This is the Customer Equity Accelerator, a weekly show from marketing executives who need to accelerate customer-centric thinking and digital maturity. I'm your host Allison Hartsoe of Ambition Data. This show features innovative guests who share quick wins on how to improve your bottom line while creating happier more valuable customers. Ready to accelerate? Let's go!
Allison Hartsoe - Welcome everyone. Today's show is about the first steps or foundation that must be in place to simply get in the game of customer equity acceleration. These first steps map to the same foundations as digital maturity, as we've discussed in the early maturity curve show and to help me discuss what this is and why you should care. I'm here with Loren Hadley the V.P. of Customer Journey and Optimization at Ambition Data. Welcome, Loren. Tell us a little bit about your work at Ambition Data.
Loren Hadley - Good morning, Allison. Yeah at Ambition Data I have the opportunity to work with executives and practitioners to connect the dots and try to really develop their capabilities in leveraging data. To me, it is gratifying to see them make strides understanding their customers and serving them better while moving their businesses forward. And I personally enjoy being able to explore a variety of business models, industries, tech stacks, data cultures. It really keeps things interesting and makes me continually think about where we're headed and how we're going to get there.
Allison Hartsoe - Yeah, I like that about consulting too. It's definitely an industry for the curious, no doubt. Thank you. You know when I think about the foundations of customer equity I think about it like a kitchen to create value or even see customer value in the first place. You need the right equipment. I sometimes think about how my husband when we were dating he made macaroni and cheese in a hotel coffee pot. Not exactly the ideal tool. So that's not the most efficient way to cook for sure. And if I were cooking up data I'd hate to be doing it with a coffee pot. Let's say that I'm cooking on a camp stove. Loren, why should I care if I'm cooking on a camp stove?
Loren Hadley - All right. First off, I don't want to disparage camps stoves because I've made some great meals while we are camping but it's not something that I would want to do on a large scale. If I'm cooking at home, I've got counter space. I've got my full range of knives and utensils. I've got running water. But when I'm cooking on a camp stove I'm generally cut off from all of that preparatory stuff, so I have to do a lot of the prep work ahead of time. The other factor is scale. We've got two basic burners there with kind of limited capacity at home I've got four burners and an oven to work with. If you are in a commercial kitchen you'd have the full range of burners, fires, grills. Yes. So, if I think about trying to serve a restaurant crowd on a camp stove I think we're all going to be disappointed. Here's an example of moving from camp stoves to commercial kitchens in the actual digital world. For a number of years, I had the opportunity to work with a Fortune 100 client who was moving through a digital transformation. We started out helping their direct consumer team to utilize data. Now we built a series of reports that were customized for each of those groups curated data that answered the type of business questions that they were really looking at. Then over time, we enhanced each of those as each of the groups got more adept at using the data that was really our camping cooking on a camp stove phase. We were using Excel and data connectors to pull in a process and display the data. It was effective it was appropriate for the time, but it really was sort of our food cart stage. Then eventually as we built these out in a number of areas we found that the number and complexity of the reports were becoming challenging to manage. Each time that there was an update a change in the system it required massive amounts of labor to go back through and find all the places where that was touched and fix that in the Excel systems.
Allison Hartsoe - I'm really glad you mention that because I don't think people realize how much time goes into things like connecting data behind a report. That's a really big hurdle for most people.
Loren Hadley - It is and the fact that you think of data as being fairly static but in my experience, you find that it really shifts a lot that it doesn't take much for somebody to add new variables and change variables out shift name a little bit. And at that point you know it becomes a real bear to try to go back through it and make sure that everything is still running smoothly.
Allison Hartsoe - So you were at the camp stove stage and then you started to move up into more advanced or better equipment. What happened next.
Loren Hadley - So the client decided to fund a major initiative shifting to a single system that would combine all of these reports into one suite of tableau-based report that would serve the whole company. So, you know essentially we were at the food court stage then we were going to move into a full restaurant mode. We were adding state of the art data prep tools and a whole array of enterprise-grade data handling systems. So, we froze work on the previous reports and all of our efforts went into building these new reports and the tableau interface. We worked with our corporate champions and with the data consumers and created a report that everybody was really excited about that brought all the pieces together that was a much more efficient way to view the information.
