How To Make Your Marketing Analytics Work For You
A crash course in how to tame the marketing analytics beast - end up with the tidy, actionable analytics reports every marketer craves.
Everybody wants an edge. We want it in business, in our personal lives, and in our marketing, too.
And so we seek it. We crave that secret knowledge others have overlooked or aren’t willing to apply.
For marketers, a lot of that secret knowledge is in the data – the marketing analytics data we gather every day. Maybe not all of our business secrets come from data (we make “gut decisions”, too), but most of them do.
Most of our competitive advantage comes from data, too. As GE CEO Jack Welch once said,
“There are only two sources of competitive advantage: The ability to learn more about our customers faster than the competition and the ability to turn that learning into action faster than the competition.”
So it’s no wonder so many marketers strive to be data-driven. Because smart companies are data-driven. Successful companies are data-driven.
Trouble is, there’s a sea of data out there, billions of pieces of information – even in a “small” business.
The job is to take all that data and distill it down. Like a chunk of carbon being pressed into a diamond.
We want clear, reliable information that will show us how to optimize our businesses. How to delight our customers. And hopefully, even how to see opportunities before our competition does.
That’s the goal of good data.
In other words, we want to tie our analytics and reporting to our business goals. We need results, okay? All that data really doesn’t mean much if we can’t get actionable results out of it.
This is exactly what the most successful marketers are doing now. Recent research from Econsultancy and Google found that '95% of leading marketers agree that to truly matter, marketing analytics KPIs must be tied to broader business goals.’
It sounds so simple. Until you sit down with your reports… and experience the meaning of that old saying, “the devil is in the details”.
Nowhere is this more true than with analytics data. You could kill a big tree just printing out the information from your social media accounts alone, much less your website analytics, off-site content analytics, your listening station.
And that’s just marketing. Then there’s tracking the financials of your business, and staff management. Inventory, product development or asset management. And we haven’t even gotten to managing the data from a brick and mortar location… or 50 of them.
And yet, organizing all that is essential to success.
We have to bring all these pieces together. Otherwise – if we just look at our marketing data in isolation – we stand to make waste a lot of time and resources.
As you can see in the graphic below, marketing data alone is not enough to make a real difference, much less keep our businesses competitive.
I know it’s daunting, but don’t give up. And don’t tell us you’re “just a marketer, not a data scientist”, either. Data analysis is one of the core skills required for marketers now.
So you need to know this data management stuff. Fortunately, you don’t necessarily need to manage it on your own. There are plenty of tools available to help, and you’re going to need them if you want to be among the top-performing marketers (or even if you just want to keep your job).
Research from Econsultancy backs this up. They found that the thing that has the greatest impact on using data to understand the customer is “having the right technologies for data collection and analysis.”
Even having the right tools and the right people and the right skills is not enough. You’ve probably got a few great tools already (you’re here on our blog, right?).
But none of that addresses the issue of information overwhelm so many marketers have. In fact, while great technology and all the rest is helpful, what most marketers really struggle with is distilling the information they have down to actionable results.
That may just be the real art of analytics – to distill the mountain of data we have into a “just right” portion.
Just-right analytics data and reporting.
Here’s the interesting thing, though: “Just right” is going to be different for every employee.
Way too many discussions of analytics skip over this. That goal of “finding the data that will help us do our jobs better” is different for every employee at your company. What the CEO needs to see is radically different than what your social media manager needs to see.
Your data needs to reflect that. Personalized reports can help.
But that’s just the first level of complexity. There are two other key challenges:
- That what your company needs to know is slightly different than what any other company in the world needs to know (and so cookie-cutter advice about your analytics is never going to be the final solution).
- What your company needs to know will change over time. Ever heard the saying “What got you here won’t get you there”? That applies to many growing businesses. Heck, if you’re at a startup, your business purpose and your business model can pivot any time. So it’s a tall order, but your analytics reports need to be flexible enough to accommodate that.
So all those different platforms, different metrics, different ways metrics are measured across platforms, different reporting formats – it all has to come together.
That’s no small thing. But it can be done.
First, you have to identify all the different data streams and assets, and label them in a sensible way.
Master your naming conventions.
It seems like a little thing, but naming conventions are the foundation of getting your data streamlined.
This isn’t something that can be done in an afternoon. Before you start changing filenames, be sure:
- You can retrieve what all the old filenames were (and maybe even restore them if necessary)
- All the key data people in your company understand, can use, and approve of the naming conventions you pick.
- Your naming conventions are built for the long-term. Any new products, new marketing channels, new content formats all need to have space reserved for them.
Obviously, this is complicated stuff. It wouldn’t be out of the question to ask for an expert’s help.
Get your data sets to speak to each other.
Integration is the secret sauce of analytics success. No wonder the ability to integrate data sources is what marketers prize most in their technology partners.
Find your core business KPIs – based on your business goals.
So now you’ve got all your files and channels clearly named and identified. All the different data sources can talk to each other, and their reporting is consistent.
You deserve some applause! But alas, you’re not nearly done.
In fact, you’re just getting started.
Now you need to figure out which primary KPIs/metrics you want to track. These are the measurements that contribute to your primary business goals.
