Intro: 00:02

Welcome back to the Response Drivers podcast, where we feature top marketing minds and dig in to their inspiring stories. Learn how these leaders think and find big ideas to push your results and sales to the next level. Now let’s get started.

Rick Rappe: 00:19

Hey, I’m Rick Rappe, host of the Response Drivers podcast. Here, I dive deep with marketing experts and innovators to learn how they approach targeted marketing and use data driven strategies to acquire and retain customers. We’ll talk about what’s working, what’s changing, and how we can stay ahead in an evolving marketing landscape.

Response Drivers is brought to you by RPM Direct Marketing. RPM helps companies develop hard hitting, direct mail, creative and utilize advanced testing and targeting methodologies to reach customers and prospects. Our goal is to fully optimize your marketing performance to drive more sales and exceed growth expectations. RPM delivers smarter, more profitable direct mail solutions so you can turn your direct mail programs into a predictable, efficient sales channel. Check out rpmdm.com to learn more.

My guest today is Bruce Meberg. Bruce brings more than 30 years of experience spanning aerospace engineering, management, consulting, marketing, analytics, and data science. He founded his company, Inbound Insight, to apply that analytical problem solving approach to one of marketing’s most persistent challenges finding and delivering the right prospects because better audiences drive better campaign results. Bruce, thank you so much for joining me today.

Bruce Meberg: 01:38

It’s my pleasure, Rick. Good to be here.

Rick Rappe: 01:40

All right. Well, you started out in aerospace engineering. And how does an aerospace engineer end up becoming a marketing data entrepreneur?

Bruce Meberg: 01:49

Well, it did take a few steps to get there. So I’ll just give you a real brief background. But yes, my first job was as an aerospace engineer designing communication satellites for General Electric. I was very fortunate to have that job right out of college. It was a fantastic job, but I knew fairly quickly that I wanted to move into the management or marketing space.

So I did get my MBA part time, and while I was there, I got my degree and went on to management consulting after that. So that’s a first few steps in my path. But then when I was in management consulting, I started realizing this was the boom of.com in the late 90s, that everyone was kind of doing their thing starting companies. So I thought, I want to do that. I want to start my own company.

But I realized also I didn’t really know how to raise the funds needed and do that at that point. So I actually joined a friend of mine who was starting his own data marketing company at the time, and we grew the business. We raised money from venture capitalists and grew the business. And eventually part of it was merged into the company. Acxiom, one of the very large data compilers that many of you may know about.

They probably have the best consumer data in the country. And so they’re a terrific organization. About ten years ago, they came out with their first, what they call data services API. So the API means that you can request one bit of information and get back an answer very, very quickly. And, and that’s become very, very common these days.

But I saw the potential of that and realized that this was something that was going to change a lot in the marketing space to be able to, to get insight like that appended or enhance your data in various different ways. And that’s a core part of what we offer today. We offer something called insight APIs. But that wasn’t really Acxiom’s thing. They wanted to use it internally for their very large clients like Citibank and others like that.

And so I felt that that would be a terrific time finally for me to, to really go out on my own, use that as the first bit of technology productize that for more smaller businesses, medium size agencies, direct mail printers to use in the context of a platform. So that’s what I did about five years ago. I took the leap and created that solution. And that was the beginning of Inbound Insight.

Rick Rappe: 04:10

That’s awesome. Well, congratulations on taking the leap and becoming an entrepreneur.

Bruce Meberg: 04:17

Thank you. It’s, it’s, it’s a, it’s a fun and challenging task. And I wouldn’t change it for anything.

Rick Rappe: 04:25

Was there anything in your career in consulting or at Acxiom that you learned that was really impactful on your direction?

Bruce Meberg: 04:34

Oh yeah. Absolutely. I mean, Acxiom, you know, they have terrific data resources, terrific ways of analyzing existing customer data to find, you know, who is the right type of prospect to be going after something typically called like an enhanced lookalike report or a or look alike report. So that’s certainly something that I, I learned while I was there and you know, beyond that, you know, working at, at a prior data company and, and being kind of the top executive there, launching new products, launching new markets. I mean, that was all part of the background that helped me prepare me for this journey.

