Rocket Fuel : Aug 22, 2017 | Data Pools or Puddles? 3 Steps to Improve Your Data Assets

By William Sears, VP, Enterprise Strategy

Data has always been the fuel that helps make good decisions. In the 20 years of my time as a marketer, researcher, and analyst I've seen many industries move from a focus on collecting more data, to using more data. There was a time I stood in a parking lot doing intercept and vehicle origin surveys to augment knowledge and inform a multi-million dollar decision of what resort or attraction to open next. Today we get hundreds of thousands of these observations from internet behavioral observations and rely on technology that translates video into data. Big Data is old news, most organizations are drowning in their data, even the terms used - Data Oceans, Data Lakes - portray it as a resource we are swimming in. But how much of it are we really using? What is the value of all this data? Most importantly where is the business changing benefit? That is the challenge we face - being good stewards of our data assets.

Here are three solid steps for all organizations to evaluate their Data Assets:

Step 1: Collect the Right Data

Data, data everywhere, but not the data you need? Might surprise you to hear my first suggestion is to collect more data. Most of the organizations I work with today are not, for example, actively collecting data from their mobile apps.

Start by taking a look at your customer journey. Are you collecting data at key points? Traditionally this has been buckets like landers, engagers, converters, but with the web of ways consumers have to engage today we need to take a closer look at our collection coverage. Some questions to ask:

• Have you worked validation points into the process? Consider creating broader buckets of actions that define those traditional engagement signals, as we all know there is more than one path to the checkout page.
• Does your flow enable data capture and set you up to make better decisions? Are you asking for too much too soon from your customers?
• Where and how do you ask customers to log-in?

Minor tweaks to your customer engagement process can net you the data you need, and enable use of the data you have. Data Oceans are great, but finding the glass you need often finds it only half-full.

Step 2: Connect that Data

A lot of organizations have data puddles - numerous independent data sources separated across the enterprise by ownership, responsibility, budget, or brand lines. Connecting your data will make it more powerful, and will provide a much richer insight into customer value to the entire enterprise. Your best customers cross multiple brands or services. They bought a dishwasher using your financing service and also shopped at your store using your partner credit card. Connecting these data signals is critical to recognizing and building customer value.

This can seem like a daunting, technical challenge, but the solution can be as simple as a universal customer ID. Best Practice is to use something that is not personally identifying outside of your network. Skip email address and instead think Frequent Flier Mile number. The second most common barrier is organizational. The good news is organizational realignment is not required. Technology can funnel your puddles into pools, lakes, and the ocean to create more value for all data users without requiring a change in ownership, responsibility or org structure.

Step 3: Act and Learn from the Data

This is all about applying your data and building your knowledge. Either one individually is OK, but just knowing is not going to help drive business, and learning should be an always-on endeavor. This is another spot where technology has created pockets of brilliance. Maybe on your website you are making great decisions, observing action, responding with customized experiences - brilliant; but, is that enabling your next email? In reality decisions are still being made by independent solutions so customer experiences are not connected.

We've all been there as a customer. Put something in your cart at home on your desktop, arrive at the actual store, fire up the mobile site… and your cart is empty. The key here is to create a master decision engine, where decisions are made, results are observed, and the next decision gets smarter based on the outcome. Provided you've already nailed collecting and connecting, you have the data ready and the customer experience mapped out. Good news is you do not need to rip and replace technology, you just need a central engine that can fire off decisions to any and all action points - your website server, your email partner, your call center engagement screen.

The best practice here is to let those solutions still 'optimize' at their endpoint, but provide them the centralized directions to get started in a smarter and more connected place. The truth is decision engines are often purpose built, so let them focus on their pocket of brilliance, but give them the raw material head-start, or guiderails to work within. This provides the coordination necessary to always be acting, learning, and optimizing toward your business goals.

Finally, it's important to properly define the value of this data centric process. Some would read what has been outlined here as a lot of work, especially when I can just buy data. Instead I encourage you to consider the cost of not collecting and owning this data. It's true that you are unlikely to create a data source that the likes of Google would offer to buy from you. Google already knows it, right? Maybe. But the value is if you collect and own this data, directly associate it to your brand customers - you will not have to pay anyone for it in the future. You will get it faster, more accurately, and yes - cheaper. Plus, you stand a chance of knowing your customers better than your competitors do. No more knife fight in a phone booth because someone is 'in market' for your brand. You knew that data point three weeks ago because you were collecting, connecting, acting on your data, and have already engaged them with your brand.