Webinar

Webinar: Callahan’s Intelligence Platform

See a demo of the Callahan Intelligence Platform and how we use it to store, organize and visualize data in order to uncover aha insights for smarter marketing strategies. 

Learn how our Intelligence Platform makes marketing planning smarter, and ensures that your strategies and tactics are focused on business outcomes (and not just typical marketing KPIs). The webinar will cover:

  • Why starting with front-end data analysis is better than using only post-campaign analytics
  • The components of the Callahan Intelligence Platform tech stack and each of the benefits
  • A live demo of the Platform, showing several use cases for various business situations
  • How the Platform allows us to plan, test, execute and measure more effectively
  • The always-on dashboards that empower clients to explore and ask better questions, faster

Webinar Transcript: 

Jan-Eric A.:
Thanks for joining us on the webinar today. We are here to talk about the intelligence platform at Callahan. By way of introduction, I’ll just introduce myself. My name is Jan-Eric Anderson. I am Chief Strategy Officer at Callahan. Sitting next to me for the webinar is Zach pike who is VP of Data Strategy and Marketing Analytics. So thanks for joining us. What we have planned to walk you through should last an hour or maybe a little bit beyond. So that’s what we have planned.

Jan-Eric A.:
Just a little bit of housekeeping. If technology works, you should be able to submit questions as we’re going through this and so we encourage anybody to submit questions as we go through this. We will answer as time allows at the end of the webinar. And we’re also going to record this webinar and make it available at the conclusion if there’s anything that you would like to revisit.

Jan-Eric A.:
So, just a little bit of context and background on the intelligence platform, then I’ll get out of the way as Zach, who will walk through this. We’ve established this platform because we really felt like at the agency, it filled a void that we had with clients about understanding how we as an advertising agency in the traditional sense, how we can help our clients achieve business goals. And looking at what this intelligence platform is designed to do is really prioritize impact on business and understanding what drives business for our clients rather than jumping to the conclusion that advertising or the things that we’ve done for other clients is always going to work.

Jan-Eric A.:
So what this is really about is, is trying to understand what drives business and how that can then inform what we do as a marketing agency. Our intention today is rather than to sell anyone on the intelligence platform, it’s really think of this more as, our intention at least is to have this more of a show and tell of what this is about. Just to give a little bit of insight and really it’s in the interest of transparency and sharing and welcome any feedback as we go along.

Jan-Eric A.:
So, part of what makes it unique as you think about, if you’ve dealt with an agency, and we talked about analytics, typically analytics are used in the form of post campaign reporting and understanding how did marketing activity perform. How did it work? And so, some of the questions that are typical that may come up in those conversations, did my marketing work, did it drive foot traffic to retail partners? How was the click through rate it? Did we drive web traffic? Which digital or traditional channels performed the best however best is defined? Is social media driving business? And so many of these questions are really coming mid or after the campaign and wondering, looking backwards, how did it do. The intention of the intelligence platform is to not just look backwards at campaigns, but to get ahead of any campaign by understanding business and understand what we can do.

Jan-Eric A.:
In this world where analytics are used for post campaign reporting, those questions are important but they’re really lagging indicators. And so, we get ready to go do something, a social media campaign. We fire in the sense of being ready to do it, we fire, we execute it and then the aiming piece, we kind of figure out at the back end, where we targeted the right people, did it go to the right places, was it the right thing to do. So it’s kind of a ready fire aim. The intention with front end analytics is to get analysis done on the front end to understand what will drive the business. And in that sense we’re kind of delivering more on understanding what we can do, aiming it to the right places before we get into execution. It’s really about informing what we’re trying to do.

Jan-Eric A.:
So that’s a little bit of context and setup for what the intelligence platform is. As I mentioned, we really want this to be a show and tell. Zach is the most qualified person to walk through that. So what he’s going to start with is really just kind of giving a peek under the hood of what technology makes up the intelligence platform and what are the ingredients that kind of go into it that make it work.

Zach Pike:
Yeah, thanks Jan-Eric. Hey everyone. So before I talk about the parts and pieces of this platform, I’m actually going to show you the technologies that we purchased and stitched together. And the reason we’re doing that is that there’s really no black box to any of this stuff. It’s all really straightforward stuff, but what we’ve done is purchased the best technologies you can find for the types of analytics and the types of data that we’re using. So the reason that you even need this in the first place, and if you are a marketer and you’ve been in this industry for a while, you’re already well aware of this. But we have a serious problem from a data perspective. And that problem is just a result of all of the different ways we can actually execute marketing.