Rick Rappe: 05:18

That’s great. Well, here at RPM, we have a performance framework that we like to take clients through new clients especially, and it sort of helps us break down the performance of a direct mail program into all the different levers that we can pull to impact performance and targeting is on the top tier, and it’s probably the number one thing that we look at when we’re thinking about direct mail performance is hitting the target audience and making sure that we get the most pinpoint targeting that we possibly can, finding those people that are likely to respond and eliminating those people that are not likely to respond is a big, important aspect of all direct mail programs. So I have a lot of questions for you about how to find the right target audience and how to make direct mail more successful. And you’ve been on the data side of direct mail for such a long time. What do you what are the things that you think make direct mail, direct mail campaign really successful?

Before the piece is ever designed?

Bruce Meberg: 06:26

Yeah. I mean, obviously direct mail is a left and a right brained endeavor. You’ve got the advertising, you’ve got the perception, you’ve got the positioning, all the creative aspects, the right brained aspects of things. And that’s, that’s another group. I mean, outside of that’s, I’m sure that’s part of your team’s efforts, but we’re focused really more on the left brain, the data side of things.

So we’re doing everything we possibly can, like you said, to find who are the right people to reach who, who, the wrong people that you don’t want to reach, which can be just as important in some contexts. And then in addition, we also try to do what we can to help with personalization by having some insight about a person, whether it’s even things like, do they own their home or have children? Or do they are they which persona do they represent? Because each customer file has multiple personas and we can help tease that apart. But then getting into the really the, the meat of it, which is the targeting, you know, there are several ways of going about that.

And that’s really been our core focus for the last few years is to find some more advanced ways beyond a traditional demographic, targeted approach, which is a good and necessary right step. First step, but certainly not the last step. You can take there’s many other ways to find the best audience that’s going to be most responsive, convert at a high rate and give you a high ROI on your campaign.

Rick Rappe: 07:53

Yeah. Well, that leads into my next question a little bit, which is what do you think marketers are getting wrong about targeting and prospecting today?

Bruce Meberg: 08:03

I think, you know, marketers today, anyone under 40 or even 45 have been brought up in the digital era. Right? So that has been a huge influence. I mean, it used to be all basically direct marketing, whether it was direct mail or even phone calls in the distant past before that industry went away. Although not for political calls, obviously, these days, but.

And then you had the pure broadcast side, right? So that was like, okay, let’s just like get a lot of cheap impressions on a TV or radio. So, but, but, you know, the last 15 or 20 years, it’s been really more about digital and that’s been good and bad. I mean, I think the bad side is that it’s it’s gotten people away from taking the responsibility of doing a lot of the targeting because you can kind of rely on the black box optimization that meta or YouTube or Google can do for you. And so you didn’t have to like think about it very much.

And they have a lot of great data. Of course, it’s in their closed box ecosystem. So, you know, it’s good and it’s bad. But the thing that I like least about that, and I’m diverging a little bit here, but is the fact that those audiences are anonymous to you at the beginning. I much prefer knowing the names and addresses of the people that were reaching out to, so that at the end of the day, when the customers inquire or convert, you have a way to match back who who from my specific audience did convert and do a very good analysis on that and see how you can continuously improve your campaign in the future based on what you learned from the first and second and campaign, etc..

Right. So.

Rick Rappe: 09:40

And in addition to having that data about their physical address and name, You also can have all these other data points in addition to that. I when I think about postal databases and the big compilers of data, I mean the. I think the amount of data that we can get to with a postal database is astronomically high. We’re covering almost every available address in the United States. Every household.

And then there’s thousands of data points available on each person or each household.

Bruce Meberg: 10:18

They really are. I mean, in terms of like the core demographics, it might be a dozen or two dozen that are really the ones that people use all the time from a demographic point of view. But then there’s a lot more information. There’s things called propensity models, which are pre-made models based on maybe known information about a sale of cars that gets published or monetized by, by a company. And you can create a model of who’s buying Fords versus who’s buying Lexus.

So you can create that model and put it out there ahead of time. But there’s many, many other uses. These databases have kind of converged into two other two other solutions. One is based on the ability to do a predictive modeling. So you have thousands of data elements.

You take your customer list, which is what we call the attainment cohort, the one we’re trying to get more people like your best customers. And you can create these predictive models based on those individuals, comparing them to the non-customers or campaign responders versus campaign non-responders. And to do that, you need to have hundreds or thousands of data elements in your database along with every name and address. So it’s kind of like the old consumer data databases as well. Now you’re recognizing who that person is and then tagging a thousand bits of information to them.

And now you can use your regression analysis to figure out, well, what’s the right formula? Weighting all these different factors that when a new name and address comes in, I append a thousand data elements and can figure out their score of are they 100? Which is the best or one which is the worst. So in terms of likelihood to convert. So they’re taking the core model of databases or they call them identity graphs.