Allison Hartsoe - So when you say efficient? Was it faster. Was it or was it still subject to a lot of those hurdles trying to get everything together?
Loren Hadley - It was much faster. We were able to pull off or at least the plan was that we were going to be pulling off of one major data source that data source was going to be quicker.
When we looked at maintaining the Excel-based reports, you know each time we made a change we had to go through check all the pieces the formulas that were in there do a manual refresh and then share that out. In this case.
Allison Hartsoe - How many hours was that? When you did the Excel how many hours was that for any change or any like that say they change page names or something?
Loren Hadley - We would often end up spending 10 to 20 hours making changes on reports particularly if it touched on a number of different reports. If you're looking at a single report and a single page name change, you know we might be looking at you know five to 10 hours of labor on somebody's part to figure it out. But even worse than that is oftentimes the changes weren't communicated so we'd end up discovering that it was broken because somebody would call up and say hey my report isn't working. What is this thing? And so, we would end up having to scramble behind the scenes because you know someone in IT changed the way that a piece was being handled. And there was no communication down about that.
Allison Hartsoe - Yep yep. Got it.
Loren Hadley - Yeah, that was that was always a bit of a challenge and of course you're trying to build people’s faith in the data. So, when they see something that comes through broken like that it raises questions about the credibility of the whole process. So, you know we were always quick to make sure that we knew what was happening and then we got that updated.
Allison Hartsoe - But that's an excellent point.
Loren Hadley - It's a major challenge yeah.
Allison Hartsoe - Having faith in the data is an excellent point. People will not take action if they don't believe the numbers and even when the numbers are right it can be hard I think to get people to take action.
Loren Hadley - Right if it doesn't say what I want it to say then I'm not sure that that's quite exactly the right data. You know, if it’s not backing up the preconceived notions or the theories about things.
Allison Hartsoe - OK, so you moved into tableau then what happened.
Loren Hadley - Yeah, we built this front-end and everybody was excited to go and then we discovered that the back-end was running behind schedule that they hadn't gotten all the tables built they were still working on ingesting and processing the data into a Hadoop system that we were going to be hitting to actually populate the front end. So, we were kind of at an impasse. We'd stopped working on the old systems the new systems were ready to go except we didn't have any data to feed into it. So, after discussing it with our stakeholders and champion. It felt like you know essentially, we had a group of people that were lined up outside of our new restaurant hungry for data and we needed to open the doors.
And so, despite the fact that we didn't have our main dataset what we did is scramble to build and new kind of comprehensive datasets that encompassed about nine different Excel spreadsheets. A bunch of them pulling in data from different data connector sources and then couple that with process and format those into tables that we can use to populate the front end.
It was a Herculean effort to get that done, but it worked.
It was also very, very fragile, so and it was very labor intensive. Each of those seven data polls had to be done in consecutive order. Then we had to go through and do the updates on the main excel sheet and there was a manual component to each of those steps. And if any of those steps failed or got out of order then you essentially had to start back over again and try to rebuild it and to make sure that everything was right once you did pull it in. So obviously not an ideal situation but it allowed us to start cooking. Essentially, I think we were working with a whole stack of omelet stations trying to serve that restaurant while we waited for them to get the kitchen spun up.
Allison Hartsoe - Yeah, like a bunch of verbatim camp stoves.
Loren Hadley - Exactly. Exactly.
Allison Hartsoe - Yeah, that's the that's and then so when you previously had maybe 20 hours per Excel report. Now, what were you operating on in terms of time to get all that data together?
Loren Hadley - Our weekly time on updating that because at that point we were updating them on a weekly basis was probably 30 to 40 hours a week. It was spread out over a variety of people in different teams that were responsible for pulling and Q/A'ing the data and getting it linked into the system. But it was it was not a minor effort and you know we were all walking on eggshells because it was very fragile, and we were just waiting for something to snap. And of course, it's getting a lot of scrutiny because this is just launched its you know we've got our people that are excited about the new data. They're trying to really dig into it. And the last thing in the world we wanted was to have it you break down not be available. You know we wanted them to really embrace this and it was going to be critical for the long-term adoption of the system.
Allison Hartsoe - But again this was a band-aid. You know it was the brigade of camp stoves. I love how you think about the camp stove analogy in order to get to what would be like a gourmet cooking station. If he'd had the right tool in place it wouldn't have been so fragile. Or as labor intensive I imagine.
