right? And you built two landing pages.
And then we actually built AI into our product.
And so all you needed to do was build multiple versions of the landing pages. So I would build as many as I thought possible that I thought could resonate with my different customers.
And then you just send that out to one ad.
And the ad already kind of, like, the algorithm knows who that customer is. They’re going to route them to the right version. So that was, like, a more affordable version of AI. It’s probably definitely an area I could dig into more.
I think it’ll keep evolving.
What I’m trying to think about, I don’t know that we’re necessarily to a place where you need to have super AI personalization. But you mentioned at Kajabi you had segments, and then sub- segments, and then micro- segments. And if, as you know more about their segment and their journey, there’s going to be even more customized messaging, it feels like a tree, where it’s like, I mean, you can have so many different messages for these different segments at different points. Is that just, like, a spreadsheet?
Like, how do you keep it straight and communicate it with the marketing team and the CS team and all that stuff?
Yes, so I think, at its heart, we have, like, a messaging framework.
And we have the messaging framework for the business overall. And this is, like, we’ll use Kajabi. This is how we speak about Kajabi. Here’s the value of Kajabi, and here’s who we’re speaking about it for. So in that case, it’s, like, creators of a certain size.
Then we had three segments. And so I had three separate messaging frameworks that were basically offshoots of those segment profiles we saw. And I was like, great, segment one is this type of a creator. Here are the things that matter to them.
Here’s the problems they’re trying to solve. And here’s how Kajabi delivers that value. I didn’t go any deeper than that.
What I actually did, I would allow, like, the customer- facing teams to do what they needed to do when it made sense. So for example, if we were launching a new product and we were focusing on two certain segments for that launch, then they would do the work to come up with the messaging that connected to them.
But it wasn’t like we were constantly branching off or anything. The ad team, like the PPC team, might be testing a bunch of different messages. And as they’re testing messages, if something’s really resonating with that segment, they’ll feed it back to the product marketing team. And maybe we’ll update our core messaging framework.
But one of the problems with AI and internal personalization, as I’ll call it, is that you almost lose the strategic piece of it.
Because we can’t understand strategically what is the value they’re looking for, because that is also not just feeding it to marketing campaigns, but feeding it to the product roadmap. It’s feeding it to acquisition strategy, way more than just, like, the slogan we put on an ad.
So I think there’s a balance.
to it. The goal isn’t to have like constant personalization.
Yes.
The goal is to experiment until you find a a value prop and a message that resonates with that customer and then to constantly be running experience to make sure that like that’s not getting outdated.
It’s actually refreshing to hear you say that because it feels like attainable to have, you know, this core message for each segment and to keep that in a place where people really do reference it and really do use it. But if you start to if it starts to get too big.
Yeah.
It borders on unusable.
I will say so many people feel overwhelmed. They’re like, I need to be doing this, this and this. And the reality is most businesses aren’t doing any of it. Most don’t even have target segments and they don’t have messages mapped back to it. If you can have target segments, even if it’s more than one and you can have clear messaging and value props, you’re already way ahead of most businesses like truly.
So speaking of complexity, let’s talk different parts of the journey. Do you mind kind of walking through?
Yes. So one of the questions I often get when I work with folks in segmentation is now what like how do we actually apply this insight? And so there’s a thousand ways you could apply it, but it doesn’t mean you need to do everything, nor do you need to do everything at once. But what I love to do is I actually will build a customer journey map as part of my segmentation work. And I go all the way from acquisition. I mean, you can even go earlier to like awareness, but I was trying to simplify it here.
Basically, like the first time you come across them all the way to keeping them as customers forever. And I’ll actually map out and for every business that’s different, but I’ll map out what are the experiences that they’re having with our business. And today, based on what’s already been created, is it a positive experience or a negative experience? And this is going to go back to those key performance metrics as well, because if you see that, like for segment one, the conversion rate is like unbelievably high.
We probably don’t need to do a lot in that stage right now. Maybe over time we want to like update it a bit. But if we see that for segment one, we’re actually really struggling in engagement or getting them like to adopt the right products in the first 90 days, then I’d probably start like actually working on the engagement phase first. And so for each segment, I’m going through this and I’m kind of saying, like, is this a good experience, a personalized experience, or is it a bad experience? Is there a metric we need to move the needle on?
And then prioritizing my efforts accordingly from there.
And we had a question specifically about the end of this. How do you I mean, in days where you’re trying to expand your customer base, how do you use segmentation for upselling and cross selling?