These days. It’s a very popular term for them and using them in extending the use of that. And that’s one way of doing it. With the predictive modeling layered on top of the database.

Rick Rappe: 12:12

We’ve had amazing results with predictive models. And yeah, it’s a very powerful tool. I think there’s a lot of things that we could get into and unpack about that. I mean, your quick overview was very good, very good for me because I’ve, I’ve experienced predictive models and kind of know what to expect. Maybe you can talk a little bit more about that.

For people that haven’t experienced predictive models, I models. I kind of want to unpack it and help them to understand the power of this tool. So you’re saying like, if I have customers, a customer base, and how many do I need? Do I need 5000 customers?

Bruce Meberg: 12:52

Yeah. 5000. Yeah. 5000 is, is, is good. Ten is better.

Three is acceptable. So that range of and you and, you know, try to go back a few years, I’ll go back 20 years because people have changed a lot in that time frame. So, you know, fairly recent. And if you have a little more headroom, like if you have 50,000 and you want to give the best 10,000, you know, pick the ones that buy, you know, kind of an RFM basis, you know, are they spending a lot? Are they buying multiple times that sort of thing?

Right. So that would be the basis of it. And, you know, the big difference between the predictive model, it’s really kind of like the old lookalike approach analysis, but with much, much more data. And it’s a multivariate. So in the old case, you might say, okay, let’s look at our customer file and let’s append age and income and other dimensions to them.

And let’s, let’s not just say, well, they tend to be, you know, this age range or that income range, but let’s look at it relative to the market that you’re in. Because if you’re in Greenwich, Connecticut, versus, you know, somewhere in some other areas of the country that aren’t as wealthy, you’re going to have very different results. So you kind of have to reference it to the local market demographics to, and then you can find what we call an index, which is where, okay, this age range or this income range index is higher than random. So you might have 5% of your, of your, of your customers or say 10% of your customers might be between 100 and 125,000 of income, but the local market only has 5% in that range. So that would be an index of 2 or 200 because you like to make it where 100 is on market research is like the average.

So it’s an index of 200. That’s great. But that’s one dimension at a time. And so then let’s say you do that analysis and you say people in this income range combined with that age range, that’s my target market. Okay.

What you’re assuming is that everyone inside that is great and everyone outside of that is bad. You know, everyone look at as dots, you know, representing each person, the green on the inside and the red dots around the outside of that two dimensional box. And it can be more dimensions, but it’s simpler to talk about it as two. That’s not how it really works. You have greens in the box, but you also have yellows and reds in the box, and you have greens outside the box.

So the best thing to do is to kind of line them up in a descending order of green to yellow, down to red, right. And that’s exactly what the predictive model does. And it does it in a multi-dimensional way. So 25 or more dimensions, not one at a time. We have to use a kind of a guess and a human, you know, input.

This is all done through regression. And then the models are created and actually the models get tested against each other using machine learning. So it’s a terrific approach to, you know, really getting you to the very best in terms of the right demographic and interests and home property, etc., values.

Rick Rappe: 15:52

And it’s the combination of, of multiple things that are predictive and putting that to putting together that scoring algorithm that gets basically put together and tested and then optimized and improved and tested and optimized and proved and tested repeatedly, very quickly with machine learning. It used to be years ago when I was working with Acxiom on the AT&T program, we had a team of statisticians, and doing these predictive models would take a month. Yeah. And now it’s much, much quicker and much more.

Bruce Meberg: 16:29

And we got to the moon using slide rules in analog. So yeah, it’s come a long way.

Rick Rappe: 16:35

It was done a long time ago, but it was not done with the speed and efficiency that it’s done today for sure. Definitely.

Bruce Meberg: 16:42

And the precision there’s there’s just a lot of data out there for sure.

Rick Rappe: 16:46

Right. How has data precision changed over the years as has that gotten better and better over time?

Bruce Meberg: 16:53

Yeah, I think so. I think, you know, I mean, we talk about Acxiom or I think they’re a great company. They have 100 separate sources feeding into their info based database. So, you know, and they’re cross-checking, they’re corroborating. I mean, there’s, I think there’s been an explosion of data.