Loren Hadley - Exactly exactly. So, eventually, they did get everything hooked up, connected and working. So, we had this great dataset. They opened the kitchen for us and we were able to dive in there get that connected up and then once that happened essentially the system ran itself. We did some Q/A work to make sure everything was flowing smoothly but we had a regular ingestion of digital data that was being brought in, processed and dropped into tables on a daily basis. We had information that was being pulled from transactional systems. So, we ended up with one great data source and one place that we could pull from to do all this and it was automated. We were basically leveraging all the powers of Hadoop at the time to pull that stuff in for us and take that labor off. So, once it was in place you know we didn't really have a lot to do aside from periodic Q/A and maintenance and upgrades to it as some questions and things came in.
Allison Hartsoe - Do you think the band-aid stage was just necessary to proving to people or showing people what could come from the data?
Loren Hadley - I think the band-aid stage was really, so the camp stove stage building up to it really was if we had jumped in and tried to build this giant system from the get-go people wouldn't have been ready for it. It would have been a huge outlay for a very unknown response. People didn't need the level of data that we were pulling initially. They had to get comfortable with that they had to learn how to use the data, how to start asking questions, and how do we leverage that.
Allison Hartsoe - How to have an eye for the data.
Loren Hadley - Exactly. So, starting out at that camp stove stage made a lot of sense as we were bringing people up and educating them about that. But then you know as they've got more advanced in their use of data it really became necessary to move to something that was more scalable.
Allison Hartsoe - Yeah, that makes sense. I think we often ignore the cultural impact of data. You know you can put great tools down but there's still a cultural response where people have to develop an understanding and know what they're looking at start taking action and then relying on the data you know from a technical perspective it's easier to think about pinning all this stuff together and then you know wow you have everything you need.
But yet if the people who are using the data don't know how to cook in the first place you end up with a lot of kids in the kitchen. Don't touch the stove.
Loren Hadley - For sure, and we find that in a lot of cases you know as people aren't ready for it it's easy to overwhelm them and then they'll either become defensive because you know it's like I don't know what this stuff is and it kind of calling my professional credibility into question or they just kind of throw up their hands and say I'm going to keep doing it the old way it was easier. So, we had to sort of step people through that process of getting comfortable with the data. Give them things that are going to answer specific questions that they really have at the moment and if they get comfortable with that you just keep feeding more and more into that. I think we went through about five iterations on a lot of these reports before we got out of the camp stove phase.
Allison Hartsoe - Wow!
Loren Hadley - And I think it was really necessary to go through that process to move people along because they really you know if the initial stages were not used to data at all they were used to making gut instinct decisions. And what I think is best and I really like that, so let's go. To the point where we were able to really look at it and say if I pull this lever this happens and start to really drive by the numbers.
Allison Hartsoe - Yep yep. Makes perfect sense. Well, this is a lot of effort to put in. Was there any specific ROI impact on this project?
Loren Hadley - So, I'd say direct ROI initially was really minimal. It's more of a long-term scale it was building the capabilities to continue to lead the industry into the future. And so, we spent a lot upfront trying to get this into place and I guess one of the pieces that's challenging about that is you know where do you credit that. Is it the I.T. infrastructure the fact that they added Hadoop systems that pulled all their data together for everybody that was using data. Was it because people were asking better questions and starting to use the data. Was it just that the executives were really focusing in this area. And so, people were stepping up to the plate more because of all the scrutiny that was going on. You know it's hard to know which of those things was the driver and probably not any given one of them was you have a bunch of different confounding pieces that are either improving or holding back the process. And it's really hard to separate out what those are. So attributing direct ROI to a specific initiative like this gets to be pretty challenging at least in the short term.
Allison Hartsoe - I can see how that would be a challenge and I think in most cases the whole our ROI angle is incredibly difficult when you're at the very beginning of the curve. You know you want to throw in a piece of technology and immediately see the ROI but your point about it being short-term versus a long-term view is really it's really well taken in the same way that you know cook doesn't become a gourmet overnight. If you're just learning to use the microwave your microwave dinner might not have the same ROI as your professional cooking school output.