Zach Pike:
So if you think about, I don’t know, just email, there are thousands of different systems you can purchase and execute email through, and each of those produces their own little data stream. And unfortunately for us in the analytic space and us in the marketing execution space, none of that data works together, right? Nothing is organized the same way, connecting this stuff is really difficult, and oftentimes, it’s extremely time consuming to pull data out of those systems. And so, that’s really the reason for these different pieces of technology. It’s all for efficiency. Let’s let computers do what computers are really good at and that’s repetitive ongoing tasks that need to happen every day, every week, every month, whatever. Let’s let our human capital do the thinking around this data and hypothesizing and things like that.

Zach Pike:
And so, we start at data. Since that’s the issue we’re really trying to solve to begin with, we start there. It starts with identifying all of the sources that are going to be important to whatever we’re trying to accomplish from the marketing side of things. But what you’ll see in this kind of little show and tell case study is the business side of things. Oftentimes, agencies like us and marketers, the business aspect of results is almost secondary. We tend to focus more on the marketing side and the execution metrics. That’s very different from the way that we do it here which you’ll see.

Zach Pike:
So once we’ve identified all that, you know, all of those sources, the dimensions and metrics that we need to accomplish our goal, we need a place to put it, and we need a way to get it there. And for us, we use a piece of technology called ChannelMix. I do not have enough time on this webinar to give the people at ChannelMix justice on just how much efficiency they can insert into a setup like this. But ChannelMix is our core piece of technology that gets data from the source system into a location that we can use it. Getting it analysis ready so we can ask a lot of questions of it very quickly and it becomes dynamic and usable long term.

Zach Pike:
So, we’re relying on ChannelMix to ETL all of that data. And we use two different database technologies. We use them for different reasons. The first one is Amazon Redshift, which is our core platform, that’s where most of our data sits. But we also use Google BigQuery. And if you’re familiar with this space, you already know these two names. They’re two of the larger names for this analytics, machine learning type world. BigQuery is one that we use for the more advanced stuff, one off type projects and things like that.

Zach Pike:
So it’s good to have data flowing, that makes us efficient, that makes our analysts to where they don’t have to be managing through data and collecting all the time. They can think about that data. But where it gets really interesting and you’re going to see this live is when we start appending data that’s from third party sources. So having all of your own kind of in-company data is really important, a lot of insight in there. But when we can start layering in things like Census Bureau data, Bureau of Labor Statistics, weather data, search activity, we expand our understanding of what’s actually happening across your ecosystem of stores or products or just sales performance in general and you’re going to see that.

Zach Pike:
So in our platform, we subscribe to several services. You’re seeing a few of them here that we’ve curated but it expands much further. There’s been times where we’ve pulled the population of registered plumbers in different areas of the country and how many building permits were actually being applied for in different areas of the country to try and answer questions and draw correlations.

Zach Pike:
So, having data, having data appended to it, we’ve got to be able to visual it. You know, when we’re talking to marketing people, we need to make the data easy to understand so we can make good decisions off of it. The tool that we use for that primarily is Tableau Software. We actually implement, we employ several other technologies but Tableau is our key platform for visualization. We do quite a bit of analysis inside Tableau as well. But it’s the people that make the analysis side really exciting. When you have your analysts thinking about the data, rather than working through the data, exciting things happen very quickly.

Zach Pike:
So, for the next few minutes here, I’m actually going to show you a little case study that we have, this is actually a real case study of a client that we went through a media planning exercise for. I’ll give you just a little bit of background on this client to start. They are a, I don’t know what’s going on with my screen here, they are a CPG company. So their product is distributed in big retailers that if I said the names you would know exactly who these retailers are. One of them is Walmart. That retailer network, Walmart’s about 4000 stores but there are other two next largest retailers double that number.

Zach Pike:
And so, when they came to us, we knew a couple of things about the product and what we had to work with from a media perspective. The first one being that this product was extremely seasonal. There are certain times of the year when this product is really needed across the country and it’s pretty consistent every year. But the real challenge with this program was you had to support a network of several thousand retail distributors. We never in marketing, sure none of you deal with this, but in marketing, we never have enough money to go around, right? We never get to spend what we would really want to spend to fully support a product so that we are beating all of our competition every day. And because of that, we had to be really strategic around how these dollars were spent. We had a certain amount of money, we had a big distribution system to support, we had to make sure our retailers felt good about the support we were giving them. So it just required a very strategic approach.

