Podcast

How to build an analytics team

Callahan Agency | August 14, 2018

If you’re building an analytics team, consider yourself lucky. Too many marketers handle data, strategy, planning — the full gamut — on their own. But as you try to figure out your top priorities and the path toward success, you may be feeling more overwhelmed than lucky.

Let us help. In this podcast we identify the questions that should be shaping your analytics decisions for establishing your team. We’ll get right to the heart of:

  1. What do we want to accomplish with all of this data as a team?
  2. What is the blind spot for your organization as it relates to analytics?
  3. What do you wish you knew?
  4. How much help do I need on the technology side of analytics versus the strategy side?

Zack Pike, VP Data Strategy & Marketing Analytics at Callahan, walks through these questions and discusses the hiring process.

Listen here (or subscribe on iTunesStitcherGoogle Play, Google PodcastsPocket Casts or your favorite podcast service):

Welcome to Callahan’s Uncovering Aha! podcast. We talk about a range of topics for marketing decision-makers, with a special focus on how to uncover insights in data to drive brand strategy and inspire creativity. Featuring Jan-Eric Anderson and Zack Pike.

Jan-Eric:
Hi, I’m Jan-Eric Anderson, head of strategy at Callahan.

Zack:
And I’m Zack Pike, head of data at Callahan.

Jan-Eric:
Today on the podcast, we’ve been talking a lot about different topics around marketing analytics. Today, what I thought might be a good topic to get into is how to build a team to handle analytics. If someone’s interested in building a team, what are the things they need to be thinking about?

Jan-Eric:
So, let’s talk a little bit about that, Zack. If I’m thinking about building an analytics team for my own organization, where do I start?

Zack:
Yeah. The first place to think about is what your objective is for the team. I don’t think it’s enough just to say we’re going to do analytics. We’ve got to think about what specifically we want to accomplish with all of this data and analytics stuff that we’re going to ask this team to do something with.

Zack:
And so, I think that can be broken down the middle in to two different pieces. And this is kind of how we even do it at Callahan, it’s the business side of analytics, and the marketing efficiency side of analytics. And those are two different things.

Zack:
Business is sales, product units moving off shelf, store performance. Marketing efficiency is more focused on budget, and am I extracting all of the efficiency I can out of whatever channels I’m running media through?

Zack:
Someone might need to know how to tell if a paid search campaign is being efficient. Is a $10 cost per action the same as a $2 cost per action, and why are those things different, and should I be a two or should I be a 10?

Zack:
Those are two different core objectives, and I think once you wrap your head around what you want to accomplish, it makes the decisions on who and how much easier.

Jan-Eric:
But you’re talking about search and things like that, that’s kind of the cart behind the horse. Which gets back to, the horse is probably the business side of things. And as you think about objectives for a team, an analytics team you’re building, I feel like it’s probably routing back to what does the organization need?

Jan-Eric:
What’s the vulnerability or the blind spot for the organization as it relates to analytics, and really tasking that and thinking about, what does the organization need? What do we not currently have? What’s on the list of IWIKs, the ‘I wish I knews,’ right? The parking lot list on the tear sheet in the boardroom. I wish we knew this, I wish I knew that.

Jan-Eric:
So much of this is really going to get back to what are the blind spots of an organization? What are we trying to understand and learn? That’s really the foundation of objectives for an analytics team, or for someone who’s trying to build a team.

Zack:
Yeah. Yeah, I think a good way to do that is to start writing those questions down. So, I wish I knew X. Here’s how I would use the answer to that question. That’s a really easy place to start. And being able to articulate that is going to make the interview process easier, because you’ll be able to articulate those types of questions to your candidates, and they’ll tell you their thoughts on that. And you’ll find someone who’s really experienced who might already have some suspicions around why things are happening.

Jan-Eric:
Yeah. So, the objectives for the organization are starting around, “What do we wish we knew? What do we not know? What are our vulnerabilities? What are the things that are critical for us to know?” Even some things, some facts or some data you think you anecdotally know, but you can not specifically articulate. For example, I own a car wash business, and I know that weather impacts my business. I know that it does. But when you ask me to tell you how much does weather impact my business, or how does it impact it, it’d be really hard for me to articulate that.

Jan-Eric:
Well, when it rains, a couple days after it rains, I get a lot of business. But that lacks a lot of specificity and understanding. So, an objective in that example may be to really understand some of the external factors and how they’re impacting my business could be a possible objective.

Zack:
Right. Absolutely.

Jan-Eric:
Yeah. If step number one is really articulating the objective of the department or the function, which really routes back … is getting back and is routed in the vulnerabilities or blind spots in an organization, where do you start? Or how do you think about what types of positions you need then?

Zack:
Yeah. Yeah. Once you understand those objectives, the questions that you need answers to, it does start to bring forward the types of people that you want.