Loren Hadley - Yeah, I mean this was a five year plus initiative to get where we were going with that. Imagine if you were in college and you were a junior and you decided to, you know what's the ROI of staying in school and you started looking at that from like the immediacy of it how much you might bring in now because I'm in school then you know that's really kind of a fallacy because you know that by completing your degree you're going to get long-term benefits. Yet a lot of companies approach it from that perspective. Well we're two years into a five-year process where does that return on investment that we wanted and so seeing it at that point there is a tendency to pull back and that's why I think a lot of companies get trapped in inertia. They don't make it forwards because they're looking at the short-term ROI. They're looking for those immediate gains when really what they're doing is building capabilities that are going to give them those long-term gains and competitive advantages down the road.
Allison Hartsoe - So when we say short term long term what are we talking about in terms of time.
Loren Hadley - Well let's see, this was typically a five year process formally that we were given a large Fortune 500 Company and it wasn't a you know that was essentially to shift them over to a culture of data to get them starting to think about data and instead of you know shooting from the hip which had got them a long ways, you know, what if we really know what our customers are doing, where they want to go, what they want to buy, how they want to buy it, what frustrates them; if we know all those things we could focus our energy in the right places and really make a better experience and improve the perception of the brand, pursuit improves the purchasing of products and basically make everything flow more smoothly for the customer and profitably for us. So the five-year mark was sort of a. And now we're at the starting line. That was the point where we were you know getting people moved from my data is I know how much I'm supposed to sell and I know how much was sold to actually having you know.
Balance kind and it tells me just like if I'm pulling these levers what's actually happening with that. If I run this sale how are people responding is it. Is it working the way I think it's going to work or are people ignoring it or is it having some sort of a negative impact and being able to do those things and react quickly.
They made a huge difference to the company, but you know the first portion of it was really getting their degree. It was building those capabilities that then allowed them to move forward in a more competitive way.
Allison Hartsoe - Yeah. Got it. Got it. So, let's say that I've decided that I'm going to build out the features of my customer data collection kitchen. What should I do first? How should I how should I move forward?
Loren Hadley - I think there's really three critical steps and a lot of times these get overlooked. The first is plan your menu. Know what it is that you're going to do with the data down the road. I think in a lot of cases companies go through and it may be driven by IT, somebody in the executive side says we need to have a data lake. We need to be able to get all of our data at once, so somebody in IT sits down and they make a plan and they build out this large data pool. But nobody sat down and said "OK" what are we ultimately going to try to do with this? Are you going to do personalization? Are we going to try to sharpen our marketing skills? You know there's a whole range of things that we can do once we understand our customer data. But if we don't plan what we're going to be delivering then we aren't going to be able to design a kitchen that's going to allow us to do that effectively.
Allison Hartsoe - Got it. Got it.
Loren Hadley - So you really have to sit down and just make that plan and move on.
Allison Hartsoe - Got it. So, step one is plan your menu. What's step 2?
Loren Hadley - Step 2 is design your kitchen. Now that you know what you're going to cook, do you need a wok or a deep fat fryer or a grill? You can look at your data based on what it is that you're going to be producing so that you can make sure that when we say we want to do personalization you don't get the answer, huh? really. OK well, our data doesn't do that, but we could start a new initiative that would move us there within a couple of years. Instead, you know that that was where you were moving. That was on the roadmap and you can start to build out your kitchen design for that. How many people are going to be in the kitchen? Are you a modest size company with a team of IT data scientists that are going to be working on this? Are you planning new democratize data or tie it out to a variety of automated systems? Once you kind of know where you're going with that, you can you know who's going to be working with the data? What it is that you're going to be producing? Then you can design your kitchen. And if you look at cloud-based computing that makes it a lot easier these days than it used to be, but it still requires some planning to get there.
Allison Hartsoe - Yeah it does. And you know what's interesting about the whole design your kitchen is I watch the movie The Founder a little while ago and they talked about the McDonald's process and how long it used to be when you would go into a drive through and you'd wait wait wait wait. Finally get your order and maybe it was right maybe it wasn't and a lot of that had to do with the design and layout of the kitchen. And they showed this scene of the founders of McDonald's actually on a basketball court with a bunch of chalk and having their employees walk through making different things and they did a race a line they'd put in something else. I mean they didn't get it right the first time they had to figure out the right flow for what they were making and then they knew when the order came in for the hamburger that they were ready, and they needed certain ad-hoc processes like calling out that they were moving from place to place so that they didn't run into each other. It's a fascinating look at the architecture that goes behind any smooth process whether it's cooking a hamburger or whether it's cooking up data insights.
