Zen and the Art of Sustainable Implementations

November 2, 2017

Zen and the Art of Sustainable Implementations

Slide 1:

I’m Jason Thompson, cofounder and CEO of 33 sticks. I’m very excited to be here. A lot of great speakers at this summit and I’m definitely honored to be part of it. This is our second year presenting so that either means that there was a spot that needed to be filled or, hopefully, there was some good feedback from our session, that it was enjoyable and we were able to provide some valuable information. Hopefully it was the second and hopefully today’s session will be just as valuable.

My talk is all about sustainable implementations. And you might be thinking to yourself that this is an odd topic to have. It’s 2017, the digital analytics movement has been around for well over a decade now. Why are we still talking about implementations? The promise is that we should be using data to do amazing things within businesses, but the reality is that we are still focused on implementations.

Slide 2:

This is an important topic for me because when I started 33 Sticks about five years ago, I was wanting to be more in the business of making the best use of data. To me, that’s the really exciting part of analytics, being able to use data to inform business decisions, running optimization campaigns, doing some really cool things with personalization. However, we quickly found that many organizations, large and small, simply did not have a solid enough implementation to build these more sophisticated things that the business was looking for. And we’re seeing that all over the place.

The gap between what businesses want from data, and what our analytics teams are able to provide is widening. And this is concerning to me. Again, we’ve had a good 10 to 15 years to get implementations right. This gap should be getting smaller, not wider. And it’s not just what I’m seeing. There are some interesting studies out there. There’s one from Duke University that’s showing a very concerning trend that year over year, we’re seeing a decrease in what organizations are spending on project related to marketing analytics. Meaning that more and more projects are going out that aren’t making use of analytics.

Why is this? Again, we should have all of this great data to inform how well our projects re performing, better optimize those and position them to target customers. Yet it’s not happening, so it’s a very real concern. I think a symptom of the underlying problem of analytics implementations that haven’t sustained the test of time and are not providing business leaders the data they need.

There’s another study out there by MIT that show upward of two thirds of executives have openly admitted that they rely more on their experience, their gut instinct, than data. Part of that is understandable, as humans we’re very much instinct driven. We like to believe that our experience is the most important factor, but it also kind of flies in the face of what we’re hearing from many executives that they want to be data-informed. They want to have data available to them to make their job easier to make educated decisions.

And again, this widening tells us that they want the data, they’re simply not getting it. And this is a very concerning trend that I think point directly to underlying implementations not providing these executives with the insights that they need.

Slide 3:

So, what I’ve come up with over my 15 plus years of experience, seeing well over 100 implementations, is what I think are the top eight reasons why. There’s no silver bullet for fixing implementations that are broken and not sustainable. And by no means do all eight of these apply to any one organization. However, you’ll probably recognize two or three that are happening in your companies. And hopefully I can provide you some insights, and if nothing more, force you to think a little bit more critically about how your organization is leveraging data, how the underlying implementation is supporting what your business need to run as a highly informed business operation using data.

So, let’s go through each of these eight reasons and hopefully you’ll get some ideas that can apply to your business and you can start working collectively on building more sustainable implementations because, if we’re presenting at this summit in 2020, I would much rather be talking about the really cool things we’re doing with data, it should be assumed that people are doing implementations the right way. It’s time we get that done. So, let’s jump right into it.

Slide 4:

Number eight. No one sticks around. This is a massive problem. If we want to build things that are sustainable, we absolutely have to have smart and talented people sticking around to maintain whatever it is we’re building. And if we don’t have people sticking around, our ability to create anything that’s sustainable is going to be severely hampered. I was recently chatting with Corey Prohens, who I believe by far runs the top analytics placement firm in the industry, in digital, IQ Workforce. He reports, and he’s been reporting a similar metric for the last several years, that the median tenure for an analytics professional is 1.8 years.

Let that sink in for a minute. Let’s look at this on a spectrum to see where that fits. If we look to other professions where we see people spending a lengthy period of time with a single firm, really develop deep expertise, understanding the client, really becoming valuable to both the company and the client. We look at architects, they’re staying five and a half years. Folks in the legal profession, 6.3 years. We have to go all the way to the other side of the spectrum to find digital analytics professionals. We are at 1.8 years.