Yes. So I’m finding definitely more and more that this is, first of all, a key ingredient that people want to use in determining their segments originally. So not just LTV, because LTV can be, you know, high if someone’s just been around for a long time and they tend to have a long tenure. But we’re actually looking at average order value or ARPU and saying or even attach rate, like some businesses are saying, well, I want to determine who my best customers are based on how many products they buy from me or how many add ons they’ve added.
And so that hopefully if this is important to you, then we’ve almost reverse engineered it from the beginning so that we have a segment who is more likely to do this. But if not, and let’s just say we use LTV to back our way into that. This could be a time where I step away from the main customer segments and I just try to understand like what types of customers are more likely to buy another product. We might have this insight out of the feature preference. And so this is where it’s really interesting.
One of the exercises I love to do is called a feature value matrix. And it’s a four quadrant model where we have table stakes, differentiators, add ons and like wasteland. And it’s really simple to produce. You basically just need to have a feature preference survey and then a willingness to pay survey. And you need to layer the two together. And so what I’ll do is I’ll do that. So if you say, hey, I want to I want to drive more upsell for segment one. I would send out this survey to segment one. I would plot it on the feature value matrix.
And I would say, wow, for segment one, all of these products seem like table stakes. These products are differentiators or these products are add ons. Can we actually now run a campaign to maybe we like move those products into the middle tier or maybe we remove them and create an add on out of it? And then we know we’re going to drive that upsell or cross sell motion because we have a clear understanding of that.
preferences. So that would typically be how I would think about this question if that’s helpful.
It is. Let’s move into more open questions and I have a couple pre- submitted but others feel free to add them to the chat. So one would be how do you know if it’s working? Like what are indicators that your segmentation is working?
So again it will go back to what are you trying to solve for. So usually this goes back to like what are the signals you even need to focus on segmentation and again goes back to the ingredients that we’re prioritizing to determine those segments. But what I’ll do regardless of whether they’re trying to drive expansion, whether they’re trying to increase retention, whether they’re trying to unlock acquisition, whatever it might be. What I do is I start to measure the performance of their business on like an average level and then on a segmented view.
And what I’m looking for is I want to see those segmented rates increase. So for example let’s say it’s like a PLG model with a trial. I’m looking today to say okay what percentage of all trialers fall into our core segments versus non- segments. And let’s say it’s like 30% segments, 70% non- segments. And we want to really drive more of our best customers. So then we’re going to set some goals. We’re going to say great let’s increase that to 35% or 40% over the course of the year.
And then we’re going to do a bunch of marketing activities top of funnel, maybe some product activities as well to drive that. And what we’re watching you know month over month or even week over week is are we’re getting more of our best customers in trial than we were before. Now let’s we also want to make sure that as we’re getting more of them we’re converting them. So then I’m looking at trials of paid conversion rate from the segmented level and the average level.
So I’m saying okay for folks in our segment let’s say their average conversion rate is 50% and we want to increase that. We’re doing we’re working with the product team and the customer lifecycle team to drive the right programs right. Do we have the right app cube pop- ups going up in app. Are we stirring them to the products and features that we know they prefer because we did the preference survey. Do we have the right support model because hopefully we also understand how they want to be communicated to.
And we have all these programs and initiatives running that will ideally increase that as well. And then you’re just going to kind of see the dominoes fall in all parts of the business. So truly usually my preference and if I’m working with a business long term I kind of break down the full customer lifecycle metrics. Customer like segmented customers, non- segmented customers. I’m tracking and measuring both and then prioritizing where we’re moving the needle and where our like marketing or product efforts are that quarter.
Awesome. Another one. How do you adapt segmentation to accommodate evolving market trends? So maybe that e- commerce example like rising e- commerce.
Yes.
Decreasing other segments is a good is a good segue.
Yeah so I think if we’re in like such a fast- moving market and things are changing constantly technology is changing. So people often think that segmentation is like a one- time activity like we did segmentation. This is something that I like to have someone in the business feel ownership over the segmentation program at all points in time. This should not be something you’re sleeping on and like waiting until you feel like something’s broken in the business before you do it.
It doesn’t mean you’re doing all of these activities all the time but I’ll have someone just kind of constantly looking at market trends, looking at competitive trends for the markets that we’re very interested in. It’s like the markets that we’re playing in today or as I like to think about it like what are the ponds we want to fish in tomorrow? How are those ponds growing or shrinking? And so the question inevitably would be well we don’t have time for that. We need to focus on the business today and we need to hit our growth goals.