Loren Hadley - Yeah, I mean if you think about it you've got a lot of different systems out there. You've got perhaps a transactional system that is capturing information coming in from your e-commerce system or your brick and mortar stores. It's feeding information to finance, it's feeding information into supply chain perhaps. Now you want to do something else entirely you want to be able to feedback through so that you can understand what a customer bought previously but do it in real time. So, you show up and I know you previously bought these products and you returned this product and this is a size that you actually ended up with in the end. If you want to build a system that does that. Now you're suddenly messing around with other people's data. You know finance they want to make sure that everything is still flowing the way they need it. The supply chain needs their piece and now you're adding additional capabilities on there. So, if you don't do this with forethought and consideration you may end up with a lot of political pushback and people who want to protect their data and they don't have the resources or want to take the time it's going to be required to really port this stuff into a more usable situation for somebody else because it’s a change to the way they've been doing business.
Allison Hartsoe - Yeah, got it. OK so step one was plan your menu. Step two is design your kitchen based on your menu and what's step 3 step.
Loren Hadley - Finally, step 3 knows what's in your pantry.
And that is if you're digging into your spice cabinet and nothing has a label on it you might get by by popping up the lid and sniffing things and you know kind of picking and choosing. But it's probably not going to be your best work. So, what if you do that with data. If you don't know where your data is or what your data is you're going to have a lot of problems. That's really sort of the first level of this, is how do you make sure that people know where they need turn to really get the right stuff. And that's really documentation and governance issue. But the second piece applies even more granularity to you know you're working with customers. How do you know who that customer is? A lot of our data is anonymous at least on the digital side. But on it's a transactional side we know who somebody specifically is. You need to pin those two pieces together in some way so that you can suddenly know this is Customer X even if they're not authenticated there's some different ways that you can go about doing that. And I think that's a critical piece the more you can move to knowing and understanding who each customer is along the way, the more you can reduce the anonymous sort of aggregate information that you have the better you're going to do at really leveraging this for useful purposes.
Allison Hartsoe - So, does that mean that I might have six cans of cinnamon in the cupboard with different ways of processing like; I've got cinnamon sugar and I've got regular cinnamon. I've got cinnamon sticks and they're all cinnamon but they're not necessarily the right tool for the job. Is that the kind of approach that we have with data? I've got different flavors of the same thing.
Loren Hadley - Or even, you've got the same you know that that is definitely one piece of that but what if you have six jars of actual cinnamon taking up space in your cupboard and really, they're all the exact same thing. Which do you pull from it? Why do I really need six of those? What you really want to do is go - "Oh, hey all six of these are cinnamon. Let's put them into one bottle and move from there."
And that's you know that's one of the challenges we've got people that are working on mobile devices. They're coming from work coming from home. I'm the same person I'm the same shopper but I'm showing up as three different bottles of unknown information or possibly three different known bottles. But you know I do you consolidate and bring that together. So, we really have a usable data set that you know that I need whether I'm at work, whether I'm at home, whether I'm on my phone on the bus.
Allison Hartsoe - Well that is a huge topic in itself and one we can follow up on. Because the whole crispness of "Who are you and how do we identify you" is just that; that's that's a very challenging and definitely an area of growth online where we're seeing more and more data sets merge together. So maybe we'll consider that for a future show because I know a lot of people would probably like to hear more about that.
So, let's let's move into summarizing and wrap up our show for today. First, we talked about why does the foundation matter, of course, the foundation is where the initial listening signals are captured. So, without good equipment to create that solid data foundation, you might as well be cooking on a camp stove right. You can do it. But the competition to nab the best customers is really becoming fiercer faster. So, get in the game, get the right equipment, get going. Second what kind of impact can you get? We talked about this and I think the impact in the foundational layers is very minimal because as we talked about Loren the time period that we're looking at is oftentimes years and like the foundation of a house or of a kitchen. You've got to have the right pieces in place before you can cook up a great meal. The stronger the foundation the tastier the meal, meaning the more insights you'll get in the long run, right? Takes time. Yep. OK so finally you know we talked about the three steps, but you know let's say that what if I don't trust my data at all. You know if I'm really at the beginning of the foundation then it's high time to get your digital house in order. I always think the website is a very rich analysis place to begin. It’s filled with lots of customer behavior even if there's no checkout or conversion. So, one possibility to get started can be to ask for an audit of your website. There's a tool that I use a lot and I often recommend it's called observe point which has a really nice crawler that can help you do this and they even cover apps you know like we were talking about there's lots of different sources of data and I'll include that on our podcast page. Loren, are there other tools from your reporting example that you would recommend that kind of you know somebody should have in their initial kitchen.