And especially when you get onto the digital side of things. And then even like first party data, which we can talk about as well, but, you know, on the digital side of things, another terrific audience because you know, what we’ve talked about here so far with predictive models, and we said, it’s great. And I have I’ve been using models for over five years, and I’m constantly, consistently seeing at least a 30% boost in conversion rates compared to the very best you can do with demographic, you know, age and income type models, or even segmentation systems like Persona or Connex or P cycle. All those, the predictive models blow them all away. So if you’re going to do one thing, that’s a great step.

And the other great benefit is that you have the full universe to work with. You know, you’re not limited. You still have every single household in the country that you can again, put in rank order and say, I’m going to start with this group and then work my way down, you know, first percentile, second percentile, first decile. Work your way down. That’s a fantastic approach.

Rick Rappe: 18:06

And so.

Bruce Meberg: 18:06

One thing.

Rick Rappe: 18:07

When you I want to just unpack that briefly. Yeah, you we create this scoring algorithm based on this predictive model based on your current customer base. And you back test that against the current customer base to make sure that it’s working correctly. You sort of build it with half of the market and test it with half of the market or half of that group, if you will, to make sure that it’s actually predicting. And then once you have this optimized predictive model, you’re able to point it at the remaining universe of prospects in the country or in your area, and you can score them based on all based on the criteria that are predictive of your customer customers.

And then it literally rank orders them and from top to bottom of the most likely to respond to the least likely to respond. And so then you can focus your campaigns on, depending on how much your marketing budget is or how aggressively you’re trying to grow, you can target the top 10%, 20%, or you can just eliminate the bottom 10 or 20% and optimize your campaign by taking out the people that aren’t going to respond anyway.

Bruce Meberg: 19:20

So exactly, that’s exactly right. And now if you’re a national company, you could probably just work within the top 1% or 2%. Yeah. Regional or local. Because of, you know, distance constraints to a physical location, you might have to go down to 20 or 30%.

Yeah. Within your radius of your service territory or your counties or counties that you work in. So. Yeah. Yeah.

That’s exactly right.

Rick Rappe: 19:41

I just to take it even one step further, we’ve even been doing marketing for a grocery store chain. And, you know, they have radiuses around each store, which is a drive time radius. Well, we were working with drive time radiuses around their stores, and we overlaid a predictive model on top of that. So we still use, and we were able to expand the radius a little bit, but be much more selective about the type of person that we were targeting. And we were able to achieve a huge lift in performance, bringing people from a little bit further away.

But, but hitting the right type of person and the right people that are going to respond. So yeah, it’s very powerful.

Bruce Meberg: 20:19

There are two of the three really important dimensions. You have the predictive nature, you know, the right demographic mix. If distance, obviously for physical location, brick and mortar type businesses. And the third dimension, which I want to get into next is time timing, right? Those are the three most critical.

And distance is, you know, obviously doesn’t apply to every single company, although I would, I would say that there are, oh, even if you’re a national firm, you’re still going to have pockets where you’re doing better than others that are not based on the demographics there, but based on word of mouth or other factors that go beyond demographics. So for the timing aspect, there’s two solutions. I’ll call it timing slash engagement. So one thing that’s come on really big, we’ve been doing it for for several years, but it’s kind of really expanded just in the last year or two is the ability to take in market activity, people browsing on a third party website for a particular product, you know, home and garden magazine, looking for all the things that they have ads about and content articles about furniture or, you know, fixing up your, your patio or whatever it might be. People are browsing that content.

They’re researching, they’re looking for, you know, if I’m looking for a new patio furniture, I’m going to be going through all these kinds of sites and that information gets picked up. It gets identified and it gets monetized. So you have these online intent signals. Now we have access to 29,000 of them, mostly B2B, but a lot B2C and but a lot of B2B as well. And so now we can tell you, for example, like we have a big news organization that wanted, they tested three different categories.

It turns out the one related to art, like art was like the best one for them. So they’re getting like 5 or 10,000 leads every month. They’re doing email campaigns for that. But they found that that particular topic worked out really well as a great audience for their news, in this case, news website. So 29,000 categories, you have a lot you could do all kinds of home services.

I mean, you could do all kinds of retail, personal services. There’s just there’s somebody there’s thousands of people right now researching your product or service and in preparation for buying it in somewhere in the near future. Yeah. And you know, we even have one, which is a retirement community in Florida, which did a terrific campaign based on a predictive model. But when you think about it, people 55 and above, even if you’re like in a feeder state to Florida, like New York or Pennsylvania or New Jersey, there’s a lot of people that are kind of in your ideal customer profile, but they’re just not ready or in market or in timing and timing for it right now.

