Ken Holladay, ObservePoint - How MarTech Management Closes The Data Governance Loop

November 7, 2017

How MarTech Management Closes the Data Governance Loop

Slide 1:

I always feel like the intros are a narcissistic hype, right? That everybody gets to feel really self-interested and self-important. I think it’s important to know a little bit about me so you can know my background. I’ve been in technology my entire career, specifically tech startups, so I’ve seen the good, bad, and ugly of data and data management.

I’ve been here at ObservePoint for two years, being at the kind of forefront of data governance. It’s been a really eye-opening experience to understand the great things organizations are doing and generally the community of analytics and how people really are truly trying to collect and give the best insights they can, but there’s a lot of big barriers to that. Hopefully we can address some of those today as we get into the topic.

Slide 2:

I think the first question to ask is: what is a MarTech stack, or what is MarTech? “Marketing technology, or MarTech stack, is a collection of different kinds of software given to execute a given brand’s marketing strategy.” I don’t know who Todd Grennan is, but a MarTech stack is just everything we use to collect the information that we try to measure our company and try to measure our businesses progress. I think the key phrase in all of this is “marketing strategy.” It is the collection of technology to execute our marketing strategy.

Slide 3:

That’s great when in 2011, this is what the landscape looked like. A lot of us might have even had our initial digital strategy designed after what was available in this landscape.

Slide 4:

Fast forward to 2017, and this is what the marketing technology landscape looks like. It is a significant difference. There are over 5,000 marketing logos on this super graphic, and it both terrifies and inspires me. It inspires me because it shows the drive other companies have to put their customer experience at the center of everything they do. There are so many companies out there that are trying to provide that through the software and services they provide. But it terrifies me because how in the world do you manage technology? How do you manage a landscape that is so vast?

Slide 5:

The question is: more MarTech, more problems, and according to an April 2017 study by Internet Trends, the average number of MarTech solutions in an enterprise marketing department is 91. 91 solutions in an enterprise marketing department. And what they said was that 97 percent of those solutions they wouldn’t deem as enterprise ready. If it’s not enterprise ready, that opens you up to a myriad of data security and compliance issues. It really becomes a concern for your organization because you’re dealing with software that’s not fully established.

Another aspect that it opens up is with this much technology, 80 percent of marketers say that data silos within marketing obscures a seamless view of campaigns and customers, and that it is the number one issues and number one challenge of their organization is to get that quality and completeness. That’s a significant issue of you’re trying to drive vision, goals, KPIs in your organization around that data when that data is so fragmented and clouded by all the quality issues that you’re facing. So that seems like a significant mountain to face already.

Slide 6:

There’s more to that. We in the data world are now asked about Big Data. How are we doing Big Data? How are we leveraging Big Data? We feel increasing pressure on using Big Data and showing insights on that Big Data. But with all the effort that we have to put into managing all the technologies, the 90 plus technologies that we have in our MarTech stack, we’re left with very little time to actually analyze and find that meaningful insight. If we take the inverse of that we say we’re just going to focus on getting the insight, there’s always that voice in the back of our minds with the doubt of the quality of the data because we haven’t had time to manage the collection points. We don’t really feel like we’ve hit that quality-to-insight balance and it feels like, with the Big Data on one side and the almost unmanageable MarTech on the other—how do we deal with this? But wait, there’s more… even more.

Slide 7:

We have this concept of dark data, which gives is a really broad perspective of everything that is in our organization. Dark data is defined as, “Information assets that organizations collect, process, and store in the course of their regular business activity, but generally fail to use for other purposes.” This is, as the name kind of implies, dark data that we could be accessing, but we don’t.

Slide 8:

I found this great image on the internet. Google has done me well and given me this image. I have no credit on the data behind this, I don’t even know where it came from, but I think it does illustrate the issue that we deal with in analytics, but also the information we deal with in business information in general. That 12 percent of the data is business-critical. That’s 12 percent. How do we find that 12 percent? How do we manage that twelve percent? How do we guarantee that 12 percent is accurate and is what we want?

Underneath that, you have 20 percent of the data that is collect, which is called ROT data or redundant, obsolete, trivial. Now that ROT data adds a lot of misinformation. It can skew your data points it can be some of those vanity numbers that really don’t add value to the business, but can distract heavily. I think it’s really impactful. And I don’t know where this number comes from, but 3.3 trillion dollars in the industry will be the cost at 2020. And I don’t know if that’s accurate, but I do know that I do think we all have felt the pain that has come from data that is obsolete or trivial that really distracts us from our goal.

And I think the key to this whole illustration and dark data is that portion, that 60 percent of that data that is untapped, that is not collected because we don’t have the tools to collect it or because it’s such Big Data, we don’t have any analytics around it to provide insights or the data is missing or incomplete. All these things can drive us crazy because we don’t know what we don’t know, and we don’t know how to sift through that or what technology we should use to collect that. Can we fix all that today? No, but I do believe that if we leverage our marketing strategy and have a deliberate focused strategy, we can better collect insight and drive our decisions based on the data, which is what we all want to do is to be a data-driven organization.