Loren Hadley - You should definitely have somebody that has that ability to see what's actually happening. I've run into a lot of situations where you're trying to instrument a site or an app and you think it's right, but you don't know. And so, the data that's flowing through is either not arriving or is coming in funky so something like a proxy tool; packet sniffer can be a really handy piece to add to your team's repertoire.
Allison Hartsoe - What would be some names of those tools? So, we can add to.
Loren Hadley - I would look at Charles & Fiddler are two of the versions that I use. There's actually some ways that you can do sort of a quick and dirty version of it with chrome just by going into your developer tools and looking at your network traffic that is going on. I would also encourage a manual audit. You run through the observe point. It's a great tool it tells you where everything is happening. But then what you want to know is am I actually getting what I think I'm getting in each of the places and that's where it needs somebody they've got background in both analysis and enough of a technical bent that they can go through and look at and say here's what you intend to do. Here's what's actually happening. These pieces are working fantastic.
Allison Hartsoe - Good point.
Loren Hadley - This piece over here there's some best practices that we should follow that we could make this a lot stronger for us.
Allison Hartsoe - And on the reporting side we talked about Excel as a reporting tool there's a plethora of new reporting tools coming out but are there any tools that you'd recommend you know for knitting data together when you're trying to combine multiple sources.
Loren Hadley - Yes, for you know if you are at sort of the camp stove home cooking stage possibly even a larger stage you can leverage Google's data studio. It's a fairly new tool that's out there but it's got the ability to connect to and bring in a lot of data from different sources. I'm a big fan of Alteryx. Alteryx allows you to connect up and easily manipulate data and build data tables, export them out to a variety of different formats and I think that's a really fantastic way to pull in you know first off it allows you to do a lot of data cleaning and then formatting and then your manipulation and output and so it's a really useful tool we don't have to be a Python programmer to run.
Allison Hartsoe - And you can push it right into Tableau too, right?
Loren Hadley - You can push it directly into Tableau. You can push it into other data sources and you're entire Tableau piece into that and you could push it together into a table and throw something like Looker on top of it. So, it's just a powerful tool for knitting those pieces together and it's probably the commercial kitchen version of the data management if you would say data studio might be your kind of your home kitchen approach.
Allison Hartsoe - Perfect perfect. So, we'll link to all of those tools on the podcast page which as always is on www.ambitiondata.com/podcast. Loren, I want to thank you again for being our guest today. I really appreciate you comin' in. It's really great to have you and hear all your stories.
Loren Hadley - Thanks, Allison.
Allison Hartsoe - Yeah. Bottom line when you use your data effectively you can build long-term customer equity. This is not magic. It's a very specific journey that you can follow to get results. Thanks, everyone.
Thank you for joining today's show. This is Allison. Just a few things before you head out. Every Friday I put together a short-bulleted list of three to five things I've seen that represent customer equity signal not noise. And believe me there's a lot of noise out there. I actually call this e-mail The Signal. Things I include could be smart tools I've run across, articles I've shared, cool statistics or people and companies I think are doing amazing work building customer equity. If you'd like to receive this nugget of goodness each week you can sign up at ambitiondata.com and you'll get the very next one. I hope you enjoy the signal. See you next week on the Customer Equity Accelerator.
Who Should Listen: Marketing, Analytics, Digital Transformation, Customer Transformation, Data Insights, Customer Experience, eCommerce, Digital Marketing, and Customer Satisfaction professionals.
Key Concepts: VOC data, CRM data, customer valuation, data organization, customer equity, customer lifetime value, CLV / CLTV, customer centricity, customer equity valuation, digital maturity calculation, customer digital data, customer equity metrics