To put that into context, an industry as we know has massive turnover problems, food services, is 1.9 years. As a digital analytics professional, we are staying in our positions within companies less time than the average person flipping burgers at McDonald’s. that is a major issue.

Why is that happening? Just as a there are a lot of reasons why an implementation isn’t sustainable, there’s a lot of reasons why a lot of people aren’t sticking around to make sure that they maintain those implementations. But there are three main factors that I hear that lead to the short tenure.

Number one: I’m not a freaking purple unicorn. This is an interesting challenge in our space in that if you mention you work in digital analytics, it’s assumed that not only do you have a broad spectrum of skills, but that you are an expert in all of those things. You’re an amazing front-end developer, you know JavaScript and JQuery inside and out. You probably can do programmatic analytics in R and Python and write amazing SQL.

You’re an amazing analyst that can really understand data and translate it into business stories. Not only that, but you can sit with the C Suite and you can help guide the business with data. You can build amazing visualizations. You’re an optimization strategist. You can build and grow teams.

It’s simply not a reality. And when these people get hired into these roles with this expectation, it’s a recipe for burnout. I’m amazed they’re staying 1.8 years, to be honest. We need to be realistic about what it takes to build out organizations and the expectations we’re putting on digital analytics professionals.

Number two is: I want more money. As we all know, there’s a major talent shortage in this space. And if you’re not creating a company that is attracting people and keeping them happy, giving them a roadmap for years of success. Companies are willing to throw money and titles at people to get them to move. So, if you’re not happy at your job and somebody throws a senior/director title at you or more money at you, you’re going to leave.

Then finally, and this is a big one that doesn’t really get talked about—I don’t know if it’s taboo or people just don’t want to admit that it’s happening—but a lot of people get to a position where they say, “I don’t know what to do with this data.” This is shocking. We have “Analytics” in our title, yet so many people in our profession are either glorified tool administers or they’re really great implementers.

And that’s great, we need those people, but we also need other people to support it. And by that, I mean, there’s only so long you can drag out an implementation before the business starts saying, “Hey, we need insights. We need analysis.” And if you’ve hired a tool administrator or implementer to also be your analyst, you’re putting them in a really uncomfortable position. And after a year and a half, they’ve probably done all the re-implementing they can do and it’s time to get out.

So, a few things to think about on why we have such short tenure, but we have to fix that if we want our implementations to be sustainable.

Slide 5:

Number seven, and this kind of goes hand-in-hand with the previous one: hiring managers don’t know who to hire. If we were having massive turnover, we’re doing a lot of hiring. And if we don’t know how to hire, we’re creating more and more problems with the problems with the people we’re bringing into the positions. It’s compounding the problem. Part of this is that businesses still today are struggling to figure out where digital analytics should fit within the organization.

I saw this on the client side. I spent four and a half years running analytics on the client side, and over time I think I reported to four or five different bosses in four or five different organizations within the company. Organizations struggle to know where analysts should sit, so often times they get pawned off to an organization where the managers of those teams have very little analytics experience, but they’re tasked with building out the team. So, when they go to hire, not only do they struggle with the skillset and the people to hire, often times they don’t even know what position they’re trying to hire for.

So, this really needs to be fixed at the top. The direction for analytics needs to come from the top and flow down. If you’re struggling to find out where analytics fits in your organization, our ability to create anything that’s sustainable is going to be very difficult. Know the role you’re hiring for. Are you hiring an analytics manager that you’re looking to set the strategic vision for the company, be a mentor?

That’s great, hire for that role. Do you need an implementer? And you do. Hire an implementer. Hire a visionary. Hire an optimization strategist. Don’t hire an analytics manager that you think is going to do all of those things. They’re not. They’re not going to be able to do it. And even if they could do it, you probably couldn’t afford them. And even if you could afford them, there’s no way that that one person has time in the day to do all of the things needed to run a successful organization.