Zach Pike:
So that was the ask. The way that we will start any type of approach like that, and that gets back to what Jan-Eric was mentioning around front end data analysis is we use data to plan. There’s no reason in collecting all this stuff and having access to everything and being able to use it if we’re not going to use it for strategic planning type exercises like this.

Zach Pike:
So we start with a core fundamental understanding of the business. If you’re a new client, we would have to build a ton of assumptions to build any plan without analyzing the data first. So that includes everything from typical sales patterns that happen from year to year, product share patterns and changes throughout the year. But also you see on a chart like this, even pricing. So understanding the effect of changes in pricing over time and what that’s doing to volume and revenue and customers can greatly impact how we execute from a media perspective. If we have areas of the country where we’re hot on price, we may take a different approach from a marketing perspective in those areas.

Zach Pike:
The other big piece of going down that path is understanding that distribution network itself. So, this product specifically, the majority of the volume comes from these retailers. The online business is a piece of it. But by and large, it’s Walmart and there are other two big retailers that were driving all the volume. So we needed to support that network at a geographic level meaning trade area. So a five mile radius around every store is the area we really wanted to hit from a marketing perspective. When you look at a map like this, this is actually the distribution network for this client, you see big areas of the country where we have no distribution. So there’s a little bit of online volume happening there but it’s probably not an area you really want to support from a marketing perspective when you’re trying to be really, really efficient with your dollars.

Zach Pike:
But as you dig deeper into that data, you find that not every store is created equal. You have very different patterns happening in different areas of the country and sometimes even cities that are right next to each other, those two stores are like night and day difference. There’s lots of reasons for that, right? It could be the placement on the shelf in that location. It could be the store manager and how they operate the store in general. You’ll see stores where their home state are their best performing stores, and you would never know that unless you had a level of understanding that we typically go to.

Zach Pike:
So, when you go down that path, understanding the business, understanding our distribution network at a very detailed level, it’s pretty simple to get to our hotspots of activity. And you’re seeing that here. And if you are familiar with geography at all, looking at this map, these hotspots you’re seeing are really well aligned to population centers. So in this specific example, we sold a lot of product where there were a lot of people. That is not that exciting, it’s not a huge insight that was like a big wow as part of this analysis. But this is where a lot of media planning exercises actually stopped. Because at this point, I know where all my activity is happening. I’m going to go hit all those areas with elevated media support because I know I’m probably going to move the most products there.

Zach Pike:
What this is missing at this point is opportunity. I don’t know if I can sell more in those high population centers. At this point, I also don’t know if there are areas where I should be selling more product but I’m not. That’s where we start trade area analysis. And so, you probably saw one of the logos on that data overlays piece in our platform that said, Esri. Esri is a tool that we use a lot for this trade area analysis stuff. And so, in this particular case, I’m going to save you all the details of all the stuff that we looked at. But we drew a five mile radius around every single location. It was about 8000 stores in total. We pull that data within that trade area. It’s not perfect but it is a really good estimation to understand the market factors that could be driving performance of each individual store.

Zach Pike:
And of course, there was a lot of correlation to high population. So if you had to trade area with a ton of people, yes, you sold a lot of product in that store. What we also found is that in areas where we had high density of our target customer, we were fortunate in this analysis that the data we were able to pull back, we could actually identify how many of our target customer were in that trade area. It was a specific type of person dealing with a specific household type issue that we could help solve with this product.

Zach Pike:
You actually found that we performed much more efficiently from a volume perspective and we had a lot of opportunity to grow in areas where we had high density of that customer base. It was more strongly correlated to density than it was just our totals. So you can understand in a place like downtown New York, you have a ton of people. But your density of customer is much lower in that area. So you’re probably jumping to where we’re going, but anything I’m going to spend from a media perspective is less efficient there, even with good audience targeting, it is just less efficient. And through this analysis, we’re able to define the areas in the country where we had distribution and we had that elevated opportunity level based on target customer density. So density meaning as a percentage of all the households in that area, how many households have this problem? Okay.

Zach Pike:
So on this map, you’re actually seeing their areas of opportunity. And in some cases, they somewhat overlap with those population centers. But you see a lot of space here that was just untapped. It had not really been targeted from a media perspective and it definitely was not a focus in general.

Zach Pike:
So that’s really exciting, right? I have a good map now, a path I can go down to say, okay, maybe I push a big marketing push in these areas or maybe I pick out a few of them and do some testing before I go in full force.