So by identifying them and it’s going to be a smaller number, it’s only like, you know, a few percentage of the total universe, but that’s kind of accurate in terms of what percentage is really looking to do something right now.

Rick Rappe: 23:21

Yeah. Yeah. When we when you target the people that are shopping. I mean, you’re also going to expect a multiple higher response rate and a multiple higher conversion rate. Yeah.

As opposed to just hitting everybody out there that isn’t or may or may not be shopping or interested in that particular thing. Exactly.

Bruce Meberg: 23:38

Yeah. So certain products timing is more important than others, but like bigger ticket items, long purchase cycles, you know, those are the kind of things where timing becomes really critical to have a great result.

Rick Rappe: 23:49

Yeah. Well, I think about these. I mean, targeting is actually, I think of the who, who are going to mail to, who are not going to mail to and, and the creative and the offer and things like that. What it’s going to be in the mail pieces is the what, what you’re going to send out. And now this third component is the when, when are you going to touch, when are you going to talk to them?

When are you going to spend your money to reach them?

Bruce Meberg: 24:13

And exactly.

Rick Rappe: 24:14

So it’s so important. And direct marketing. Let me ask you a quick question about that. Yeah. What are the signals that you can look for that they look for to find people that are interested in a particular topic or in a.

Bruce Meberg: 24:29

Well, I mean, it’s, it’s, it’s all sorts of data that, like I said, you have these identity graphs, which are picking up literally billions of signals every day from across the internet. So there’s a lot of information to work with there. And then, you know, the key then is to be able to take that and figure out, well, who is that person that I can then reach? So, you know, that’s what I said. There’s kind of an explosion of this kind of data really coming from the digital side, but now it’s translating back into the traditional direct mail email space, just in the last year or two.

Like I said, it’s been there before, but the number of categories was more like in the hundreds, not the tens of thousands. And that’s been the big difference. So that’s something we’re very excited about, to do more projects like that to help with the timing aspect.

Rick Rappe: 25:18

Yeah. Would you say it’s still very important to do testing of these?

Bruce Meberg: 25:23

Oh, absolutely.

Rick Rappe: 25:24

And targeting ideas. I mean.

Bruce Meberg: 25:26

Testing.

Rick Rappe: 25:26

Is testing.

Bruce Meberg: 25:27

Testing is the foundation of everything. I mean, you know, a lot of times you might try something, but if you haven’t done it in the context of a test plan, like I’m going to do, you know, this audience with creative a and that audience with creative B and, and the number two audience with creative a and make a grid and, and really plan it out, you’re just not going to be able to learn. I mean, the first time out is, you know, is not going to be your best optimum campaign. It’s almost impossible. It’s 1 in 1000 chance.

So you need to like work with all the factors that are important and narrow them down, see which ones are truly important, which ones are not. So yeah. I should have mentioned that sooner, but from a direct marketing direct mail, especially point of view, which is such an expensive medium, you have to really focus on the right way to launch the campaign in a, in a thoughtful way like that with testing built in.

Rick Rappe: 26:17

Yeah. And our performance framework, we’re we talk about the things that drive performance. But then at the bottom of it, it’s all it all has to be built on this foundation of structured testing and detailed response analysis and detailed analysis on the back end. But you can create a system with the right approach that gets smarter and smarter and smarter over time. But a lot of people get that wrong.

They get that part wrong. They throw out one thing or three things and they they test one list to one creative and another list to another creative. And they can’t figure out what is it that’s impacting performance. So it’s a, it’s a common mistake that I see on the agency side that people just don’t structure their testing very well or correctly.

Bruce Meberg: 27:06

Yeah, absolutely. I, I almost, I almost switched careers when I was a management consultant to be a quality consultant. I actually studied under W Edwards Deming in graduate school, who basically taught Toyota about quality in Japan after World War two. And, continuous improvement. There’s a term called Kaizen, which is continuous improvement that is so important.

And that needs to be built into all of this. Absolutely.

Rick Rappe: 27:35

Absolutely. Well, we’re singing from the same hymn book, I guess. Yeah.

Bruce Meberg: 27:41

Our brothers.

Rick Rappe: 27:44

And I know you also do a lot with engagement and targeting the people that are engaged and shopping and, and showing interest. Talk a little bit about that. Yeah. How you capture, capture people as they’re as they’re showing interest.