Jan-Eric:
We talk about, like a business analyst for example, if I’m going to need to understand what I’m routed in from a business standpoint, I know I need to have a business analyst. If one of my big vulnerabilities is understanding how marketing is impacting my business, I might be looking for a campaign analyst, or a channel analyst, and bringing in these types of functions. If I’m going to be looking at massive pieces of data, I need a data administrator. These are different types of roles we might need.

Zack:
Yeah. Yeah, I would start at the objectives. Once you get there, it’s almost like how we have our platform set up today. So, data coming in is a big piece of the equation that this team is going to have to deal with on a day to day basis. And sometimes, a person who is really, really good at thinking about a business, isn’t always the best at going out and collecting all the data to be able to answer those questions.

Zack:
Many times in their careers they’ve leaned on much more technically minded people to do that. And so, understanding the balance of, “Okay, how much help do I need on the technology side of this analytics program I’m getting ready to set up?” And how much help do I need on the thinking and strategy side, is a big factor.

Zack:
If you’re starting form zero, no technology in place, you’ve kind of answered some questions, you’re probably using Excel or something like that. You’re probably going to look for someone who’s either well rounded, meaning they’ve kind of got a little bit of the technology side in their skillset, and they’ve done the analysis and strategy side, which those people are extremely difficult to find, unfortunately.

Zack:
Or, you’re going to maybe lean a little further into the technology side. Hiring someone who can come in, who already has experience in the data databases, the tools that are necessary to do this stuff, have dealt with the mess before. Maybe lean on someone on that side to begin with, hoping that they will grow their skills on the strategy side as the technology comes into place.

Zack:
That’s the first real decision point, is technology versus analysis. Past that, if you’ve got the technology side covered, maybe there’s a database infrastructure already in place, maybe Tableau is already a part of your organization somewhere else. Then you can lean harder on the strategy side, the true analytics professional side that, in the terms that we typically think about it.

Zack:
Someone who’s got some experience in the tool you’re using, whether that’s Tableau, Power BI, Domo, whatever the platform is. Maybe they’ve got some experience in there, and they’ve done some reporting, they’ve done some analysis and kind of gone that route.

Jan-Eric:
Yeah. So, it’s a variety of skills really, that come in to this. You’ve got to build a house to house the data, and someone who’s got the technical expertise to be able to combine different technologies that exist, be aware of the technologies that exist, what’s available, and how to combine those technologies into the right type of stacker to give you the capabilities that you need. That’s one thing. The ability to capture, and structure, and clean up data, and make it useful. The ability to visualize that data in useful views.

Jan-Eric:
And then like you were saying, the analysis piece, you kind of see this different role start to emerge on different steps along the way of establishing a house, and then to house all the data, and bringing it in, structuring it right, and being able to analyze it and present it back to the recipients.

Zack:
Yeah. Critical that those bases are covered. And then you can start thinking about alternative skill sets. It doesn’t matter how many different ways we do this, and how many times we do this, communication is absolutely critical anytime we’re dealing with data.

Zack:
And you will find data people who are great at the technical side. They can take some crazy data set and build an amazing answer to it, but they have a really tough time communicating that to an executive team, or a marketing team, or any team who is not experienced with the data. They have a hard time with that translation from numbers to insights.

Zack:
And that’s another hard skill set to find, but it’s really important if you’ve got the technology side covered. Pushing on someone who can articulate start to finish what they did with the data and what the result was, or should be as a result of whatever recommendation they’re making is very important.

Zack:
And there’s a lot of people that would argue that that skillset is even more important than some of the technical stuff. Because if someone has already mastered the presentation skills and the communication side of analytics, that person learning the technical side is an easier transition than trying to take a technical person and teach them the communication side.

Jan-Eric:
Which is an interesting perspective, because to many people, analytics can be intimidating and seemingly difficult to understand. Where, having worked with you and other people that have the ability to, in plain speak, and I mean that as a compliment, in plain speak be able to articulate complicated ideas or analysis from analytics is invaluable. And I think it’s been critical, and at least in my experience, it’s been critical for marketing analytics to be universally understood and accepted within and organization.

Jan-Eric:
So, that’s really critical. Any other thoughts on that?

Zack:
Yeah. The last thing I want to mention is the title data scientist. That is the hottest title in the whole analytics industry. It’s all over the place, and you’ll find differing meanings of it in different areas. In some places, a data scientist is essentially what we would deem as a reporting analyst. And in some places, a data scientist is this magical human that knows all this amazing stuff, and can teach a computer how to look at pictures and tell if it’s Zack or Zack’s son in the picture.

Zack:
And so, just be careful with that need and that term. If you’re just getting started, you’re probably a good two years away from needing a true data scientist. And if you get someone coming in with that title, probe into their background on the data cleansing side, on the communication side, and what their leadership skillset is. Just make sure you’re teasing everything out.

