Doing Data Like a Boss

It’s late January. The resolute go-getters from a couple weeks ago are starting to thin out of the gym already. You may be noticing that your career aspirations for the year are beginning to wane in the face of the daily fire drills.

You’re not alone.

For a lot of us, all those good intentions are starting to turn into just that — good intentions.

Here’s a little mid-week pick me up to help you reclaim that fire you had back on January 2nd. You need an action plan. If being a data innovator in your organization was on your agenda, pull out your to do list, friend, and start planning your next steps.

Step 1: Define the metrics that matter most.

Data for data’s sake starts to feel like an academic exercise. Don’t go down the winding path of following interesting breadcrumb trails without tethering yourself to your key goals. What’s more important this year — increasing acquisition or growing purchase frequency with repeat customers? Plug into the yearly corporate goals and pick one to chase down, not two or three.   What are the yearly targets and how are you trending now? Where is improvement needed the most? Invest where there’s the most urgent need.

Step 2: Choose one channel and understand what’s currently driving strategy.

Everyone is talking omnichannel marketing. It’s a great destination, but not a good place to start on your data journey. Instead, start small and pick a single channel where you have the most influence and where there’s the greatest opportunity to create change.  

Next, identify the key data inputs that drive that channel strategy.

Say you’re looking at building acquisition through social ads. What drives the segmentation strategy? Is it purely demographics and Facebook look-alike modeling or are you using purchase data and cohort analyses to find commonalities among your most profitable customers? How much tolerance do you have for customization of messaging and what dictates the message that different segments see? If you’re tackling growing repeat purchase through email, figure out how those lists and messages are built. Are large lists getting blasted on a regular basis or are website behavioral data and purchase trends informing lists and messaging?

Step 3: Define benchmarks, look for gaps and pilot some experiments.

First, define what good, great and lousy look-alike for each channel. What are the metrics that matter in your organization and what do your historical trends show? How does that look against industry benchmarks? The DMA Industry Benchmark Report is a good resource and offer product industry reports specialized for each channel. For example, Nanigan’s quarterly Global Facebook Advertising Benchmark report is a thorough deep dive.

Now that you know inputs and success benchmarks, step away from your desk and go for a walk.

Forget how things are done today and think like your consumer.  

What is data that might help you better identify them or predict their propensity to engage? Do you have partner data that you’re not using right now?

Say you’re a sports team and you play in a stadium that offers a wifi experience. Can you mine that wifi data to find the people who visited merchandise areas and target them with future email merchandise offerings?

Can you look for the stand-out social brand affinities of your customers to find a cheaper audience to target on social than what you’re currently buying? You know your customers best. Walk in their shoes and look for clues that might be in the data you or your partners have.

Pilot a program where this data is incorporated and look for uplift. Look for simple cost effective ways to incorporate data from new sources. You don’t need a data warehouse or a DMP to run some simple tests. Go manual, scrappy and pragmatic (and of course, Umbel can help here).

Step 4: Quantify the results and share your learnings.

Share your results among your team and experiment, iterate and try new approaches until you start to see the uplift you want. Brainstorm cross functionally about what pockets of data might be useful even though it’s under utilized.

Hold lunch and learns where you share your pilots and start a conversation in your organization.  Start breaking down the silos by sharing information and new ideas will start to generate.

Step 5: Start looking to incorporate another channel.

Are the insights that you’re gaining throughout your experiments useful to other channels? For example, are the people who interacted with merchandise at more two or more games last year and clicked through to your emails also a good target for season tickets next year through social or display? How can you find a way to extend success throughout the customer journey across channels? Start talking with other groups in your organization.

Step 6:  Make the whole organization smarter.

Can you mine customer insight from your pilots? Share them. Get on your soapbox. Better yet, look at a way to codify them for everyone to use. Can you update your CRM or ESP with the under utilized data? How can you inform the models that the entire company uses.

Good luck, data boss. This is a long journey. We’re trekking it with our clients everyday and would love to share stories from the road as you make your way.