Jan-Eric A.:
There’s another element that comes into kind of evaluating the opportunity here, which is another piece of data that kind of sits outside of the platform, which is understanding variability immediate costs, or how much it costs to pursue business opportunity in each of those. So that starts to really lay good groundwork for projecting return on ad spend. If we understand what the growth potential is in a market and we know how much the relative cost of media is to pursue that growth opportunity, we can set up a model there to see where’s the most efficient pursuit of business growth.

Zach Pike:
Exactly. And that’s actually kind of a preview of what you’re seeing here. It’s one thing to know where my opportunity is and identify that, but it’s a whole other world to segment that opportunity based on whatever factor makes sense. And to Jan-Eric’s point in this case, it was efficiency of the media. You really can understand a payback estimation or at least a hypothesis around that payback once you know your media costs that are associated with going down that path.

Zach Pike:
So opportunity is great. Now I know where I’m going, I know I have a good idea on what I need to be doing from a targeting perspective. But I haven’t yet identified when. That was a big factor for this specific issue. We had a seasonal product. But the understanding was it is a general season across the country. There’s not a ton of variability as you dig deeper. We were able to dig deeper because we had done the legwork to wire up all of this data on the front end. What you’re actually seeing on the screen here is just a quick state level view of that product seasonality.

Zach Pike:
So, this makes sense to anyone who’s really dug into data before and thought about this at a deep level, but on a seasonal product that is you’ll see later, somewhat driven by weather. Texas has a really long drawn out season. It starts early and it really ends late in the year compared to a place like South Carolina, Georgia, Ohio. Tight seasons. They start at different times

Zach Pike:
So looking at this, why would I ever have one flight window across the entire country? I would never do that, right? I would never treat Texas the same way that I treat Ohio because if I run the Texas flight strategy in Ohio, I am wasting a ton of dollars across that type of flighting window.

Zach Pike:
So again, this data is analysis ready, we can ask a lot of questions really quickly. If I want to look at this for every single store across the country, I can do that. So, this is the same view you’re just looking at but you’re actually looking at every store inside Georgia. You can see the areas of high sales activity. And Georgia would be a pretty easy state to build one plan around, right? I’ve got a pretty defined season throughout the year. But you think about a state like California, Southern California and Northern California are two very different climates. They get warm and cold at different times of the year. When you look at all of your stores across California, it’s much more sporadic.

Zach Pike:
So you’re going to see a little bit more of this, but this was really the point where we said okay, traditionally, this product, the marketing plan has been executed across the country on a traditional flight schedule that was really stable everywhere. But we’ve got the data, why don’t we just let math build our marketing plan for us. We can use this data to define the logic around our start and end dates. We can use this data to find the logic around our budget ramp up and ramp down. So let’s let math do what math is really good at and build this out for the 6000 locations that we were targeting here.

Zach Pike:
So that’s exciting, right? I now have a really good understanding and I can be super efficient. I’m not spending money when customers are not thinking about this issue, which is the times when we waste those dollars. We want our dollars working as hard as possible. But again, this awesome platform, we’ve got all of this data, we’ve got this whole data library that we can pull attributes from. So why would we just use our own data? Let’s use some data that other people have that they make available that can be proxies to when customers are really starting to think about their problem that this product can solve. And so, what you’re actually looking at here is simple data that Google makes available. This is actually search activity by week for a very specific search term, that if someone’s searching on this term, it’s very likely they have the problem that this product solves. And you see there’s a high seasonal aspect to even the search activity around this term and really the problem that this product solves.

Zach Pike:
And so, this does a couple of things. One, it helps validate your sales patterns. Because the question is always, well, sales, yes, it’s driven by customer demand, but it’s also driven by is the product on shelf. Is it where it’s supposed to be inside the retailer? Are people able to find it? Did my competition do a big in-store promotion that shoved my product kind of behind the scenes while they were doing all their fancy stuff either at the beginning or the end of the season. That can substantially impact your sales numbers, which could hide the real seasonal aspect of this product. And so, this view here shows you a little bit of a simpler view of that.

Zach Pike:
So in this case, this is Louisiana. On the top, you’re looking at search activity. On the bottom, you’re looking at sales activity. And yes, you can just see the hotspots there. It kind of validates what we’re seeing. But if you really break apart this data and look at it from a statistical perspective, you find that there are weeks in here where search activity is hotter than our sales activity, meaning that on the tail end of the season here, we actually should be selling more product. Even on the front end, in some states, there’s areas where we should be selling more product if we believe that this search activity is a good proxy to sales which at this point we did and we still do.