And it is just a balance like there are tools that you can use like there’s trend tracking tool even something as simple as Google Trends. Truly you can track really interesting things like certain phrases that people are searching for, the rise of certain job titles, whatever it might be for your market. And then oftentimes my business, the size businesses I work at, don’t have like a ton of resources internally to do our own market research in terms of like surveys.
So what I’ll do is I’ll find the people who are doing that research that I really believe in and I will follow their surveys and reports and I’ll try to track them year.
because I’m still following the same methodology, right? I’m, even if I don’t 100% agree with their methodology, I’m getting the signal of, are they thinking that it’s growing or increasing or decreasing? I’m looking at new trends that they’re surfacing. And ideally I’m validating them across a couple of each other. So at Kajabi, we found three different reports that we believe that were well- respected.
And what we did was we actually saw three different reports had slightly different numbers, which is fine and to be expected, but they were all within a range that we found to be acceptable. And so it’s a little bit of like knowing your miss, like kind of agreeing on like, here’s what we need to get perfect and here’s what like we know we don’t need to get perfect and finding a balance. Awesome.
Another interesting one, the examples feel more SMB. How is this different if you’ve got enterprise and there are a bunch of different people within the company, different departments, different people who are part of the sale?
Yeah, great question. So couple situations, but the process is the exact same. What I have found typically is that in enterprise level businesses where the end customer is enterprise, usually the data set is smaller. So I actually worked with a client earlier this year where they had a data set of 200 customers, right? And that was over a course of a couple of years of data. That’s okay. Like 200, we’re not getting stat sig with 200, but let’s acknowledge that.
But we are getting way more signal and insight than if you never looked at the data set to begin with, right? Like usually in those businesses, people say, well, our data set doesn’t matter. Like, this is what I’m hearing on calls. And it’s like, it does matter, right? Even if we’re just validating your hypothesis and proving it true, that’s fine. So the data set is smaller, which means we probably need to spend more time doing one- on- one interviews than if the data set had stat sig.
So that’s like my first one, but the concept between a segment and a persona doesn’t change. So this almost goes back to the question we said before. What we’re not doing here, like in the example of an enterprise that has different departments, those aren’t different segments.
Those are different personas, right?
And so a different segment might be mid- market versus enterprise. It might be B2B SaaS versus an e- com business.
Like those might both- What about like a less than lead that’s trying to sell into sales and customer success?
Yeah, I think that’s fine. And then basically what we’re doing is we’re looking into the different personas within those. So we’re saying, okay, if you’re selling, if you’re a company with sales needs, what’s the personas within that? It just means that more often than not, the buyer is different than the user and you’re gonna have more personas than you are if you’re an SMB business, but you should not have more segments. Like the recipe is the same.
Got it, okay. What if you have really strong fit, but in a niche and you have to grow?
Yes. So I love to use the analogy of like the game of risk. And so in the game of risk, we’re trying to have a world domination, but let’s say like you have captured Australia. Australia is the easiest place to capture in the game of risk, right? So you have Australia, you’ve fortified your bases, but you wanna expand. You’re not gonna expand overnight to North America. It’s on the other side of the risk board. You have no way to get your men there. And so I was like to say, what are the properties adjacent to Australia?
And so I use this metaphor to say, look at who has things in common with your niche segment today. So this could be similar problems, jobs to be done. It could be similar business side. It depends on how you’re segmenting your market today, but you wanna find someone that like your current product will be able to meet their needs. Your marketing is not gonna have to change drastically. And by including them into your segments, you’re not gonna have to fundamentally change your homepage or whatever it might be. And so adjacency is your friend here.
And then it is sometimes a bit of experimentation to know which adjacent market is going to be the most lucrative or interesting one to you. But I would say, don’t try to go after something completely different. Like use your core competencies to win that next market.
So often when you think about expanding, you can expand, you can sell more things to your existing clients or you can sell to new clients. I guess if you’re addressing a new segment, this is new clients. Do you have perspective on when you do each?
Yeah, I mean, I think that sometimes comes down to what you are selling and if it’s even possible, like if you want to be a platform, for example, then yeah, you’re probably sticking with a fairly niche segment and you’re building multiple product lines. And that works for many businesses, but then some businesses want more of like a horizontal play and they’re like, we could be for anyone. So we wanna keep capturing more markets. And they have like one kind of limited product that could just be used for different use cases.
It’s hard to say without knowing the product, but this is where really like the product vision, like you cannot make a separate market decision here than a product vision. People often ask me like, who should be leading these types of exercises or work? And generally for me, I think product marketing needs to be like the leader of this, but product management needs to be at least along for the ride in the passenger seat. If not sometimes, you know, really kind of building the map and steering the driver too.
