Big Data is all the rage these days. It’s also confusing, intimidating and potentially misleading. Not so when you work with Empircal Path, a longtime MWC partner. We thought it would be helpful to ask Peter Howley, founder of Empirical Path, a few questions that may be on the minds of our CMO friends and let him riff.
Why do you get excited about working with marketing data versus other types of data? I think it’s because there’s just inherently a story behind it: we tried this, and that happened. A client puts out an ad, or an email, or changes a landing page, and their marketing data shows the results. Or could show the results: often our job as analytics consultants is to suss out that story.
Frequently with marketing data (and other types), the data is incomplete, so there’s almost a detective element to some projects: how do we estimate or proxy a missing clue–missing because someone forgot to, say, add campaign tags to a link, or put tracking code on every website page. When you’ve attributed a success to the guilty party, you’ve solved the case.
On a personal level, I also like marketing data because I kind of abandoned the world of verbiage and presentation when I got out of journalism, and I miss the puzzle of, “how do we convey this to an audience?” Now I’m helping the people who are still working those same problems in new platforms, most directly with our media and content clients but really for all of the marketers we serve.
What are the top things brands should keep in mind when analyzing marketing data? Well, probably first, the incompleteness of it, and how that doesn’t have to stop — or subvert — the analysis. If we know some data is buggy, can we still make comparisons for one buggy period vs. another, or for some useful, less-buggy subset of the audience or campaign? Struggling brands might not have invested in the infrastructure to gather the most pristine data, but there’s always a way to use what you have collected.
Second, we always emphasize putting analysis in context: instead of reporting “the number,” how does it compare to competitors and peers? Or to the category as a whole? And to your clearly defined marketing objectives (which might spark the question, “do we have any?”)?
Finally, brands should focus their analysis on data that most closely reflects their business goals. Reporting the number of social followers is a classic example: that’s a proxy for a proxy for a proxy of revenue, which in turn is just one part of the formula for profit, so how do you push further and further down that equation so that you’re reporting on something that c-levels care about? When you can’t seem to make a decision, or you tend to over-think things, then posing the choices in terms of profit can clarify quite a bit.
How do you see the roles of CMOs changing over the next few years? The CMO is like the dance partner of the agency and they’re both learning this new dance; only the CMO has to lead and do every move backwards! The stakes are just higher: an agency that loses one of a few clients is worried, but a CMO that stalls the company’s brand is done. So CMOs have to master cobbling together these services, from many providers, or a few, and with some in-house component as well. They have to get involved in the coordination and turf battles to get all of these players to deliver something on time and on message. So they’ll need more technical skill and measurement skill to evaluate all of these providers and increasingly technical tactics. But CMOs still have to sort out who, if anyone, to listen to about the big-picture stuff that every provider wants to influence, or have their own strong beliefs…and be right. Feeling kind of bad for the CMOs now…
What is the biggest challenge you face today? Finding people who can do this digital measurement stuff! Which is in one sense great, because it is a result of so much demand for what we do, especially web analytics. In a perfect world, every engagement would have not just veterans like my partner and me, but mid-level and junior staff, which is where we see the crunch.
Not only is it, “the more brains on a problem, the better.” One brain should focus on generating findings, which means others should ensure data is collected and reported right. We need talent that can deliver on the established stuff, but also play with startup tools, so we can do some value-add and spec projects, write case studies, and keep our offering cutting-edge.
What advice would you give to newcomers in the marketing industry? You gotta bring both sides of your brain. And keep developing both sides. I was lucky that publishing was hurting when I graduated college, so I stumbled into a data-driven strategy consultancy and built some quant skills to go with my appreciation for white space and, I don’t know, the semi-colon.
Know that the aesthetic and empirical can feed one another: if you want to take some creative risks and get colleagues to embrace change, you’re more likely to get buy-in if you make the up-front case with data, and reassure the doubters that you’ll be measuring the whole time, so you can pull the plug if the numbers say so…or double down!
Finally — and here’s where I sound like a crotchety old man — just because you’re new to the marketing industry doesn’t make the marketing industry new. Delivering value that makes people want to part with their money is an age-old problem; we’ve learned a lot about it…I’m not saying “be stale,” but more like, “stand on the shoulders of giants.” Build on what worked and didn’t in old-fashioned channels when opening up a new-fangled social/digital/interplanetary channel, and don’t lose sight of the value of consistency among these tactics.
Peter Howley developed Empirical Path’s data-driven approach while at strategy consultancy Bain & Company and in the MBA Program at Harvard Business School. He started the Business Research and Analysis practice at social marketing firm Noral Group International. Peter led market research and web analytics at washingtonpost.com, then headed the advertising effectiveness product line at the internet’s largest ad network. He has earned Omniture and Google Analytics certifications. Oh, and he’s really smart.