Marketing as a Data Science
Hello world! This is my first blog post, so an introduction is customary. I have worked in the digital marketing, social media, branding and analytics consulting space for product as well as services businesses. As a result, I have had the rare opportunity as a digital marketer to also learn from and advise other businesses on their marketing strategy and analytics.
But as they say, more experience can lead to either more enlightenment or more exasperation. (Don’t ask me who said that, you know who just made it up.)
Truth be told, most of us marketers do not really value our data as much as we claim we do. We might have “Analytics” and “ROI” on our strategy slides and also a painfully constructed data warehouse and reports. But is that all? Is marketing really data-driven or is data still treated as important but used mostly post-mortem? Traditionally data has been used to answer questions like:
- Are we doing enough?
- Are we doing things right?
- How can we improve?
We ran a little promotion in Kentucky retail stores – did it help us go past competition? Should we have gone for better placement as opposed to the price discount? Note the past tense in all these questions. This method of using data and analysis to reach to useful business conclusions is hardly enough in the digital era. Here are some questions I’d love to answer based on data. But how?
- If I invest $5000 on this Facebook app, will it get me 10,000 active and engaged Likes from my target audience?
- If I replace the text content on this particular webpage with a video, how will it affect the average time people spend on the website and in turn will it improve awareness about our products?
- If I move a part of my email marketing budget to banner ads, will it result in revenue improvement? How much should I move?
Sounds much like a Kipling poem, doesn’t it? But the ability to have answers to such questions, in my opinion, would be the epitome of data-driven marketing — the ability to infuse more and more mathematics and logic into a practice largely based on hunches and creative skills. At its most mature form, digital marketing should be a science where one can predict target appropriation, awareness generation and have a constant tab on channel returns. We’d then have very little scope to go wrong then, wouldn’t we?
As a result, my maturity model for data driven marketing has only three levels:
Level 1: Data What?
Here’s the digital marketing practice that represents quintessentially everything wrong in terms of using data – not knowing where the right data is, unable to integrate it, investing time, money and effort with not much method, proof or reason. They haven’t figured out most of the P’s of data driven marketing as listed here.
Level 2: Digging for Gold
The marketing team that sits on a neat and healthy stockpile of (non-explosive) data. They have what they need, know where to go, but are not answering the right questions yet. Campaign analysis, channel performance, web analytics are all set up and handy but not used to answer marketing questions and impact the future course of business. They are yet to realize which is more important for them – more data or more analysis?
Level 3: Excel before Outlook
The “epitome”, as I’ve already called such an environment, leads the way and rides every wave be it Big Data, predictive marketing or Data Activation (I hope you’ve read David Karnstedt’s blog Data Activation & the Future of Digital Marketing). In these teams, the digital marketer is also a data scientist who opens up his Excel worksheet before sending emails in the morning. This team can answer any of the following questions at any time:
- What does or target audience think of us or say about us?
- What kind of information are my users looking for?
- How many leads am I able to generate per $ spent on a channel?
- What is the quality of lead per channel? Quality being assessed by something like (revenue + pipeline)$ per lead
I like to think the future of digital marketing lies in the intelligent use of data and the ability to inspire desired digital behaviour based on an unparalleled understanding of content, channels and the target audience. This maturity model might not be a jump-out-of-the-bath kind of discovery, but it ought to help us assess where we are and what we need to do to be the best.
What am I going to do this week? Analyze differences in user behavior on the website across North America and Europe. Hold on till my next blog!
Parth Mukherjee is a global product marketing manager for Adobe’s technical communication portfolio of products. Parth loves technology and also the use of technology in digital marketing. On this blog, he shares his trials, tribulations, successes, and what he reads, hears or experiences in the world of digital marketing. (Spolier alert: Parth uses the Adobe Digital Marketing Suite and loves it.)