Bruce Meberg: 28:04

Absolutely. Well, the like I said, the in-market leads, as we call them, that’s people searching elsewhere on the web. But there’s another tool, which is the first party data from your own website. So we have a tool called Visitor Connect, which is how you can putting a pixel on your on your website capture, who is not filling out that form because only about 5% do or who’s not buying. Or maybe you don’t have an e-commerce website.

So that’s not irrelevant. But, you know, we’ve had some really great success with companies in many, many different industries from small to very large. We had one, one client who generated 99,000 leads in one day. So wow, that was a big one. Sorry.

Rick Rappe: 28:45

From their.

Bruce Meberg: 28:46

Website. Yeah, it was, it was from their website. It was giving Tuesday, and it was one of the top ten brands in the kind of charitable organization space for children’s charities. And they had a huge day on Giving Tuesday, which is the day after Cyber Monday. And so that was 99,000 leads captured in one day.

But we can we can typically capture most of the most of the people that visit the website And, and also it’s super important because you mentioned engagement, you can have like a minimum floor of engagement, like how many minimum pages or time on site you can, you can have it be so that the pixel only fires on certain pages or gets overridden on certain other pages. So you have a concept of the abandoned cart, which would only start when they get in the cart on that URL, that page. But if they, if they actually convert and buy, then it overrides the pixel from firing because you’ve already got that person as a new customer, you don’t need us to give it to you at that point. The same thing works with charitable donations where you have a donate path abandonment concept. But, but really any, any pages on the website that are more likely to convert than others can be, can be programmed.

It’s a very simple setup with the pixel and you can get started very quickly to generate leads that way.

Rick Rappe: 30:00

And so the benefit there is obviously if someone starts shows, shows up on your website, doesn’t convert, you know, they’re interested and you can hit them quickly. Retarget them. And a lot of retargeting happens in the digital space with emails and digital ads and things like that. But should companies be considering using direct mail for retargeting their visitors? And you’re able to get to their postal address as well as their digital information?

Bruce Meberg: 30:31

Yeah, the basic record is a person’s name, their home address, and their personal email. So that’s what you start with. We have some additional B2B data. If you’re a B2B company, we can get you their work email, their LinkedIn ID, D, the company they work for, their job title, all that, but the basic record is it’s really direct mail, email centric. But as I mentioned before, that data can now be onboarded to any platform.

So you can onboard that to basically means uploading the names, addresses, emails, whatever information you have onto Meta or Facebook. And then that becomes now a custom audience within your accounts on that platform and you can target those specific people. So that gives you the advantage of knowing who’s in your campaign. And also getting I mean, multi-channel to me is the best. You know, no single media is going to be optimized.

If you have direct mail, send out a drip email campaign and do a digital or YouTube video campaign alongside that, that’s going to give you your best overall ROI mix because the direct mail is physical, tactile lasts longer, sits on your, you know, your living room table longer. So that’s great. But people need to be reminded with email touches or digital impressions as well.

Rick Rappe: 31:42

Yeah, the digital touches can go out extremely quickly, which is an advantage and that in retargeting. But we’re also doing a lot of trigger direct mail programs, which can go out extremely quickly as well. And they really complement the digital approach, I think totally, because not everybody is paying that much attention to the ads they’re seeing on Facebook or, and, and it’s a different it’s a different experience, right? Getting a mail piece is very, like you said, tactile. It has staying power.

So it and it’s more trusted these days, I think than a lot of digital channels.

Bruce Meberg: 32:19

Yeah. I mean, I think, you know, digital can be frustrating. You really have to go with the large numbers because people are just simply not registered. It’s not registering in your consciousness, right? Email’s is tougher and tougher these days as well.

So if you really need to get the message through, direct mail is definitely one really good way to do it. And you have to be very selective because you know, you’re spending a decent amount of money for each each piece, but that’s why you need tools that give you the right level of engagement, the right pages overriding if it’s not. Also, I didn’t mention you can actually target just your geographic area. So if you’re in a 3 or 4 county area, that might be 500 or 800 zip codes, you would only get leads of people who live in those ZIP codes. And if you have an opt out list for privacy, or if you have an existing customer list that can be used to exclude the lead so that if we see someone that’s on that list, we simply throw it away and don’t charge you or give it to you.

So, you know, you’re really getting net new high engaged prospects. And that’s the key.

Rick Rappe: 33:19

That’s amazing. What an amazing tool.

Bruce Meberg: 33:23

Yeah. It’s, it’s, it’s a, it’s getting very, very popular. It’s, it’s worth any, any company really should consider testing it. And again, you can do it with direct mail. You have the email to.