So, hiring managers, please ask for help. Ask for help internally. Reach out to someone in your network that you trust. Bring in consulting firms that specialize in hiring. This is a huge decision and one that is going to be critical to the long-term success of your analytics implementation, but your analytics practice as a whole. So definitely figure this one out. Understand who you’re hiring, what you’re hiring for, and hire the right person.

Slide 6:

Number six: industry gurus inventing new ways to do the same thing. Again, we’re 15-20 years into this and we’re still doing the same thing. We have experts in our space that come up with a new and better way to collect the same exact data. And not to say that that’s wrong. Is the data better? It could be. Is the way of collecting it more sustainable? It’s possible. But we have to ask ourselves, if this has been the norm for more than a decade now, is the model really working? If we’re still focused on implementation, if companies still haven’t got implementations right, is this the right path to go down?

Unfortunately, I see two major problems with this. One, existing implementations that are solid, are sustainable, are often picked apart and torn down because there’s a better way to do this, we need to start tearing this apart. And often that takes a solid, sustainable implementation and puts it on questionable foundation. Again, not to say that we shouldn’t be thinking about new and better ways, but we have to think critically. And we have to ask the question: is this right for us? Is this right for our business?

Another new trend that I’ve been seeing as analytics is hot, are new agencies entering the space and they’ve introduced what I call the “burn and churn” model. They move in, and regardless of what your existing implementation looks like, they say something like, “We need to rip this down. We need to rebuild it. We can do it in two or three months. And oh, by the way, we can do this without any help from your side.” I have yet to see this model work and I have yet to see an implementation done using this model that’s anything close to sustainable. Let’s be careful.

I’m not saying experts aren’t important, they are. They are very important and they can put you in a much better position to be successful over the long-term. So absolutely, make use of professionals, experts that are inside and outside of your organizations, but hire the right ones.

A quick example of this. I recently flipped a home because I watched a few shows on HGTV, and of course that makes me an expert at home renovation, and I quickly found out that it was way more complicated that I had realized. I was in over my head. I reached out to an expert to help, and my decision—going back to the hiring managers who don’t know how to hire—I didn’t know how to hire. I brought in an expert who wasn’t the right expert for me, he cost me a lot of money, set my budget back, set my timeline back, and it was very frustrating.

I reached out to someone who has been flipping homes for a long time and I said, “Dude, I need some help. Help me figure out how to find the right expert. I know I need the help, I want to learn, but I need someone that’s the right expert for me.” He was able to teach me a lot on how to identify the right expert to use, I found another expert who was a little bit more expensive, he got me back on track, produced beautiful results, and taught me along the way so that I came out the other side of it ahead of the game.

The experts in our industry, and there are a lot of them, absolutely, let’s make use of them because they can be very beneficial to us, but let’s be careful and make sure that we’re selecting the right experts for what we’re trying to solve.

Slide 7:

Number five: analytics in a silo. It may sound like I’m beating a dead horse here, and I am, because if you’ve been to a conference over the last 10 years, I guarantee there was a breakout session where someone was talking about, “We can’t do analytics in a silo.” And it’s true. For this slide, I want to focus on analytics being siloed from our product and our development teams. This is a leading cause of implementations that are simply not sustainable. And you can see this happening. I’m sure a lot of you have heard similar feedback: “I guess we can get analytics into this project, but it’s going to blow up our schedule,” or, “Yeah, if we really need to put analytics into this page, we can do it, but I’ve got to tell you, our budget is going to be totally off now.”

If I’m an executive in a company and I’m hearing those things, I’m upset. I’m upset because what we should be talking about is: why wasn’t analytics included as a critical part of that project from the beginning. When we were thinking about project timelines, when we were thinking about budget, why didn’t we include analytics as a core piece to that? But we’re still not. It’s often an afterthought, it’s often a bolt-on. And what happens is, you can take the most beautiful implementation and if every downstream project is a bolt-on for analytics, we scramble. We put on quick-fixes and patches and hacks to make analytics work, and over time, that can corrode even the greatest of analytics implementations.

