It still surprises me how many companies underutilize the customer data they have. I suspect even if you polled companies who are “doing great” at leveraging data, they would say they could do better. Even the big corporations who are amazing at leveraging data miss a trick every now and then.
Here’s an example: For one client, we collect rebate submissions, which means we see somewhere between 150,000 and 500,000 customer activations through our system. We know the product lifecycle; and, on paper, retargeting those customers when the product typically needs replaced with a customized offer to ensure they purchase from our client’s brand again *seems* like an obvious approach. Yet here we are, over five years later and we’ve yet to maximize that data, despite repeated attempts.
Good customer data is data we can act on, turning it into a behavior change or an incremental purchase—or (ideally) both.
This means using the data to inform your engagement strategy, which motivates your audience to do something different or repeat desired behaviors. But how do you know what is good data? In short: data that is clean and complete.
But if that’s not clear enough, continue reading below for examples (good and bad), uses and key advice for each type of data.
Data Type: Clean Data (Best)
Examples of What Can Provide Clean Data
- Stores with universal point of sale (POS) data
- Manufacturers who get complete sellout from their distribution or stores
- Distributors and stores who track end user purchases
- National account info
- Well-conducted surveys
How Clean Data Can Be Used: Clean data can be quickly segmented to target a variety of audiences with a variety of motivators specific to their segment.
Key Advice: Before you go huge, leverage the data to test and learn. Fail fast, pivot and focus on what drives revenue and ROI.
Data Type: Usable Data (Good)
Examples of What Can Provide Usable Data
- Stores with widely available POS data
- Manufacturers who get some end-user data
- Manufacturers who get some sellout from their distributors
- Claim or rebate programs
- Warranty registration
How Usable Data Can Be Used: This data type allows you the ability to segment and target groups with multiple motivators. It also provides the option for an impromptu approach with some personalization.
Key Advice: Leverage test and learn approaches like above, and work on ways you can enhance the data to make it cleaner.
Data Type: Inconsistent Data (Poor)
Examples of What Can Provide Inconsistent Data
- Inconsistent POS data at any level
- Manufacturers with antiquated reporting structures
- Franchise scenarios where the agreement doesn’t require transactional data be shared with the franchise (unbelievable; this is more common than you think)
- Independent channel reporting
- Channel claim programs
- Disparate data storage at any level
- Poorly conducted surveys
How Inconsistent Data Can Be Used: With a little analysis, even inconsistent data can provide the ability to do some partial segmentation and targeting.
Key Advice: When working with this data, balance the time it takes to clean it up with the ROI. Aim for big targets first to justify further efforts.
Data Type: Unusable Data (Worthless)
Example of What Can Provide Unusable Data
- Anything handwritten
- PDF or handwritten reports from the field
- PDF or handwritten reports from manufacturers
- Text box surveys to large groups without a plan to consume and act
- Poorly structured data
- Incomplete records (e.g., address but no city/zip)
- Out-of-date data that no longer applies to your current products or services
How Unusable Data Can Be Used: Great for testing out your new shredder. Otherwise, dump it because it’s time to reevaluate what you’re collecting and the form it takes.
Key Advice: Walk away until your systems and collection methods are improved. The cost to capture and clean this type of data is probably more expensive than investing in better collection and storage means.
Now that I’ve brought you up to speed on what I mean by data quality, it’s time to think about how you leverage the data you’ve collected.
Whether you are new to the game of leveraging data, or not yet seeing the benefits and success you feel you should with the data you’ve collected, start with how you would address the following questions before jumping into large-scale data utilization.
1. What Do You Want to Achieve With This Data? (Tip: Start With the End in Mind.)
If you know what you want to accomplish, you can identify what data you need in order to pull off your plan. Nothing is more disheartening than getting to the end of a project and realizing you didn’t actually capture the right information in order to do the things you want to do next. When that happens, you either have to go back and ask more questions of the audience (which they’ll hate), run an impromptu approach lacking strategy (which will discourage you and your analysts), or do better next time—none of which are ideal outcomes. Also, when I say think through what you want to achieve, I don’t mean “grow sales” or “change behavior.” I mean, specifically, how do you intend to leverage individual data points or responses to target customers? For example, “We will ask people who enter our sweepstakes what their occupation is so that we can target them based on their response with other products and services across the enterprise.”