So you can do that and you can onboard it for digital. So it’s not even like you have to change your current media strategy. You’re just finding new leads or really the most engaged you can possibly hope for better than any, you know, cold prospecting list you could, you could use.

Rick Rappe: 33:50

Right, right. How is AI changing what you’re doing these days? And how do you see AI shaping the future of targeting and direct marketing. Yeah.

Bruce Meberg: 34:05

Well, I, I love, I, I mean, you know, between grok and ChatGPT and me were a great team, you know, but it’s like having some very, very smart assistants, although they can’t, they can be kind of dumb sometimes too. And I don’t usually just rely on one’s word or the others. I always make them like compete against each other. But yeah, I, I don’t think it’s gonna ever really replace the core of what we’ve been talking about. I think some people might have that idea, but I don’t believe it will because everything we’ve talked about has extremely proprietary databases and first party knowledge that the AI simply doesn’t have access to.

I mean, I suppose you know, somebody, maybe one of these big database compilers will start licensing their data to the AI companies at some point. But, you know, you know, the specific knowledge and the proprietary knowledge that goes into the modeling that goes into the intent leads. And obviously the first party data. Nobody knows that except you. So AI will not know that, and they won’t be able to identify who’s on your website.

So that’s beyond the realm of what they can do. It’s not like they know everything about the world. They’re not, you know, deities or their, their, their programs that get fed data and spit out answers. They don’t know everything. They know what you can, are able to teach it and tell it.

But some things they just won’t be able to get access to, especially, like I said, you know, website visitor data, that’s something that only you as a company have that and we help you enrich or identify and, you know, figure out who that person is.

Rick Rappe: 35:35

Yeah, I’ve always found it extremely helpful to find and have experts like you that I can talk to about data sources and data targeting ideas, because there’s so much in the list world that sounds good in the description, but it’s all kind of a black box. And, you know, really we don’t, we don’t really know where that data came from. How it was compiled. You know, how much of it is inferred data versus actually collected data or recorded data. Right.

And so the advantage of working with an expert on targeting is often that, you know, they’ve tried things and they know what works and what’s worked and what hasn’t worked for other clients. And so they obviously lead, lead you in the right direction and help to make the choices. But there’s a lot of things that sound good on paper that may not perform as well as you think they would. So it’s very interesting.

Bruce Meberg: 36:38

And something we’re looking into now. It’s been kind of around for a few years, but it’s not it’s it’s to the point where now it really makes sense is to basically take all our knowledge, all our product knowledge and, you know, experience and get it into the AI. So train it on our knowledge, on our knowledge base and our data and everything we’ve done, so that you can use that so that you don’t have to ask me, you can ask my AI surrogate and get a pretty good answer, and probably a lot faster than I can type it out or write it down or spit it out in my words, you know, it’s like you type it and it starts. You get the answer almost immediately sometimes. So so that’s something we’re working on now.

I’ve seen that before, but not to the extent that I have this vision of it that I want to get out there in the next couple of months. So keep an eye out for that tool on our website.

Rick Rappe: 37:32

Will do on predictive modeling. I want to go back to predictive modeling for a second because, because our conversation just made me think of something else that I want to ask you, which is. The database that you’re using for predictive modeling. I mean, obviously Inbound Insight has access to and brings together multiple different data sources, and you have lots of options as far as what kind of data you can use it. Is it important for marketers to think about testing different predictive models against each other?

Or I mean, when I think of predictive model, I think like, let’s create a predictive model. But are all predictive models created equal?

Bruce Meberg: 38:17

To a degree they are, but there’s different ways you can approach them. So for example, if you have a national client, like we have a lawn care client with, with different brands and different parts of the country. So you might have, you know, something in New England might be different from something in Texas. So you might want to have two separate models created simply because there’s a lot of different factors that you’re not sure you can’t predict if you kind of blended the data together. Instead of having two really good models with their own peaks, you’d have kind of one blended one with a lower peak in a sense, if that makes sense.

So it’s so there is a concept of creating more than one model for a client for various reasons like that, or even a product based model. So like if you’re a bank or a credit union, you might have, you know, you know, different product sets and you might want to model each product set to create its own model for that. So, there’s different ways you can do it, but, but the underlying technology of the models and many of the underlying databases are not that different these days. I mean, you get to like a point of diminishing returns. Like remember when we used to buy computers based on how fast the chip was?

And every, every, every six months, oh, it’s like 30% faster. We got to do something. It’s not like that anymore. And this is actually it’s pretty mature technology. I mean, it’s regression.

