Data, Predictive Analytics & Mama's Lil Pillz

So, my mom, Maria, had three strokes last year. Yeah, that’s right, three. Last year, she was a regular at the emergency room like it was free burger night, every night. The reason? Her brain was doing its best impression of a T-pose glitch. She’s now in a daily wrestling match with her new reality, which involves a pharmacy's worth of medications. And by “pharmacy,” I mean she’s on enough meds to stock a small Walgreens. A quick story about Mama's Lil Pillz:

My sister Susan, who clearly read every book in the Nancy Drew series, went all stealth mode and discovered Maria had been repurposing an old Tylenol bottle, filled it with… not Tylenol. So every time she thought she was popping a Tylenol for her hip pain, she was actually taking a blood pressure pill. Fun, right? And by “fun” I mean “terrifying.” Early indicator of a problem.

Then I moved in. Cue my daily observations of Mama’s lil pillz routine. Picture this: she’d carefully take her pills out of the pill dispenser, arrange them on a napkin next to her coffee, and spend the next hour picking through them like a kid going through their Halloween candy. Occasionally, I'd find a stray pill that had rolled off the napkin and onto the floor under the table. I would carefully pick it up, blow it off (hey, I raised a couple of kids, this actually works), and pray it wasn't the one that keeps her happy, or worse, the one that keeps the blood flowing. Dashboard indicator warning lights flashing.

The latest episode in our medication sitcom involved Maria sleeping all day and saying she felt just horrible. I have no judgment about camping in your bed all day if that feels like the right thing to do, no judgment here, but obviously I was concerned about her feeling horrible. Later that evening I discovered that even though it was only Tuesday, her Wednesday and Thursday pill slots were empty. I’m guessing she lost track of the days—retired life… it happens, am I right lucky retired friends? Attention all military personnel, we are now at Defcon 1.

Now, if this were a business, these behaviors and data points would scream, “Your processes are failing! Code blue! Change needed, tout suite!” I would throw on my business process management hat, rally the stakeholders, review the situation, and together we would create a new process for predictable, improved outcomes.

Gut Feelings, Anecdotes & Tribal Knowledge

Back in the day (meaning pre-data analysis), marketers like me relied on a mix of small client sourced data samples, anecdotes from sales and account reps and a healthy dose of 'gut feeling' to guide our Go To Market strategies. To design an effective campaign we were always ferreting around the organization searching for those individuals who were the keepers of information; they had the tribal knowledge and answers to key questions:

• Why customers bought from us, and what made us unique.

• What features do our customers love, and why?

• What would make their jobs easier, more efficient and less frustrating, and why?

• What will save their companies time and money, and why?

• What are our win/loss reports telling us about our blind spots?

• What are our competitors' sales teams doing and how are we doing it differently, and why?

• What is the customer journey after we drop a red-hot lead into a trial?

• How do we shine a light on our expertise, differentiate ourselves and position ourselves as THE experts?

• What knowledge is our customer hungry for and what is the best way to deliver the answers to them, and why?

Yes, we’re like an annoying 3-year old with all of our ‘why’s. But the point is that this information was all over the organization and mostly undocumented. It took valuable time to get this all sussed out and organized.

Enter AI, Stage Left

When you dump all of the operational data (CRM, ERP, finance, customer service tickets, win/loss reports, online comments, client scorecards, reviews, you name it) into a big ol’ data warehouse you are then able to create a comprehensive view into a report that incorporates all of the information about customer behaviors. (Can I get an amen and hallelujah?) Theoretically this is when we ask smart questions that help us better understand how to run faster and jump higher. We then thin slice the data into reports and analytics, so that we get a view into the past in order to predict the future. We are now able to tell the right story to the right person at the right time in the right way. It’s like magic, but with data visualization.

Using a content creation framework in ChatGPT

But even if you don’t have a data analytics team, you can still benefit from AI. Here is an example of a ChatGPT prompt that, when answered, helps to tell your story in a new way:

"Using the 'Storytelling' framework, please write a marketing campaign outline that tells a compelling story about how our [product/service] helped a [customer] overcome their [problem]. Use vivid details and emotional language to create a connection with the reader and showcase the effectiveness of our product."

Because it’s a language learning model, the more information we put in over time, the higher the quality of the output over time using your brands tone and voice. Think of it, every piece of content that your customers and prospects are exposed to are now written in a singular brand voice. Brilliant!

Association Rule Mining

Here’s a fun fact: data shows that people who go to the store and buy bread and butter often times also buy milk. French toast for dinner, anyone? This insight can guide product placement in a grocery store. Do you put these items on the shelf near each other for customer convenience or far apart to tempt customers into buying Newman-O’s, Haagen Dazs, and Fritos along the way? (Okay, that bit is about me… please put in the comment below your grocery store vices). Embedded in this answer is a strategy to increase sales by having more items in the shopping cart then they came in for. And what works for a physical buying experience is what works for a digital one. That's the association rule.

Customer journey mapping is a great example of how to apply the association rule to drive upselling and cross selling. Finance data tells us who’s buying what and when. CRM data shows engagement with our content and where they are on the buying journey. Mash it all together and you can now cluster the data to better understand the personas of your buyer, and then you're able to use AI to find more of those people. Using AI, Marketing then designs campaigns to move customers along a prescribed path designed to increase revenue by acquiring new customers and by extending the Lifetime Value (LTV) of an existing customer. Can you say ka-ching?

A Call to Action for Managers

Managers today need to not only have strong vertical and domain expertise but they also need to be data savvy so that they can be asking better questions to yield better answers from their data analytics. Their ability to ask the right questions allows the analytics team to pull the right source data, review it, throw it into the salad spinner and spit out a unified picture that tells a cohesive, data-driven story. But it all starts with managers knowing what questions to ask that, when answered, helps them make informed decisions faster.

So, while I didn’t need big data, visualization charts, or fancy analytics to crack Maria’s pill-popping caper or to create a fail-safe process for her, it’s a prime example of how a little Sherlock Holmes-style observation, asking smart questions, connecting the dots, assessing the Defcon level, and lovingly instituting a new pill management system is basically the same as tackling business problems with a data-centric approach. Just swap the pills for KPIs, and voila, you’re a corporate hero!

Now if you'll excuse me, I've got a pivot table to create for my parents doctors appointments... or a pill to find on the carpet.