Zach Pike:
So, imagine the value of that, if you’re walking into a meeting with a buyer at one of these retailers, and you can show them the data that says, hey, here’s my sales trail off period inside your store. Here’s what’s happening in Google from a search perspective in the states where I think I have more room to move. That can draw discussions around, okay, well, what’s happening on shelf during this time period? Are you guys loading up the next season a week or two early and that’s affecting my sales or am I not doing something from a marketing and execution perspective that keeps customers coming in because they obviously still have this problem later in the season than what I’m actually experiencing in sales.

Zach Pike:
So that stuff gets really powerful. Across all of our clients that do this, it’s always received really well. Retailers, buyers love it when you bring them data to one, help them understand their own seasonal aspects better, but also to show them how you’re supporting them by driving customers into their store.

Jan-Eric A.:
And the feedback we’ve heard is that this is a unique kind of look at what’s influencing business and is building credibility for clients as they’re trying to get increased distribution. You know, a lot of times, the effort to convince the buyer may not end up ultimately impacting how long the product stays on shelf because the powers that be, there might be other things that are coming in, they’re making seasonal overhauls. That’s a decision maybe that Walmart’s making. But we can still use this information beyond that because from a marketing focus standpoint, we can shift our attention to online sales via Amazon for example or other e-commerce opportunities, e-retail opportunities when we know that the distribution is going away, but based on the search information, the search query information, there’s a clear opportunity for sales, we need to make sure that we are ramping up our visibility in areas where the product is still available.

Zach Pike:
Yeah, definitely. One more thing, and this is like the coolest piece of our kind of data library that we have out there. But there are so many businesses that are impacted by patterns in the weather. You think about any type of product that’s seasonal, typically, there’s a weather aspect to the beginning and the end of that season. Even products, even things that you wouldn’t expect. Restaurants are impacted by weather patterns, right? So, we also process weather. We subscribe to the best weather data that you can get, which is offered by IBM. It’s the same weather data that you see on the app on your phone, on many websites across the country. It’s very widely leveraged. And we leverage that same data to help predict when seasons are going to start and stop.

Zach Pike:
There are sometimes in some products that we work on where a season can swing six weeks from year to year, meaning the start date of the sales season next year could be six weeks later than it is this year. And you can probably imagine, if you’re inside this company during that six week lag period and having no visibility to why that’s actually happening, it can be extremely stressful and it can cause a lot of fire drills and a lot of work that maybe wasn’t necessary. So this is where we really leverage weather data and this is a much more involved process where we’re using data science and actual statistics to get to the triggers that are actually driving the start and stop of a season.

Zach Pike:
In some cases, we see, like you can kind of see a little bit of it here in this case, this product, when it starts getting really hot, the sales trend really, really dies down very quickly. There’s a couple of very specific reasons for that. But in some cases, you see things like the low temperature has to be above a certain point for a number of weeks before the problem actually presents itself that the product can actually solve. And once you understand that, now you’ve got the firepower, one, to help your retailers make good decisions. You’ve got it to help drive a good marketing plan, which I’m going to show you next. But you’ve also got it to help with internal discussions and conversations and planning to say hey, based on what we’re seeing in the weather, we feel like this season is going to start late. It gives you a little bit of breathing room so that you’re not wasting money by starting your media budgets too early.

Zach Pike:
So, that was all of the analysis that happened through this in kind of a simplified fashion just in the effort of time. And really what it all led to was a plan from a media perspective that had a couple of aspects to it. The first one being, of course, everything was targeted to just the trade areas of these retailers. So, with programmatic advertising, Facebook advertising, all this digital stuff out there, we can be extremely targeted in the geographic locations that we’re going after, in addition to the typical audience targeting that we’re all probably familiar with. So that was the first piece of this media plan. But then we used all of this data and all of these attributes that we identified through this process to define our flighting schedules for every store, to define our budget for every store and then to influence the channel selection at a market and store grouping level.

Zach Pike:
And you’re seeing just some views here that kind of give a bit of visibility into how that plan was set up. But these were actually views that we published for the client as well. And the client was able to walk into these retailers and show them specifically, hey, in southern Texas, I’m starting media here and I’m turning it off here. And here’s how much money I am allocating to that flight window, and here’s how many stores that’s supporting. You can imagine a discussion like that, gets the retailer thinking, okay, well, what else do we have going on from a marketing perspective during that time period should we help this manufacturer of this product move more products? They’re sending people into our stores, maybe we should reciprocate a little bit. We’ve actually seen that happen a few times.