The bottom line is that sustainable analytics cannot be achieved without buy-in and support from your in-house development team. It can’t be done. Analytics teams, if you think it can, please, stop thinking you can do this alone because you can’t. More importantly, executive teams, stop pretending that your analytics teams can do this alone. They cannot. You have got to invest in making sure that your organization as a whole has analytics prioritized as critical to your business. You have to prioritize with your project and development teams that analytics is to be included as part of the project and not as an afterthought. Again, I’m beating a dead horse here, but these things are critical if we want to create things that are going to sustain over time.

Slide 8:

Number five: rinse and repeat. And if we combine this one with the high turnover rate that we see in our industry this one gets magnified. And you can identify rinse and repeat when you hear yourself or others saying something like, “This is how I did it at my last company, so we’re going to recreate that exactly here. Because it worked there, it’s going to work here.” That isn’t to say that we shouldn’t use our past experience. Absolutely, our past experience is very important. But if we’re going to take the route of “I did it at my last company, so we’re going to do the exact same thing here,” you are going to fail.

The fact that these companies are drastically different. They sell different products. They have different target customers. They’re in different industries. Those things do matter. And they make what you’re trying to build extremely important to understand. So please take what you’ve learned at your last job, but understand the landscape of your current company and tailor your implementation to their needs, not your last company’s needs.

Slide 9:

Number three: to a hammer, everything looks like a nail. This one is one of my favorite ones. It’s human nature, we want to capitalize on our own strengths. And as analytics professionals, we obviously have our strengths. Some of us are implementers, others of us are true analysts, even others focus more on data visualization or optimization. However, where this becomes problematic is if we’re hired to run an analytics practice as a whole, our job is to set the vision for analytics, which we’ve already discussed is a broad spectrum of things, however, I’m going to play to my strengths.

If Jason is hired as an analytics manager, and my background is in implementation, there’s a high likelihood that that’s where I’m going to put my focus and not on the other areas that I may not be as comfortable in. and why is this a problem? Well, I can take over an implementation that was already good before and built to be sustainable before and I start hammering holes in it simply because that plays to my strengths of optimization, that can ruin something that was already working before.

We need to be really careful about this one. If you want an operation, ask a surgeon if you need one, they’re probably going to say, “Yes.” If you really want a new analytics implementation or to bang holes in your existing one, go out and hire a new analytics manager who has a background in implementation.

I don’t put all of the blame on analytics managers, even though a lot of the blame should be burdened by them, they have a much bigger focus than just implementation, but I put a lot of the blame on organizations and the executive leadership in these organizations. They just don’t have the willingness to invest properly in an analytics practice. Again, stop pretending you can hire one person to run everything you need to have a data-informed organization. It just does not work. Be careful of this.

If you’re hiring new analytics managers to set the vision for your company, it’s great that they have expertise in one of these areas or another, but be very careful and even more careful if this is you, be critical of always using a hammer because that’s what you have in your tool bag. Look to expand your set of skills. It’s going to make you a much more valuable manager, a much more valuable employee, and it’s really the right way to look at building sustainable implementations.

Slide 10:

Number two: lack of planning and design. We’re really getting down to the meat of the issue here. These last two issues that I’ve identified are really at the core of why so many implementations are falling into disrepair and the businesses are not getting the data that they need to properly run a data-informed organization. I get it, it is not glamorous.

I watch a lot of Adam Savage on YouTube, on his Test-it channel. That guy loves the planning and design step of projects and it’s amazing to watch how meticulous he is in the setup. He often times spend more time in planning and design than the actual build of his projects and it’s a beautiful thing to watch, if you ever watch any of his shows on YouTube. And if you have an Adam in your organization, if you have someone who really loves planning and design, please hang on to them because they are rare, but critical. It’s not only important that you invest the proper time in planning and design before you begin building your infrastructure, but it’s absolutely necessary. Without proper planning and design, even the best thought-out implementation is not going to stand the test of time.

