I want to thank everybody for joining my session today. I’m sure you’ve already heard many great sessions and I hope to continue on with my session today as I talk about what I call the truth about marketing analytics. That title might be in your face a little bit, but the idea is that—we’ve been around since 2007— so in today’s session, what I really want to share with you are three primary principles that we’ve learned, being in this business for so long, that anyone who wants to be successful in the marketing analytics space really needs to recognize and overcome. Those three principles are very near and dear to us here at Alight, and again, I want to take some time to share that with you today.
Speaking of Alight, again, we were founded in 2007. We have a platform called ChannelMix, and it’s what we call a full-service data aggregation platform. The whole premise around ChannelMix for use, when we began, was to create this single source where every media and marketing channel could flow into to allow us to do these things that you see on this slide. From strategic services to providing Google Analytics services, to making sure that we’re tracking all the inbound media effectively, to doing analysis and modeling, to custom reporting, and obviously having database developers and data architects work on the platform to be able to really drive what we call robust marketing analytics.
There’s couple highlights that you’ll see out there, ChannelMix, we believe, is the first data aggregation engine that was built all the way back in 2008, was when it launched. We have a services history, so I think one of the things that makes us a little bit unique when it comes to marketing analytics is we’ve lived and breathed a lot of the pain that agencies are living through and marketing folks and brands are living through, which is trying to coal through all the data to find out what works and what doesn’t work. We’ve lived it. Services are a huge part of what we do each and every day here at Alight, so we understand the context. The other thing that makes us unique is having all of these skillsets under one roof. I think to be able to attack the marketing analytics challenge, you have to have really smart people that bring different skillsets to the table surrounding a single source of truth, as we like to call it, which is ChannelMix.
That’s a little bit about us, and that’s not really what I’m here for. What I want to talk about today is, first and foremost, when you think about marketing analytics, what should it deliver? We talk about analytics in the context of dashboards and putting things in motion and attribution modeling and all these catchy phrases, but if you step back and think about what marketing analytics should deliver, we think it should deliver three things. First and foremost, it has to deliver efficiency. As you think of your marketing analytics strategy, think about your current world—for us, I’m going to share what our process used to be—but typically, when it comes to marketing analytics, we’re very, very inefficient. Because of the way that data exists, where we have to go get it, we have to piece it together, we create this QA mess. So if we’re going to implement an analytics marketing strategy, it has to be efficient.
Secondly, it needs to drive value. I’m going to ask you today to stop thinking about the term “reporting”. I want you to think about the term “performance storytelling” because that’s really what we should be doing. We shouldn’t be spitting out reports that have metrics associated with them, we need to be able to tell performance stories. When we do that. We’re delivering a higher level of value than what we have typically done in the past. Lastly, it needs to deliver revenue. Data in and of itself is meaningless. But when we bring context, in an efficient way, to that data, that helps me tell a story—I can use that data then to drive revenue for my organization. If you’re an agency, the means better client relationships. If you’re a brand, that means you’re maximizing those budgets to build stronger brands and drive sales and maximize budgets. At the end of the day, when you think about marketing analytics, these are the things it must deliver.
Standing in the way of this great hypothetical slide, are three principles. The core of what I want to share with you today is: if I want to deliver efficiency, value, and revenue through my marketing analytics strategy, I can tell you that—having been in this business so long—you need to overcome these three principles. Number one: we have to solve the data problem.
I use this phrase a lot when I speak around the country and the phrase is this: “Marketing doesn’t have a dashboard problem—We have a data problem.” To that end, I thought I’d share with you what we used to when we first started the company. Many of you will probably recognize this process. You have all of these sources, and these sources could live—TV, radio, print, it could be email, it could be ad serving data, paid search data, social data. We have data that’s everywhere. The story for us was, our first client was an agency and our entire job was to deliver succinct marketing analytics dashboards that helped the agency tell their clients what’s happening.