Zach Pike:
So that is the media planning exercise. We have a couple of additional views to show you though. So that’s all front end data analysis. At this point, we know a lot about the business. We have a really good understanding of what’s going to impact sales. We have a good understanding of where we have the most opportunity and we have a really solid media plan. But by doing all of that pre work, your measurement mechanism almost flows out of that automatically. So this is where we get to the back end data analysis that comes out of our front end work. But it’s much more informed than any other way you could have done this.

Zach Pike:
So I’m just going to show you some sample views. Again, this is just kind of a little bit of a show and tell to show some of the horsepower in here. And this is horsepower that Tableau gives us but this is also horsepower that’s a function of having the data clean and organized and flowing in as quickly as possible in an automated fashion. So this is, I mean, this could be a typical executive summary of business performance, right? I’ve got my core critical metrics there at the top. These are the things that everyone wants to know every day or every week. But then, I can overlay all of my promotion activity. So promotion happens across lots of channels. Facebook, display media, paid search, broadcast, radio, email. I want to understand that whole picture. So this is a simplified view, we could break this out however we want it to. We could throw a little filter down here in the bottom that lets you see specific marketing channels.

Zach Pike:
But I can get that visibility really quickly. This is kind of a comfort piece to make sure what I expect to happen is happening. But then, because you control this data at a level that it’s probably never been controlled before, we can bucket things specific to whatever’s happening from the media marketing execution perspective or any other type of breakout that is necessary to answer questions that are happening in the business. In this case, for this client, we broke it into four primary marketing periods. The reason for that is there were specific things happening in each of these periods that were meant to impact stores in a very specific manner. So when you’re doing that, I want to understand year over year what’s happening in store associated with that type of marketing execution. But I also want to be able to compare period to period. So I can understand, did marketing execution A, work better than marketing execution B?

Zach Pike:
So all of that stuff gets possible. It’s super efficient. It’s not like we’re waiting months to build a measurement strategy like this, since we did all that front end work. But of course, many products are very geographically driven. In a view like this, I can see what’s going on across the country. I’m seeing how much revenue I’m moving in each state. Just stay here for demonstration. We can do sales territories, DMA, MSA, whatever the market classification was. But then down below, I’m laying at sales over my marketing promotion activity. This is the view that everyone wants is, hey, when I bump promotion, do I see a resulting impact in sales.

Zach Pike:
This stuff isn’t like a big statistical analysis. This is a trend chart that I can look at and get a handle on what’s going on. But because we’ve got the data in a nice clean format, it’s really flexible, I can see this for just Colorado. I click on Colorado, and now I’m seeing my promotion activity happening in that state and I’m seeing my sales activity resulting from that. I can click on any state here.

Zach Pike:
So that is getting a little bit deeper into our ecosystem of retailers or sales. But what we often find is that any client that has a large retailer network, if you are only moving product through Walmart, that’s still 4000 locations across the country. We can give you visibility into the detail around that, right? It’s not difficult to break that apart from a market perspective or sales region, or DMA and show you what’s happening from a growth or decline perspective. This one actually shows you media support levels on the vertical axis. So even in a little snapshot like this, I can see my markets that I have the highest level of media support and what’s happening to them from a year over year product growth perspective.

Zach Pike:
But then down below, I’ve got a dot for every store. And this is sample data, but if I click on one of these, like I want to know what’s going on with this one, this market. I’m giving them a ton of media support and they don’t seem to be growing. So I can look in there and see what’s going on with each of the stores inside of here. I’ve got a couple of stores that are really far behind. I’ve got a couple that are only down two points, three points. And now I’ve got a really good understanding of what’s going on in that market. So you can imagine this can go much deeper, which is this next view I’ll show you here and then we’ll get to questions.

Zach Pike:
So, what often happens is we started a high level of business understanding. We take a couple of months, we really break things apart and that gives us as a marketing agency a lot of benefit, but it also gives the client a lot of benefit. It gives them a deeper level of understanding about the business that they may not have had before. Not because it was for lack of access to the data, it was for lack of understanding of the data because it had never really been visualized and synthesized in this manner.

Zach Pike:
But what this view does is breaks down everything to the minutiae of understanding in this case, retail store level performance. What we see every time is, as I get a good handle on my business, I get a good handle on my markets, I start wanting to go deeper into products. And then once I have all of that stuff set, I really want to understand what’s happening at an individual store level because I’m starting to do strategies on a store by store basis, different types of promotions.