Often, segmentation is a great platform for account- based marketing. If you’re doing account- based marketing, how do you think about, I guess, building a database, enriching it, segmenting it? I guess, how can you take your segmentation and feed it into an account- based marketing program?
Totally. So I have worked with folks where this is the case, and this isn’t really… So when you know that this is where you’re headed, it’s an interesting game too, because we can say, well, what tools might you use for ABM, and what are the fields of data available in those tools? Because we might define a segment, and then it’d actually not be targetable, right? And so there’s a few things I look for in segments, like it has to be measurable, it has to be targetable, it has to actually be clear and distinct.
And so if you’ve gotten to this point, but you actually can’t target them, because ABM is your motion, and let’s say you’re using a LinkedIn navigator tool, or whatever it is, so usually I know kind of like the go- to- market motion of my clients before. So let’s say you’re doing ABM, we’ll actually reverse engineer and say, great, we need to make sure that we know this about our segments. And then ideally, if we’ve done the segmentation exercise right, we’re just giving you now the targeting criteria that you can put into the tools.
And so it really is about just making sure that that target segment is large enough, again, to sustain your growth ambitions, and that we know where you can reach them, whether it’s LinkedIn, whether it’s a different platform. But I see ABM as just a campaign that comes out of segmentation.
Awesome. And then if you apply that also to paid marketing, what are your favorite ways to apply segmentation to paid digital marketing of all different types?
Yeah.
Paid marketing is harder, because the more you target, the more expensive it gets. So I think about this a few ways. When I have bucketed my customer segments, I work with my acquisition team in general to really understand the acquisition channels, to understand where are they coming from. Are they coming from paid? Are they coming from SEO? Are they coming from events? Wherever it might be. And then if they are coming from paid, in a world where we have attribution and we can go back and actually see what campaigns they came from.
I say this because it’s not everyone’s reality, but I have worked places where we can see that. And we’re like, wow, they came from this campaign. Then oftentimes, it’s more just saying, well, this campaign is more effective than this campaign for this segment. Let’s put fuel on the fire. Let’s shift budget here. One example from Kajabi, another SMB example, unfortunately, but just happens to be in my sweet spot.
What we found was that paid was working for all of our segments, but when we actually dug into the segmentation, we saw certain paid channels were more relevant. Some people were only on meta, and other segments were like, wow, meta sucks for this segment. So we turned off budget for that segment immediately, and we reallocated all of those funds to a completely different go- to- market motion, to YouTube, actually, paid ads.
And so it’s that level of strategic conversation where it’s not just saying, well, how much budget are we spending in paid advertising this month? It’s saying, who are we trying to attract, segment one or segment two? Great. Let’s actually work back to how many of them we need, and we know that segment one is going to come from meta. Great. Do we have the right budget to get the right numbers? We know segment two is going to come from YouTube ads. Let’s make sure we have the right numbers there to drive our targets at the end of the quarter.
So that’s helpful. Another thing I like to add is you can work with the paid team to test value propositions. And so I will feed them certain value propositions or messaging that I want to see, and because the paid ads get such scale, you can waste some dollars if the message doesn’t resonate, but you’ll get that feedback loop within a couple days, and then you can quickly shift budget. And then I like to incorporate that on the homepage or in other campaigns. So that’s a fun way to play with paid.
You mentioned at the beginning, this was interesting, meaning this answer, that the more you segment your paid advertising, the more expensive it is. Is it ever better to not give meta as much detail and spray it wider, or is it always better to focus more tightly?
I am not a paid expert, so I’ll say this with that caveat. I think it depends on what you’re trying to do and where in the buying journey. So now it depends on how much you’re able to do your buying journey. I have seen for some businesses that meta is really important and we need a meta touch point to convert that customer, but actually it’s more in the awareness stage than the like consideration stage, in which case I just go high level message and I am just going really generic, honestly.
I’m not going to anyone, but I’m going like the highest level customer that could ever buy our product and it’s a high level general message. But for customers where I, or segments where I know that it’s like more of a consideration, it’s maybe the last click before they hit on our website, that’s where like I better be pretty targeted. It’s gonna cost me a little bit more money, but I just, because I know the role in the buying journey, I’m more willing to spend that.
Awesome. All right, possibly the last question here is, how does segmentation feed into sales enablement?