Say increasing loyalty is a primary business goal. So your primary KPI for that might be number of orders placed per customer. Secondary KPIs – stuff that helps to make those additional orders happen - might be:
- Return visits to your website
- Interaction with social media posts
- Leaving a review of the products they have ordered
- Referring another customer
All those events contribute to how loyal your customers are. But not everyone needs to track them. Your CEO might want to know the loyalty metric of repeat orders. They might be interested in those secondary KPIs, but it’s really your marketing manager that’s hovering over those secondary KPIs.
A marketing manager would also know the tertiary KPIs, too – those metrics and measurement that contribute to each secondary KPI. Take “return visits to your website”, for example. There are a couple of ways that can happen:
- Customer clicks through from an email
- Customer clicks on a retargeting add
- Customer responds to a print mailer
- Customer attends a live event and goes to your website as a result
You get the idea. All those metrics, measurements and ways to return to the website need to be identified, named in a way that you can use them across your data reports, and then wrangled into line, like data ducks in a row.
Now that those tertiary measurements are cleaned up, your data is layered in a rather elegant way. With primary business goals, secondary goals that support those primary goals, and then the tertiary measurements that contribute to those secondary goals.
Your business goals are your data’s mission statement.
This almost starts to sound like we’re describing a business mission statement. Those are, of course, a one to two paragraph statement that defines what your business does, and how it does it.
A good business statement that’s embraced by a company will be visible in everything you do. Down to the emails you send to your coworkers, to how you ship your products, to how you handle a customer complaint. That’s the goal of a really great business mission statement – it is like a geometry proof that you apply in every action you take.
Data can be like that too.
Once you’ve got that short-list of core business goals, and the primary KPIs that drive them, then you define your secondary and tertiary data over that.
Note that some of those secondary and tertiary measures can apply to more than one business goal metric. Like for our loyalty goal, we’ve got the orders per customer KPI. That’s supported by (among other metrics) a customer responding to a call to action at a live event.
So live events are tied to loyalty. But maybe another one of your business goals is to build engagement with your audience – even if there isn’t a sale tied to it. Email clicks are good proof of engaging with your audience. So that tertiary KPI is tied to two primary KPIs.
Mapping your analytics like this is a far better way to understand what’s going on in your business than just looking at a few disparate reports every week. It also gives meaning and purpose to every metric you manage. No one’s going to think “who cares if no one clicks this email” when you’ve so clearly mapped how every email click contributes to loyalty.
Data visualization helps tell a data story.
Sometimes, it’s not murky data that’s inhibiting better understanding. It’s the design of the data itself.
For instance, here are two views of data changing over a period of time from Adweek’s webinar, “5 Steps for Scaling Data, Analytics & Insights: Move Beyond Countless Reports and Dashboards”. This type of timeline trend gives a status of how long-term marketing channels are performing. It’s far easier to read than trying to compare four different sheets of paper from four months’ worth of reports.
Data visualization is, of course, an art unto itself. But it’s an art with very practical uses. If you’d like to get better at it, check out “The Wall Street Journal Guide to Information Graphics: The Dos and Don'ts of Presenting Data, Facts, and Figures” or “Storytelling with Data: A Data Visualization Guide for Business Professionals”.
They’ll help you break out beyond pie charts and line graphs. There are many more elegant ways to tell a data story – and to reveal where a business needs to focus its efforts.
How you interpret the data matters, too.
Every marketer needs to understand the concept of “confirmation bias”. It’s the idea that we prefer information that supports our worldview, and that we dismiss information that challenges it.
All humans do this to some extent. Most of the time we’re not even aware of it. But who hasn’t looked at an analytics report hoping to see proof of something? Proof that one campaign beat another… proof that one content format beats another…. proof that a particular customer service policy isn’t working?
Whatever it is that you want to prove or disprove, be aware of your bias. Sadly, we humans have a long and glorious history of taking books, information – any kind of data – and bending it to our preconceived notions about things. Often to disastrous effect.
It takes courage to challenge our confirmation biases. Hey – after you’ve worked three months on a product launch, who doesn’t want the data to show a success? But let’s face it: Confirmation bias is a very subtle way we lie to ourselves.
So be careful how you interpret your data, even after it’s all nicely organized. Try having teammates play a friendly game of devil’s advocate with you, so you can challenge each other’s assumptions.
Before you make a silly mistake and assume correlation equals causation. Sometimes data just looks like it gives meaning.
It’s a data-driven world, especially for marketers. And that’s actually a good thing. Because most of us actually like data, and we’ve got a ton of tools and disciplines to help us to gather and organize it.
But there’s no way around it: Data management is a big job. It requires serious, deep, long-term thinking. It also demands technical skill in order to produce clean, reliable data in the first place.
Which is why data analysis skills are in such demand now. And while it might well help your career to develop your own data skills, it is possible to do enough “continuing ed” to still make sense of all your analytics data, and to find opportunities in it.
We just need to treat this like the skill it is, and understand that analytics and data management needs attention. Just like you have a system to keep your company’s finances in order, you need a system to keep your data in line, too.