Zach Pike:
So in a view like this, I’ve got a quick snapshot of all of my retail locations. I can scroll through here and see what’s going on. I can compare these stores to each other. But then down below, I’ve got the detail into the core critical factors that are affecting the performance of an individual store. So the way that this particular view is broken out is I’ve got three or four, four or five lines of store sales type performance, total customers, product volume trends year over year. Then down below, I’ve got visibility into what’s going on from a year over year perspective for my marketing support. How much broadcast media am I giving them this year versus last year? What’s my digital media look like this year versus last year?

Zach Pike:
So I click on Miami because I’m curious about them, I’m thinking about running a specific promotion for that store. I can see very quickly what’s going on with them and just a quick snapshot. Think about taking this to that store operator or thinking about having the discussion with your field marketing team about this store. So I can see they’re hurting on sales performance. Now they’re doing okay on pushing the product that we’re really focusing on. But I’m not supporting them like I was from a media perspective. We’re down 3% as far as 13% on the digital side. Compare that to a place, like let’s click on Myrtle Beach here. It’s a different story, right? They’re doing much better from a sales perspective but I’m also supporting them from a media perspective at a different level. I’m really killing it on the inbound call volume.

Zach Pike:
So this gives you that insight, and really this isn’t to answer every single question that you could ever possibly ask about an individual store. What it’s here for is give you a snapshot of what’s going on. So you can start to have those discussions to say, okay, why do we think this is happening? Why is this store doing better than its store down the street at similar levels of marketing support? What’s going on in store that could be causing that?

Zach Pike:
So that’s typically the process we see it go through and that’s why we build views that are at this depth. A couple of things I’ll mention then we’ll jump into questions. It’s all available at your fingertips. So 24 hours a day, seven days a week, accessible anywhere in the world. This is just the benefit that Tableau gives us with their piece of technology but then it’s all wired into that database. All that database is being filled by the data flow processing that’s coming out of all of those source systems that we’ve talked about on that overview slide at the beginning.

Zach Pike:
So hopefully that was helpful. I think we were going to pause real quick and let some questions come in. I’m going to hop back over to the slide here.

Jan-Eric A.:
Yeah, we’ve had some questions coming in which has been great. So if you have some questions and haven’t submitted them, please do. I’ll start going through some of these that we’ve gotten. This first one that came in is a great question and one that we’ve actually, we’ve gotten a lot from clients. The question is we already use Tableau, so what does your system provide that I don’t already have?

Zach Pike:
Yeah, that’s a really good question. I actually think everyone should be using Tableau. So Tableau though is only a visualization and data synthesis tool. Okay, there’s a lot of fancy analysis you can do in Tableau if you really understand the tool and you’re good analyst. And there’s a lot of fancy visualizations. Every visualization I just showed you was built inside Tableau. But the horsepower that is part of this platform is the data processing that empowers Tableau. So Tableau software is nothing without data. What we see a lot of times our clients who have Tableau, they purchased it thinking it was going to solve a lot of their problems. But unfortunately, they had not fixed all of the issues in the data before they bought Tableau. These companies are really good at selling their product and making it sound easy. But unfortunately, it’s just not easy.

Zach Pike:
And so, we’ve done the legwork to figure out that whole data mess. And it’s different for different clients, right? So, yes, Google AdWords data is consistent across every client. But the way we want a slice that data is different for almost everybody. That’s where the real benefit of kind of an all inclusive platform like this comes into play.

Jan-Eric A.:
Cool. Another great question that we’ve also worked through with some clients. Will the platform work for a client with limited data and inputs?

Zach Pike:
Yes. Depending on what works means. So, if we’re trying, we have to have the data to answer the questions that we’re trying to answer. If you’re, let’s just take a simple example, if you’re trying to answer, hey, how much revenue did I move through the Walmart location on Third Street and we don’t have the data to see what happened in that Walmart location on Third Street, we obviously can’t answer that question. But we do have several clients who have limited data sets. We have clients who are heavy in email distribution and CRM. In those cases, we don’t always have good sales data to tie us back to hey, I sent this email and I drove this sale and I can that attribution path there.

Zach Pike:
So in many cases, what we’ll do is we’ll find proxies. So we’ll find proxies in the data that we do have available that we can link over there with some assumptions built around it that everyone feels good about. But really, it’s a question of what we’re trying to answer. The other thing I’ll say on having limited data is sometimes the data is limited only because we haven’t really tried to acquire the data that we need. So these big retailers, for example, they all have really good data packages. And sometimes that data, whatever setup today to feed the data over is really messy. Maybe you’re getting an EDI and you’re not quite sure how to use it. Sometimes we can ask the question to say, hey, what are the other options to get this data or I’m looking for X, can you help me do that?