The problem is we have all these sources and, literally, we would have to go out and manually download all this data. We would use marketing’s best friend, Excel, where we would assemble all this data. Then we’d get crazy and we’d build some charts. Then, like every good marketing analyst, we would take those charts and we would stick them into PowerPoint. That’s where I say your data is officially dead. No more can I use that data because it’s now frozen in this power point that may be 5 slides deep or 80 slides deep, but I can’t use it anymore. When the other co-founder, Michelle, and I would go through this process and hand these reports over to our client, we would pray for one thing. And that prayer was built around: “Please don’t ask me any questions.” Because we did not want to have to go back through this gauntlet to try and figure out the answers to those questions because this process is not efficient.
Some of you probably recognize this process. I call this process the Data Death March. You go through this entire marching process only to have it end up in PowerPoint, and you’ve officially killed your data.
The effects of this process—as you think about your organization—as it related to us, as we started hiring more people and bringing people into the Alight Analytics family, what we found was: our teams started operating in silos. So team one wasn’t talking to team three because everybody was creating their own separate data sets. Data began to get stored everywhere. And we had no ability to be proactive. Ultimately, we were spending more time gathering data than analyzing results. Remember, we’re a marketing analytics firm. What we should be doing is have the ability to analyze and deliver really powerful stories. We simply weren’t able to do that.
In fact, this is what our process was. We stepped back in 2008 and said, “Where are we spending all of our time?” 82 percent of our time was spent, literally, gathering data. 18 percent was left to try to analyze and tell a story. We knew this wasn’t a recipe for success, long-term.
What I tell everyone today in the industry—forget about tools, forget about everything that you know and understand that, in order for you to be successful in marketing analytics, you have to have a two part solution. What we call data aggregation; the ability to centralize all this data into one single source of truth. If I can do that, I can do anything I want on the data utilization side. Once I have all that data together, and am able to create that single source, it truly sets me free to be able to do anything I want on the data utilization side.
So principle number one: stop the Data Death March. Rethink your problems. If you think about that manual process, it’s rooted in the fact that we as a marketing industry have a data problem. And we need to focus on bringing all that data together in an automated fashion, and then use that data however I want, with whatever tool I want. So aggregate the data first, then report it. Principle number one.
Now let’s talk about principle number two. I mentioned this a little earlier, and frankly, reporting metrics isn’t enough. If you think about the process we have to go through today—for us it was that 82 percent of our time gathering all this data with only 18 percent left to truly analyze it—what the outcome of that process is, is really bad reporting.
A typical campaign report for us may look like this, where we would talk about total impressions and totally clicks and maybe some Google Analytics stuff on a separate chart or we talk about sessions and bounce rate. Then we use very technical terms like majority, percentage of things broken out by media. This reporting doesn’t tell me anything. All it’s doing is justifying what I did. And if you’re in a place where you’re using reporting to justify either the budget that you have in marketing or the services that you’re providing, then you need to stop thinking that way and start thinking about, “How do I use analytics to drive results rather than justify my existence?” I can tell you—in talking to thousands of people across the country—most of the time, because of our data problem, we’re forced into reporting like this, which turns into a justification discussion. We don’t want that. So stop thinking about metrics.
What I want you to think about is a performance story. Just image, if you will, that all this data is together in one place. Maybe I’m using a data visualization tool like Tableau, which is the tool that we’ve grown up with here at Alight. We’ve been using that product since 2008. Tableau is a fantastic product. We have the freedom to use whatever we want, but it allows us to tell performance stories. What I mean by that is something like this. Let’s say you wanted to have a marketing funnel where your media costs and your impressions and you’re talking about what your media mix is up top. What I the average cost of an impression? And how many impressions does it take to generate a visit to my website? And how many visits does it take to generate applications started or applications completed?
You can begin to build the metrics dynamically that help me tell a story to do things like: what is the media impacts funnel to understand investment performance? I know in this case, I spent 383 thousand dollars to generate 231 applications. Is that good? Is that bad? I don’t know. Let’s look at the media mix and start to ask questions. The job of a dashboard should be, number one, to help you ask better questions. In this case, if I continue down, I can identify key performance behaviors within that media funnel and align my spend and engagement to principle conversion events or actions. It’s not talking about media share or percent share or button clicks. I’m talking about: how many conversion events did I deliver with this media mix and what is my ROI? That’s a performance story.