And regression has been around for 5100 years. What’s changed is the databases. But even those after a certain point, with all the data sharing that goes on, are starting to reach a point of everyone’s like 95 to 98%, you know? So like, when’s it 80%? When’s it 99%?

So. So it’s really how you implement it almost more than which specific model you would draw from. Yeah.

Rick Rappe: 40:08

That makes sense. We’ve seen some big differences in terms of direct mail performance, like for acquisition, for prospecting campaigns where, you know, it depends on a lot of it depends on the data that you choose to put into creating your model, the customer group that you select to, to, to model after. You can certainly, you can grab just a random selection of your customer base. You can grab a very targeted group of your customer base that responded through direct mail. That’s great.

If you’re targeting people with direct mail, you want people that are direct mail responsive. So, right, picking out the people that came in through Yes, Facebook or email or whatever helps to make the direct mail performance better. And of course, you can also look at your customer base and say, what are the high value customers in the mix, like you said, and go after those people that are not only responsive, but buying the right type of products and right contributing the right lifetime value.

Bruce Meberg: 41:13

Right? Because there’s two components here. One is conversion and one is maximizing lifetime value of the converters, right? They’re slightly different. You can have a lot of people who buy a small amount, or you can have a slightly smaller number that are buying a bigger amount, right?

Rick Rappe: 41:26

Yeah. For one of our clients, we even created a predictive model that we use within their former customer base. So we’re doing a former customer campaign reacquiring lost customers, but we’re able to, you know, be very selective and rank order that universe and be very targeted and selective in how we you can even.

Bruce Meberg: 41:48

You can even predict which ones are going to lapse. So even the stage before you lost them, you can say, oh, this one’s going to lapse. I better optimize or offer to give that person to keep them as a customer.

Rick Rappe: 41:59

Head off the head off the problem. It becomes a problem.

Bruce Meberg: 42:04

Exactly.

Rick Rappe: 42:05

It’s a great idea. Wow. That’s great. Well. One last question.

How do you see the economy influencing consumer behavior right now? And how do you think that should also influence prospect targeting?

Bruce Meberg: 42:23

Well, I mean, I’m actually kind of bullish on the economy. I mean, I don’t have a great answer for you specifically related to, you know, what we do. But I do see, you know, as gas prices are starting to come down a little bit and I think interest rates will drop next year, I think I think we’ll have a very strong growth in the US GDP next year, which is going to lift all boats. So maybe that’ll open up some marketing budgets allow people to try some new things. And, you know, create more optimization plans and testing strategies and test cells.

And, and really, you know, I think it’s a good, it would be a good time to propose some of those things so that when they are, you know, in that growth mode, they kind of have a things in order and are on the right path and they can just really ramp it up at that point.

Rick Rappe: 43:11

That makes sense. Well, thank you, Bruce. This has been a very informative interview, and I want to make sure that people know how to get a hold of you if they’re interested in talking with you, more about list targeting or what you do at Inbound Insight, how should people reach out to you?

Bruce Meberg: 43:28

Probably the best way is to send me an email. It’s simple. It’s bruce@inboundinsight.com. So look forward to hearing from you.

Rick Rappe: 43:37

Okay, we’ll put that in the show notes and people can find it in the transcript and, and get Ahold of you. And I know Bruce, you’re also on LinkedIn. I’m sure that people could also reach out to you on LinkedIn.

Bruce Meberg: 43:49

Absolutely.

Rick Rappe: 43:50

Heck, you can probably go to Bruce’s website and it’ll probably capture you as a lead and reach out to you. You never know.

Bruce Meberg: 43:56

May just do that, may just do that. Yeah, obviously that would be inboundinsight.com. But there’s a lot of good content on there, industry specific. So I do encourage you to poke around and see if there’s something there you like.

Rick Rappe: 44:07

Yeah. Well, thank you for being a resource for RPM, Bruce. We really appreciate that. And thanks for your time today on the podcast.

Bruce Meberg: 44:15

All right. Thank you so much, Rick. Really appreciate it. Thanks, everyone.

Outro: 44:18

That’s a wrap for this episode of Response Drivers. Thanks for tuning in. If you found today’s insights valuable, make sure to subscribe so you never miss an episode. And if you’re enjoying the show, we’d love it if you left a review. Got a question or a topic you’d like us to cover?

Just drop us a message at responsedrivers@rpm.com. Until next time, keep driving response and making your marketing work smarter.