Jan-Eric:
Yeah. So, give me an example of a question you might ask in an interview. So if I’ve got a candidate, how do I know if they’re worth it? I’ve got a candidate. And how do I know if they’re worth it? If they’re as good as they seemingly are on paper?

Zack:
Yeah. If I’m looking for someone who’s going to be focused more on business analysis, someone who’s going to help me figure out of the pricing action I took last year, how good was that, or how bad was it? I’ll ask them to tell me about an example of where they’ve done something similar in their past.

Zack:
So, tell me about your … the project you worked on that you are most proud of. And the goal with a broad question like that is, you’re looking for three key points in their response. The first one being, how they acquired the data, and what they did with it. They should be able to explain, “Hey, I pulled data from four different places. It was super messy. It was disorganized. Here’s the methods and tools I used to clean it up and get it ready so that I could actually analyze it.” That’s step one.

Zack:
Step two is what they actually did with that data. So, if this person’s going to be asked to do technical analysis, they need to be able to explain the methods that they used to get there. It should not just be an Excel chart and grid and things. It should be true algorithms and methods on the analysis side. They should be able to articulate that in a way that you understand it, probably coming from no data background, in the person that you’re hiring.

Zack:
And then, the last piece, which is the most important, and oftentimes is where you see the true nature of most of these people, is in the business result that that activity actually produced. So, they need to be able to articulate what I did with the data, how I got it, how I analyzed it. But then, what was the impact? And typically what I’m looking for are quantitative measures. $7,000,000 in revenue over the past three years after the program was implemented. 50% reduction in waste, in cost, and here are the different pieces of that.

Jan-Eric:
Of course these are just examples, but that’ll give you at least good reason to believe that they’re connected enough to the process and understanding of the information, that they’re actually tracking real tangible results, and are truly leveraging the data to the best that they can.

Zack:
Yep. And the way that they explain it, if it’s too hard to understand, then they’re probably not your person who’s going to be explaining this stuff to other people. Asking someone to go through that progression of an activity, I’ve seen to be super valuable. It identifies a lot of holes for a lot of people.

Jan-Eric:
One last topic or question related to this. If I’ve got a team in place, or I’ve hired somebody, what’s realistic to expect in the first six months, first year, first week? What should I expect? What’s reasonable to expect back from this team?

Zack:
Right. Part of that is contingent on how much technology’s already in place. Back to what I said earlier, if I’ve got a data infrastructure already in place, and I’m going to have them come in and use at least some version of that infrastructure, they’re probably going to get up to speed quite a bit faster on the actual impact to the business side of things.

Zack:
But if they’ve got to come in and set up all of the technology, one, we’ve got to make sure we hire the right person or people. But two, I need to give them time to do that. Hopefully you’ve hired someone that’s coming in with some experience of doing that before. And if you give them the autonomy to make the decisions they need to make, and they’re good enough at managing costs and the plan and everything, having a technology stack up and running in two months is not unreasonable.

Zack:
Now, different organizations run at different speeds. But from our experience at Callahan, one to two months is really what it took us to get stood up with the minimum viable product of our technology stack. But once that’s up, and if you’ve hired good people, they should be getting integrated with the teams inside the organization, their clients essentially, creating those relationships in the first couple of weeks, and be asking questions around what are the stuff that we need to have answers for?

Zack:
Again, if you’ve hired the right person, if you are looking at this person to drive, which I believe we all should be, that person should be coming to you with, “Here’s what we need to do next.” So, I’ve sat down with procurement. I’ve sat down with the finance team. I’ve sat down with all three brand marketing teams, and here are the biggest problems I’m seeing today, that I feel like data can help with. And then you two can work together on, “Okay, here’s the stuff that’s highest priority, let’s start tackling these things.”

Zack:
And then again, hopefully you can set that person loose, they go out and collect all the data, they get it all integrated, and they start producing results inside three, four months.

Jan-Eric:
Well, this has been great. Hopefully our listeners have found this valuable. Again, just to kind of recap, we talked about the key pieces on building this team are really around understanding what’s the task of the team? What does the organization need? What are the objectives of this team? And based on that, then you start to understand what are the types of roles that you need to fill, and where do you start with that?

Jan-Eric:
We’ve talked about the types of questions and areas to probe in an interview process, to make sure that you’re identifying people that have the right skillset. And then finally, touching on what are realistic timelines, and what we can expect from them, and how soon.

Zack:
Yeah.

Jan-Eric:
So, thanks for your input on this. Hopefully you all have found this beneficially useful. Thanks.

 

Thanks for listening to our Uncovering Aha Podcast. Callahan provides data savvy strategy, and inspired creativity for national consumer brands. Visit us at Callahan.Agency to learn more.