It’s like sales enablement is one of the biggest outcomes of this. Again, understanding that buying journey, understanding for that segment, and this goes back to the enterprise question, for that segment, who’s involved in the buying journey? So I’ll actually work with sales to map out what are all the personas? And say you have three personas, that’s not unusual in a large enterprise deal, but where people get it wrong is they don’t understand who needs to know what, when.
And so I’ll actually map out, great, who is the core person who’s coming to you? They’re instigating kind of the sale. That’s oftentimes the person who’s gonna be using it, or their teams are gonna be using it. And I’ll help them understand what are the criteria they’re looking for? What are the messages they need to hear? Then who else needs to be involved to capture the sale? So usually there’s someone in finance who needs to sign off on it. And they actually don’t need to know, they’re never even gonna go to your website.
They’re just going to double check to make sure you’re a real business, and then they wanna know risk, they wanna know budget, they wanna know terms. And so I’ll build sales enablement to educate that person in a different way. Because what happens is if you don’t understand those personas within that segment, you might over inundate that person, and they might feel like, there’s too much, this is too complicated of a deal.
I was just at a conference last month, I’ll share this one thing, because I thought it was such a fun trick, where they were a software company selling to enterprise teams, where there’s usually three buyers, or three people involved in a sale. And they did this exact thing where they’re like, okay, the core buyer in the segment, that’s who we’re building our website around, that’s our pitch deck is for them, all of that.
And then they would run paid ads on LinkedIn, targeting that financial buyer at the business, just so the financial buyer, super, there’s no CT, just so they saw it, so that by the time the paper came across their desk, they were already familiar with the business. And I was like, well, that is genius, I’m gonna borrow that technique.
So I think- So this was after they had already, there’s an open deal, we’re gonna start advertising.
Exactly. I love it. It’s so smart, right? So I think there’s a lot of fun, clever things you can do. And the reason they were able to do that, is because they knew that that buyer is in a preferred customer segment, they knew that they were willing to spend more money on them. So segmentation plays into all of this.
So with sales, there’s always a keep it simple, stupid component. How do you make sure that this is actually consumed?
Yeah, depends on your sales team. I mean, I like to co- create the sales enablement with someone from sales, because otherwise it’s never gonna be used. It’s gonna be like, I don’t even know, I’m not an expert on sales, you’re the expert on sales. And ideally, they should also be the expert on this buyer. If you were coming out with segments, and your sales team is like, you’re out to lunch, you’ve probably got it wrong somehow, right? These should be real buyers that are coming across your plate already.
And so, first of all, I’ll try to involve a representative from sales for this whole activity, so that they’re already bought in and like pumped about this by the end. And then I work with them to say, which segments are harder to enable than others? Like, what do we actually need to close these deals from beginning to end? And we’ll just like start chipping away on the playbook for them. And again, you can’t do everything at once, but I kind of let sales guide what they need.
And then I work with them to make sure that the message is right, the visuals are right, the product features that we’re highlighting are right at the right time based on those insights that have come our way. So it’s a partnership.
Awesome. Well, thank you so much for taking the time on a Friday. It’s been awesome talking to you. We will share the recording of this full thing with registrants as well as some recap notes. So thank you so much.
Thanks for having me.
Slides
Key Takeaways
- Strategic Alignment: Segmentation should align with business goals and customer needs, helping to prioritize efforts and resources effectively.
- Customer Journey Mapping: Mapping the customer journey is essential to understand customer experiences and identify areas for improvement.
- Paid Marketing Strategy: Targeted campaigns can be more effective but also more expensive. It’s crucial to balance targeting with cost-effectiveness and adjust strategies based on performance.
- Sales Enablement: Segmentation plays a crucial role in sales enablement by helping sales teams understand buyer personas and tailor their approach accordingly.
- Market Trend Adaptation: It’s important to adapt segmentation strategies to evolving market trends, staying informed and adjusting strategies as needed.
- Segmentation vs. Personalization: While personalization is valuable, constant personalization can be overwhelming and may lead to losing the strategic focus of segmentation.
- Focus on Core Message: Having a clear core message for each segment is crucial, and over-segmentation can make messages less impactful and harder to manage.
- Measurement and KPIs: Using segmented metrics to measure performance is key to understanding the impact of segmentation strategies and optimizing them over time.
- Experimentation and Adaptation: Constant experimentation is necessary to find the right value proposition and message that resonates with each segment, and to ensure that these messages don’t become outdated.
- Balance and Prioritization: It’s important not to try to do everything at once. Prioritize efforts based on segment performance and focus on what will have the most significant impact on business goals.