It’s interesting, my son is a huge fan of Simcity and when he first started playing the game, I would watch him and he would get so frustrated. He had this rush to just build, build, build. He wanted to throw houses down and sports stadiums and railroad tracks, and before he knew it, he was in trouble. His town was on fire and I said, “Bud, what’s happening?” and he said, “Oh, my town is on fire and my roads are a mess. My fire trucks can’t get there to put out the fire. Now I’ve got this crying family that’s moved in and people are moving out of my high-rise apartments. It’s a mess. I’ve got to start over.” It’s all because he didn’t tackle the time to properly thing through how to design the city he built.

Just as we wouldn’t start building a neighborhood without the critical things like grating and roads and utilities right at first, the same is true of our analytics implementations. Let’s not jump right into our tag manager and start building rules and sending data to our analytics platform and building dashboards. We’ve got to stop and think and plan and design. We do a really good job of this when it comes to our data collection design, our underlying data layer strategy, but it’s so much more than that. There are so many other factors that need to be thought about before we ever put any code on the page.

We need to think about the systems and people that need to be in place to support this from across the organization. How do we get plugged into our development sprints? What’s our request for new tracking so that we don’t ruin a brilliant implementation because we don’t have any way of maintaining it? What does our QA process look like? What does our change process look like? There are so many things that we need to think through before we ever start putting code on our pages to collect data.

Slide 11:

Finally, the number one reason. And if you take nothing more away from this presentation, please take this: everyone wants to build and no one want to maintain. Maintenance really is the driving factor for things that sustain over time. Again, we can have the best planning, we can have the most well-thought-out implementation, but after it’s built and no one wants to maintain it, it’s quickly going to fall into disrepair. It’s going to crumble and it’s going to fall apart. I want everybody to commit that if you’re going to build something, you’re also committing to maintaining that.

And that’s not just initial implementations, it’s anything. If we’re adding a new marketing pixel or we’re plugging in a new marketing technology, I want you to commit that if we’re going to build something new and push it out that we’re dedicating ourselves to maintaining it. It has to be a dedication because it’s not fun. It takes a lot of time. It takes a lot of money and sometimes we just have to grind through the maintenance. But if we we’re going to build something that sustains, we have to maintain it. If you truly want to build something that is a legacy, that outlives your 1.8-year tenure at a company, please commit to maintaining everything that you build.

Slide 12:

That’s it. Hopefully there was some value here. Here is a random picture of me when I thought my music career was going to take off. But seriously, thank you. Thank you for attending. It is very humbling that you would take time out of your day to attend my session, to listen to me speak. It means a lot to me and hopefully you were able to take something of value away from this presentation so that we can collectively start working on creating things that are sustainable so we can start seeing the promise that analytics told us many, many years ago that we can run truly informed data organizations.

To be honest, I was hard on a lot of people. I called out a lot of people and roles. The fact is, there are a lot of organizations and people that are doing really, really great things. I think a lot of you who are listening to this presentation are in that group. You’re doing great things, but we as an industry do a really horrible job at giving you a voice to that we can learn from you. We need to fix that. We need to give you a voice and a platform so we can learn from you. We can learn a lot from the great things that you are doing that often go unseen.

If I can be of any help to you, if I can help to give you a voice or a platform, if you want to reach out and tell me that my presentation today was horrible and I was wrong about everything, I will make time for you. I love interacting with people in this industry. I love helping. Please reach out to me on twitter, send me an email, I will make time to talk with you and I look forward to hearing from you.

Thank you again. I appreciate you listening to my session. Enjoy the rest of the sessions today. We have a lot of great speakers, you can learn a lot. Thank you to ObservePoint for including us.

Previous Flipbook
Securing Your Analytics for a Stress-Free Black Friday and Cyber Monday
Securing Your Analytics for a Stress-Free Black Friday and Cyber Monday

Learn how you can lock down your analytics implementation in time for the coming holiday shopping season.

Next Video
Fundamentals of A/B & Multivariate Testing
Fundamentals of A/B & Multivariate Testing

Watch this introduction to the fundamentals of A/B & multivariate testing for those who were afraid to ask.