Slide 9:

Now we’re going to talk a little bit about spending. All of this is a little daunting, but to add a little more drama to this image that I’m painting, in the past year, 57 percent of survey respondents said that they expect their marketing budgets to increase, which is great, right? We have more budget to put the right data in play or the right tools to collect data and store and all of those things. But a similar study shows that increase is going to be incremental in 2018 and that we’re going to see less focus or less budget to those problems. Those problems are doing to continue to grow exponentially. So that puts us in a really hard place that we have to do more with less. That’s difficult because it seems insurmountable, these issues, but I think you can put together a really strategic plan to be successful in your data governance and in your marketing technology strategy.

Slide 10:

I think there’s these four points that really help us drive through the strategy around building this strong practice. First, you want to assess your technology stack. You’ve got to know what you have, what you don’t have, and understand what is there. You need to assess your strategy. Document your findings. And then you have your ongoing strategy. And this is an image that comes from Forrester that talks about building that success in digital intelligence. And I think that strategic approach that we’re going to talk about here, it shows that it’s a foundation and I truly believe that if you have a solid foundation in your MarTech strategy, you’ll be able to more accurately manage your data and make sure you’re doing things adequately.

Slide 11:

First thing, assess your tech stack. First and foremost, I think auditing your marketing stack is so easy with ObservePoint and ObservePoint’s tools. If you run a quick audit, you can get back an understanding of what technology has been living on your digital assets. It gives you a baseline. It’s difficult to dig through this source code or dig through your TMS to see what’s being loaded, what’s not. Not everything is loaded in TMS, so using an audit is a great way to get that base understanding of what’s out there. And it also helps you manage and measure how much you’re using your technology.

It gives you the baseline around what pages do or do not have it, and it lets you see you have X technology, but you don’t use it at all. I think it’s something that’s very underutilized and overlooked, but can become a huge pain point in an organization is when we don’t know who is responsible for the technology when something goes wrong. So, I think the first step in a sound strategy is to see what you have, see how much you use it, and then know who is in charge of it if something goes wrong. And that provides that go-to point when we have unexpected issues.

And lastly, I think it really helps, even at the analyst level, to understand how much we’re paying for the technology that’s in our stack because since we see that there’s going to be an ever-limiting amount of budget that goes around, data management and data collection, and just your marketing technology in general, being more frugal as it comes to technologies we use and utilize and how much we pay for them can help us use our budget to the full effect.

Slide 12:

The second step of having a sound MarTech strategy is to assess your strategy, to understand what you have and what you’re’ trying to do with it. I think with anything, a goal that is unmeasurable is not really valuable. It really doesn’t matter if a particular variable is firing, if it’s showing up on a page, or if a technology is even there if you don’t know what you’re doing with it in the first place. If you’re trying to collect eVar 22 and you’re evaluating whether eVar 22 is firing, but you don’t really know why you even want eVar 22 on your page, it really becomes kind of a burden and part of that ROT data. It could be obsolete, it could be redundant, but you don’t really know, you’re not measuring it correctly and you’re not attaching that to something of value.

I think first and foremost, you connect your technology, your MarTech strategy to your KPIs. I think we do this in some ways with an SDT tagging plan, but I think we need to be more deliberate about how we do that. And we’ll talk a little bit more about that in slides to come. Lastly, your tagging plan, which is kind of the nuts and bolts of what should be where. But I think more than just having what should be where, to document who owns each of those technologies, what that technology is supposed to be doing for your organization, and what that technology is connected to. So that you can easily reference what KPIs can be effected if there’s an outage. You can easily contact a person on your team to deal with errors or inconsistencies in the data. I think tagging plans need to grow beyond just: this should be that and should be here. It should grow much further into a strategic document.

Slide 13:

Let’s talk about documentation. I think with anything, the first question you ask is: why? Why do we collect this data? Why do we care that we have X visitors coming to the site and so forth? The first thing to do is start with your business vision, understand it. What do we do as an organization and why are we doing it? Then document the company goals that are associated with that vision. You can sue spreadsheets, something we use in the industry as a standard, but any documentation, whether it’s a Google Doc or whatever it is, just writing it down and connecting it to your vision will help you have clear goals and help your organization be more focused in what they’re trying to achieve.

Then connecting those goals to KPIs so you can measure them and see whether you’re succeeding or failing. I think all organizations at some point or another have done this, but I think the real key in our world, in the data governance space, is to bridge the gap between business goals and business measurements and connect those into your tagging plan. I think this is the gap that will really help you do better data governance when we have a strategically planned Marketing stack. When we know what we’re trying to collect, where we’re trying to collect it, what it should look like, but we also know why we’re trying to collect it. We know who needs this piece, what KPI is using this portion of data to calculate the outcome.

