The 3 Essential Truths of Marketing Analytics

James Hart Alight Insights, Analytics for Marketers

Fun Fact: In 2011, there were “only” 150 marketing technology solutions available in the marketplace.

Last year? That number had soared past 3,500. Which suggests that many marketers are drowning in the data generated by such a large number tools and systems, and still hungry for a solution that helps them make sense of it all.

Alight Analytics — and our solution, ChannelMix, the first media data aggregation engine built for marketers — have been active in this space for a solid decade, long before the vast majority of players today.

In that time, we’ve learned three essential truths. They make the difference between building a data operation that drives your marketing — and struggling along with one that devours your workweek, wastes your budget and generally blights the earth.

Truth 1 – Marketing doesn’t have a dashboard problem. We have a data problem.

5Of those 3,500 tools, there’s no shortage of tools that will generate marketing data. The majority of them, however, aren’t reporting options. They aren’t built to help you communicate how your campaigns and messages are performing.

Even solutions that are supposed to handle reporting will fail because they fundamentally misunderstand how reporting works. We believe it should be a two-step process: data aggregation first, then data utilization (visualization).

Visualization is essential. Industry-leading solutions like Tableau Software are worth every penny. Nothing illustrates how your marketing is performing like a well-built dashboard.

Before you can build those visualizations, though, you need to aggregate all your media data from tens or hundreds of channels. You need to screen for errors or outliers that distort your results. And you need to make sure your data is in a format that’s ready for analysis.

Too many vendors — especially end-to-end tools that promise to do everything — sweat the visualization while failing to devote anywhere near the resources they should to consolidating and perfecting the data.

And that’s why their users end up with gorgeous dashboards populated with inaccurate numbers. Or those users are forced to waste hours sniffing out errors and manually compiling data.

We take data seriously because it’s part of our origin story.

When Alight Analytics opened for business in 2007, we didn’t have a technology solution. We were simply consultants, helping brands and agencies interpret their marketing analytics.

And we ran face-first into a huge problem. To compile reports every month for our clients, we spent hours and hours manually downloading marketing data from dozens of media channels. Then we got to cut and paste those numbers into too many spreadsheets to count.

We called it the Data Death March. It consumed hours and hours of our lives. There was just too much data to manage.

The Data Death March drove us to create ChannelMix, which automatically aggregates marketing data from pretty much any channel you can imagine. Instead of being trapped in silos, that information is stored in one place, creating a single source of truth that’s easily accessible by multiple users in an agency or enterprise.

Truth 2 – We must rethink our analytics approach to drive value through our most important asset: people.

One of our early mistakes — one that others keep making — is that we tried to hire unicorns.

That is, we tried to hire people who could do everything: write code, wrangle data, understand marketing tactics, have a strategic discussion with clients, etc.

This is a bad idea because unicorns are nearly impossible to find or replace. More often, you’ll end up hiring someone who excels at one half of the equation and limps along doing the other part of the job.

We realized we needed to better align our resources. That’s why we hire marketing people to do marketing, and we hire data people to do data. Not only are they happier, it’s also easier to find more of them if you need to add staff or fill an open position.

Right now in the marketplace, there are several end-to-end analytics solutions that bill themselves as being entirely self-service — something that any marketer can use straight off the shelf.

But they aren’t, not really.

Marketers sign expensive, yearlong contracts for these tools. Then they discover they have to know how to write SQL statements in order to generate the answers they need. These poor souls either learn to code, or they get to stage a unicorn hunt.

A better solution for marketers is to use people who excel at the job you need to do, and give them tools customized to that job.

With ChannelMix, for example, you can hire our team of data and technology experts to manage your media data for you. We’ll hand you a collection of consolidated, perfected data. Then your marketing pros can use Tableau’s best-in-class solution to create illuminating dashboards.

Truth 3 – Reporting tells you what happened. Analytics tells you what to do.

A funny thing happens if you acknowledge Truth 1 and Truth 2: Your reporting process (and your life) will stop being quite so miserable.

In fact, you’re going to unleash the awesome, untapped potential of your marketing data and your people.

Because your media data has already been perfected, you aren’t devoting hours and hours to collection and formatting.

Now your marketing experts can skip the rote reporting (“Our website had X visits last month”). They can spend more time putting those numbers in context and telling performance stories (“We had X visits, and that led to 20 percent more conversions — hey, let’s do more of that!”).

As your team builds its capacity for this kind of analysis, all kinds of opportunities will open up. You’ll be able to ask better questions and make smarter decisions. What if you segment your marketing results, region by region, so you could develop a media mix that performs best for each individual area?

As a result, your organization’s performance is going to evolve, and your revenues are going to reflect that. And that’s truly the point of marketing analytics.