Here’s another example where I may be able to construct my overall media coverage, both online and offline to core markets. Maybe I’m in a retail environment or maybe I have a B2B sales team throughout various parts of the country. How do I establish the media mix trend that supports sales and engagement—either in my brick and mortar or with my sales teams—in those geos? And then be able to align my market response and my market need to proactively change how I’m spending those dollars? Performance stories help me visualize and really tell a story about how we are spending our money, where we need to get better, what’s winning, what’s not working as well, and what are we going to do in the next quarter, the next month, the next year, for example. That’s what performance stories allow me to do.
The last thing I’ll share with you is, when you think about constructing that performance story, you need to think in terms of this hierarchy. All of these systems that we have—AdWords, Facebook, YouTube, Twitter, and Instagram—they all produce channel reports. That’s where I should be optimizing, so this lower level is the detailed reports, cost-to-value metrics, etc. However, the examples that I just showed you talk about two really important things. In that middle level, we’re aggregating all of my data by channel and by campaign. When I do that, I can have a clear insight into my performance, I can drive planning, I can drive forecasting—once I gather all that data together through a platform like ChannelMix—to be able to then tell the one to two-page value story.
If you’re an agency or you’re in the marketing department of a large brand, you want to be able to walk into a room, with your client or your stakeholder, and say, “Here’s what the campaign did. Here’s the results. Here’s the interactions that we created. Here’s all the results in a one to two-page value story.” When you are able to do that, it changes the way you are able to drive marketing strategy within your organization. This pyramid is extremely important. It talks about the data flow of building it from the ground up all the way to the value.
Principle number two: reporting metrics isn’t enough. We have to break free from that. We have to stop thinking about clicks and impressions to things like: how many impressions does it take to do X? That’s when we start telling performance stories. That’s principle number two. So principle number one: we’ve got to fix the data, principle number two: we’ve got to tell a performance story.
Then principle number three: this is a culmination of bringing both of these principles together into one overarching process, or I should say structure.
I call this the four Ps of marketing analytics. It’s a play, obviously, on the four Ps of marketing. But the four Ps of marketing analytics utilized this two-part solution: data aggregation, data utilization. Then the four components, or the four Ps within this structure are: platform, process, people, and plan. These are the four things that I can tell you, you have to have to be successful. These four things are shared across the data aggregation side and the data utilization side. On the platform side, we need data aggregation. We need an engine to pull all this data together. When we pull all of that data together, we need technology, not people to monitor and automatically capture all this data.
Then we need people who understand how to work in database and develop data structures and know how to code. I can tell you that I talk to marketers all the time and they ask me things like, “Well how do I do a union statement?” or “How do I create an outer join with this dataset to this dataset?” My response to them is: “Why are you doing that?” Marketers should not be tied down to trying to figure out how to code and SQL and Java. That’s not where you should be. You should have people who understand how to code in this space. So the people side is data management resources. Then you’ve got to have a plan. When we talk about bringing all of this data together, it’s not enough to say, “I want Facebook and YouTube and Twitter and Google Analytics and AdWords and TV, radio, and print.” Great. I’m going to bring this all into a single platform.
But what are the questions that you’re trying to answer? What is the taxonomy that I’m using? How am I going to serve data across multiple channels and have the same campaign name? You have to have a plan around your data. These four elements are absolutely fundamental, foundational to being successful in marketing analytics. When you do these things, you can start to think about, “Well, what is the story I want to tell?” I know the questions that I want, but now how do I want to tell that story in my plan for reporting?
Then people—who’s actually going to build this reporting? I have four different people working on one report. That’s not efficient. That’s not going to be valuable long-term, and it’s surely not going to help me understand the revenue proposition that marketing is delivering. Who are the right people to be working and doing analysis? Then how do I build dashboards that are automatic? Even if I use Excel, how do I build charts and graphs inside of Excel that I don’t have to touch, that I just have to refresh the data and they automatically update? Lastly—and I encourage everyone—if you’re living in Excel and PowerPoint, please stop. Please start to look into the future and think about the right business intelligence software that can help you tell that performance story. They will inherently bring with it all of the automation as part of that software package, but you’re ultimately going to need people—and clearly you’ve got to have a reporting strategy.