I think this is a puce that should not be overlooked and is really valuable to get business buy-in to really have the whole organization focused on your marketing strategy, your whole organization focused on finding good quality data through data governance. Once you’ve documented what variables correspond to which KPIs, I think the next piece is to hook that into the ObservePoint Rules Engine because the Rules Engine does something that humans can never do which is contact thousands and thousands of pages and check variables and do all these things at scale, repetitively, and really helps drive the data quality effort to a minimum because you have a machine that’s constantly validating that the rules that you’ve set are going to be measure accurately based on the data being collected.

Then you run your audit. Audit all the digital assets you have. Use all those rules to validate the variables that are coming in. what this will do is provide results and those results then are easily identified as: this failed or that failed. Then you can take those results, measure them against your tagging plan, and you can see this trickle effect as it comes back out. You can see which KPIs are going to fail because the data is inaccurate and which ones will provide a poor measurement of the goals and it kind of has this very cyclical effect that really drives value.

If you were able to, through documentation of your tagging plan and the connection to your data governance practices, if you were able to show the greater organization that we say we will do X and this is our vision and here are our goals, but the things that we’re using to measure that are wrong and failing, it really drives that message home of the importance of data governance.

You might say to yourself, “This is over the top, I’m too busy. I’ve got my boss breathing down my neck for all these reports.” The first thing I would ask is: how can you afford not to do this? How can you afford not to connect the measuring tools to the goals that are trying to be measure? I know it’s difficult, I know it seems like an undertaking, but I think that the second point I would say is it doesn’t have to be a companywide project, it doesn’t have to be every single goal or every single KPI your organization is trying to measure.

It really can just be one. One critical piece, one very visible piece, something that the organization—when you’re able to build this very focused strategy around one KPI—can get behind because they can see that measureable distance in the accuracy. They can see the correlation between the goal, the KPI, the data collected, the quality of the data, and that brings it all together. The purpose of this marketing technology in the first place is to be able to understand what’s happening and to really change behavior if needed. To drive better customer experiences. To improve the business through the data.

You may be the only one doing it at the beginning, but I would bet that if you did this for one KPI and if the leader in the organization saw that someone was actually trying to connect the strategy to the data that it would catch on. They would get excited because they care about those KPIs, they care about those goals. And if you’re able to give them a clear understating of where it’s coming from and why it’s coming or how good or bad that data might be, I think everyone can get excited around the possibility of that clarity. Plus, it’s a great way to cover yourself if something was to go awry and you say, “Well actually, I have this document that shows that this KPI is measured by X, Y, and Z. So, it can’t be these data points because we don’t measure that KPI with these data points that you’re claiming.”

The result will show you where your fail points are and allow you to see what KPIs will be off and the underlying broken data. I really think everyone can easily understand the value that this strategy would provide and it really doesn’t have to be a massive undertaking.

Slide 14:

The last piece is governance. Governance is obviously kind of this daunting thing. A lot of organizations have governance, some don’t. but I think that the most important piece to understand that governance is an automated validation and not artificial intelligence. You can’t set up governance, whether you’re using ObservePoint or you’re trying to do it manually, you can’t just say this is the way it is and just let it roll. You have to be maintain it from time to time to keep it up to date with your strategy, with your vision, and really have some proactive behavior and some very measure effort towards your data governance so it’s more valuable to your organization and doesn’t go stale and start being counterintuitive.

Like I said, most organizations don’t have data governance or data governance boards, so an automated system is really important to provide that measured audit and quality balance to what you’re trying to collect and measure against. If you can show the connection between the goals and the quality of the data you measure, I think you’ll have more buy-in from the organization, you’ll have more people interested in the value of data governance and you’ll have that excitement around the possibility of measurable strategy and data and insight.

Slide 15:

I want to leave you with this very cheesy thought: how do you climb Mt. Everest? I think it’s intuitive, right? One step at a time. I would liken that unto data governance and MarTech strategy. How do you do MarTech strategy? You do it one step at a time. You take one KPI, you connect that through the full cycle, and that will drive energy and excitement around it. And I promise you that if you, in 2018, focus on the why around everything you do—why we collect this data, why we use this technology, why we care about this KPI—that will drive more insight. It will cut out that noise from Big Data, it will help pull out some of the dark data that is found in organization, and help you focus on finding better insight and data that you may not have previously collected, and it will help you drive a better marketing technology.

Thanks for your time. I feel really fortunate to be part of this conference and I think we’ve had a lot of good speakers and hopefully you’ve enjoyed some of the insights I’ve shared. Feel free to reach out with any questions. I hope you enjoy the rest of the conference.

Previous Video
Adam Corey, Tealium - How GDPR is a Prime Opportunity for Data Professionals
Adam Corey, Tealium - How GDPR is a Prime Opportunity for Data Professionals

Unpack how to leverage GDPR and the work required to help establish data as a priority in your organization.

Next Video
David Rogers, Slalom - How to Build a Testing Optimization Center of Excellence
David Rogers, Slalom - How to Build a Testing Optimization Center of Excellence

With so many moving parts across an organization, how do you get started thinking about a best testing prog...