Zach Pike:
We actually just had a client two months ago who had really no visibility into shipments into their distribution network. They only knew the amount of product they were sending to the distributors themselves, not to the retailers. So we worked with them to help find with those distributors the source of the data and now it’s flowing. So now they actually do have that information. So a lot of times it’s just because we haven’t asked the right questions.

Jan-Eric A.:
Another question. You showed weekly data. What if I needed it updated daily? That’s really easy. I think that’s based on the availability of the data. If the data comes in daily, it can be reflected daily. The frequency of reporting is really based on how the data comes into the platform. The platform is not based on weekly or monthly or quarterly or daily. It could be hourly or by the minute. That’s just really about the minutiae, the data and how it comes in, correct?

Zach Pike:
Correct. Our limitations are only the limitations of the source we’re getting the data. So, one of the reasons I said at the beginning we use computers to do things that computers are good at, which are repetitive tasks, if we need to suck data out of a tool every day, we do that and we can represent the data daily if that’s important to your business. Or in some cases, annual data is what’s important. You can probably see how that could be rolled up pretty easily.

Jan-Eric A.:
Okay, here’s a good question. How do you get access to data from outside parties like retail partners?

Zach Pike:
Yeah, really good question and one that we deal with on almost every client. So, it’s a combination. A lot of larger retailers, your Walmart, your PetSMart, Petcos, Home Depot, Lowe’s, these big organizations out there, they want you to have your data because when you have the data in a good clean format, you’re making better decisions that’s only going to help them sell more product.

Zach Pike:
In many cases, that data is already floating around your organization somewhere, whether it’s feeding into an ERP system or it’s just sitting in a database somewhere that your finance team or your IT group had set up the feed and it’s been flowing for years. No one’s really leveraging it at the level it should be leveraged. But sometimes it requires getting creative. I’m not going to rehash it but that’s the example we went through a couple months ago where the client was distributing their product through a distributor network. And then that network distributed the product on to the actual retailer. So there was a break in the data. We worked with the distributors and actually with a third party company who already had these relationships built to acquire that data. So, it’s kind of varying solutions.

Jan-Eric A.:
A couple more questions here. You showed consumer data. What about B2B?

Zach Pike:
Yeah. B2B is of course possible. Many times, the insights you can get from B2B are really exciting because oftentimes, we know who specifically we are targeting. But as far as the platform and the database and all these connections are concerned, it doesn’t care if it’s Zach Pike or if it’s Callahan. It doesn’t matter if it’s a person or a company. We are looking at attributes and trying to figure out which of those attributes are impacting what’s going on from a sales or marketing perspective.

Jan-Eric A.:
Then other than what you showed, what other types of data can go into your system?

Zach Pike:
Again, there really are no limitations from a platform perspective or a technology perspective. The limitations are what data we have access to. So, I think I may have mentioned at the beginning, if we need to go out and find the population of construction companies in every city across the country, typically that data is available somewhere. If that’s a factor that’s impacting the success or lack of success on your business, we’ve got to go out and try and find it. It sounds simple but there really are no limitations. As long as we can get the data, get it collected and get it organized, we can load it in here and start using it.

Jan-Eric A.:
Well, I think we’re about to wrap up here. Those are great questions so thanks for submitting. Just kind of enclosing, for us at the agency Callahan, the intelligence, we’re not a data analytics company but clearly we place a lot of value in what it is and how it is, frankly it’s quite liberating for everything else that we do. As a full service marketing agency and we talked about what we deliver to clients is data savvy brand strategy and inspire creativity.

Jan-Eric A.:
Front end data analysis is really the beginning of that process. We’re taking data and connecting it to desired business outcomes and what we’re trying to produce. But it really ends up informing all the other things that we do. When we start to combine that with consumer insights and cultural insights and market insights, understanding what really drives business in the context of what’s happening with consumers and the world around them, we’re able then to start to develop that brand strategy and creative execution and media plans and also as we’re talking about starting to inform marketing analytics and that back end reporting and it really brings it full circle.

Jan-Eric A.:
I can say as someone who’s been doing this for a while and starting my career in media planning, having a resource like this at our disposal has been very liberating about how we go about this.

Jan-Eric A.:
I want to thank you all for attending. Hopefully you found this show and tell to be helpful. If you have any additional questions or follow up, please reach out or visit us at Callahan.agency. We’re also going to make a link to this webinar and the recording available to you once it is available. So again, thank you very much for joining.