So when you think about the four Ps, they line up to give you eight components to make this successful. Here’s the problem: sometimes we don’t follow this process that I’m going to show you. You’re going to select your data aggregation platform, you’re going to figure out how to bring all this data together, who’s going to manage the data, what the right measurement strategy is, then you can tell stories, find the right too that you want, and it becomes this circular process all built around these four Ps. Too many times though, I will tell you that people start in the wrong place, and I was one of them.
I bought Tableau software in 2008 without fixing the data problem. If you start with data utilization software, then you start to ask all these questions. How am I going to get all this data in here? Who’s going to actually use this thing? How did we get this product? I need somebody to write some code for me to make this work. Maybe I get a dashboard one time, but how am I going to automate it? Then I have questions about what does this platform actually do? Maybe we should step back and create a strategy around this data and so on, and so on. It becomes very chaotic. I can tell you 9 times out of 10, people call me and say, “I can’t get marketing analytics to work.” It’s because they started in the wrong place and they blame it on the tool.
As I said in the beginning, marketing doesn’t have a dashboard problem, we have a data problem. We have to fundamentally fix the left side of this equation, move over to the right side of the equation, and build our process and structure around these four Ps. When you do that, I promise you, you’re going to be successful long-term.
Principle number three: it takes more than software to succeed. You’ve got to embrace the four Ps of marketing analytics. When you do that, when you understand that this is more than a software tool problem, you’re really equipping yourself to be successful long-term.
To round up the end of this story, here’s how we solve marketing analytics at Alight. As I’ve mentioned, we built ChannelMix, and ChannelMix is available to anyone on the call today. ChannelMix is this Big Data engine that brings all this data together, that collects it, cleans it, stores it, blends it, and prepares it. We have a dedicated support team that manages all of this data that ultimately gives us one connection to use whatever tool we want. Inside of Alight, our reporting team Excel, Tableau, and we use R for all of our modeling. The reporting team has the flexibility through this one connection to really spend time where they need to be spending their time, which is telling performance stories. So I know this process works, it’s been proven. I’ve been doing this now almost 10 years, as I mentioned. This is how we’ve solved it.
The other thing that’s super exciting when you centralize your data, you can then organize around it. Again, sharing our story, we have multiple teams that are now working off of the single source of ChannelMix, using all of these products, the tools that they want to be able to tell the performance stories. Then we have Marketing Analytic Academy, where I spend a lot of my time today actually teaching academy about best practices and things that I’ve done wrong and things that we’ve learned as an organization and really helping equip marketers to be successful in the marketing analytics space.
Let’s go back to this and I’ll wrap up today. When you think about marketing analytics, these are the three things that it must deliver: efficiency. And efficiency is the first thing that I need to focus on, so I would ask you to think about—data is what is going to drive my efficiency. Think about the condition of your data, think about all the sources that you have, think about a process by which to bring all of that data together into one place because when you do that, and it’s being actively managed, you’re going to have better use of resources, significantly reduce your QA risk, and you can have a standardized approach to marketing analytics.
It all begins with the data. Each on of these then builds on themselves. As I become efficient, I drive greater value. My quality of reporting will change drastically. No longer am I reporting on just impressions and clicks, I’m actually telling performance stories, which shows that I am being better stewards of my budget and it helps me be proactive. I can’t tell you the biggest challenge in the value component here is as marketers, we just aren’t proactive because we can’t get to the data. So when we become efficient, we can become proactive. Then the last thing, of course, when we’re driving great value, we can ultimately drive revenue for our organizations. We can maximize our budgets, drive sales, and build stronger brands, which at the end of the day, that’s what marketing analytics should be delivering for any organization.
That being said, I want to thank you for attending my session. I know that you’re enjoying a lot of great information and you’ve probably attended a lot of great sessions. I appreciate you sticking with me through my session. Hopefully, I’ve shared some things that are meaningful to you. Obviously, if you need to contact us, we’d love to hear from you. You can reach us either through our email, our phone, or visit us on the website at alightanalytics.com. Thank you again, and I hope you enjoyed our session today.