2. Do You Currently Have Access to the Data You Need or Want?
It’s okay if you don’t. If you are new to leveraging data, there are ways to capture sales and behavioral data through strategic agreements with your channel, claims programs, surveys and market research. Knowing you don’t currently possess what you need is important to determine up front so you can make a plan to collect it through one of the means listed above.
3. What Will It Take to Get the Data You Need? Is It Even Possible With Your Current Go-To-Market Arrangement?
This depends a lot on where you are in the channel (manufacturer/distributor/store/mixed) and who you want to target (B2B, B2C, etc.). If you are a manufacturer, you may not have great end-user data, but that’s okay. You can put things in place to help you collect usable data. If you are a distributor, you might have better access, especially if you also have owned stores on a universal POS. But keep in mind: disparity exacerbates cost. That doesn’t mean you should abandon the project. Just know that the more difficult it is to collect and aggregate good data, the more it will (likely) cost to get a collection system in place.
4. Do You Have the Right Partners?
A good market research firm and a good incentive partner can really help expedite data collection and validation of your data. It’s even better if one agency offers both services. If your data is inconsistent or lacking value, there are tools and companies who can help glean usable information but you must decide what’s best for your company. Is it better to pay someone to clean up data, or invest in better systems allowing better collection and more flexibility? Or are both important? A good partner can help you identify the financial risks and benefits for each and arm you with facts to make a confident decision.
5. What Is Your Solution to Disparate CRM Systems?
Believe it or not, I can name at least five Fortune 500 companies who have disparate CRMs for multiple business units targeting the same customers with similar products. Just amazing!
You don’t have to have your CRMs completely sorted, but think through a plan to get there, especially as new state and federal privacy laws take effect in the next few years. What are you going to do if a customer tells two of your business units they can be contacted, but tells a third unit that they can’t be contacted? What is your enterprise solution to adhere to these new laws?
After you’ve addressed these questions it’s time to put planning into action. When you’re ready to start leveraging your data, keep the following few tips in mind:
1. Put the Right Team on the Project
The right team on the project is the team who is going to embrace quick failure. That’s right, failure. Companies who are experts at leveraging data can run multiple tests and learn scenarios in a month, a day or even an hour. Granted, they’ve worked hard checking all the boxes to get to this point, but you need a team to lead this who understands failure is simply one step closer to the magic bullet of success.
2. Arm the Team With the Right Tools
If you need to dig a basement, you aren’t going to have a crew of people use garden trowels. You’re going to call in a backhoe and dig it in a matter of hours. The same thing goes with leveraging data. There are dozens of tools and partners you can use to help manipulate and extract the key points from your data set, and your investment in the right tools should more than offset the labor you would spend trying to execute big data plays with low-end tools and manual processes. If you have thought through how you want to use data and have a clear understanding of the cleanliness of your current data, vetting a tool (anything from a better version of Excel to AI-powered software) or partner should be relatively easy.
3. Create Realistic Benchmarks (Tip: Don’t Be Afraid to Ease Into This for the First Few Tests)
Leveraging data can seem insurmountable, especially if you have any of the hurdles above. Don’t be afraid. Start with the cleanest data you have, run a few tests, and see what happens. At first, you’re probably going to be below industry average for open rates and conversions. But, if you stick with it, you and your partners will figure out what works for your audience, and then you’ll be off to the races, the races being where you can tie a sale to every marketing initiative instead of telling the story in terms of click-through rates.
These aren’t hard and fast rules you have to follow. These are just some of the learnings I’ve had through my career. Frankly, many are concepts anyone should consider before any project.
And for those crunched for time, here's the TL;DR version: Get a solid understanding of what kind of data you have and how you want to use it before you jump into customer-facing activation. Once you get that sorted, start small and move quickly to eliminate what doesn’t drive results so you can invest more into what does work.