We’ve already covered two core data challenges facing most companies in my last two posts: aggregating your most valuable data, and setting your organization up to capitalize on this new well of information.
But just organizing your data and finding the right people or technology to manage it isn’t where your big data planning ends. If you stop there, you’ll just end up with a bunch of very nicely organized, important data points. And that’s not insight you can act on.
So how can you turn your data into action?
“The Meaningful Use of Big Data: Four Perspectives – Four Challenges” (PDF download) highlights different approaches to make the most of your data through a disciplined approach of interaction, to best understand and capitalize on your data.
And I couldn’t agree more. The research underscores the need for process and intent in order to make the most of your data. Without a solid plan for action, a bunch of numbers won’t help any company move forward. So even if you have the data warehouse, even if you have the technology and the data scientist, mastering big data doesn’t stop there.
Here is a three-step translated process for online marketers and retailers:
1. Develop data objectives that tightly align with key business objectives. Before diving into your pristine new data or turning to a brand new data scientist, identify how your data strategy aligns with your overall business initiatives. Think about how this data ties to the organization’s major goals.
This is important because many times individual people or departments will come up with great ideas for the data. But if those ideas aren’t closely aligned to your overall company goals, it’s a wasted opportunity. For instance, one department might want to reduce bounce rate for a certain segment of visitors. But if your overall company objective for the year is to increase average order value, that should take priority.
Ask: How does this query of our data help us reach our organizational goals? Will this query of data help capitalize on a business opportunity? Will it help us close an operational divide with our competitors?
The answers to questions like these will help you make sure deep dives into your best data aren’t a waste of time and actually impact your company’s bottom line.
2. Choose your approach to the querying process. How you approach your data will have a huge impact on what you can do with it. There are many perspectives when it comes to your data approach, but let’s take a look at two of the big ones:
- Retroactive analysis. This approach is used for looking at past customer behavior to uncover lost or missed opportunities. For instance, you might query the data to uncover the opportunities missed during a macro campaign, then determine how you can tweak that campaign to quickly and effectively change those results.
At a basic level, this approach is focused on the behavior of customer segments within your data, and it’s a rich and vast vein of analytics that can allow you to drive incremental performance in future campaigns.
Ask: What are the driving factors or attributes about the overperforming or underperforming customers that, through isolation and further scrutiny, could produce a better result on re-execution?
- Predictive/behavior modeling. This approach takes into account the isolated attributes that, across segments, prove to be potentially valuable when aggregated. Similar to multivariate testing, predictive/behavior modeling focuses on understanding individual element performance out of many elements, but you haven’t yet modeled them together as an aggregate.
For instance, you might look at your data to uncover your best conversion performers. Perhaps your data says there are a whole bunch of great performers from Michigan, a whole bunch from Facebook, and a lot who are around 25 years old, according to Census Bureau data.
Now, it’s time to take a closer look at people who are 25 who live in Michigan and come to your website from Facebook. Using that information, you’re able to create a commonality that can cut through isolated aspects and compile disparate elements to create a customer model and uncover ways to find elements you want to reproduce.
3. Make sure you have the ability to execute. Whether you decide to focus on retroactive data analysis or predictive/behavior modeling, or another approach altogether, the ability to execute a new campaign or plan with this information is absolutely critical.
A well-thought-out and disciplined process is the only way to drive data success and make a meaningful impact on your organization. Don’t get bogged down in complexity. Your technology and analytics resources should directly relate to the sophistication of your efforts. Don’t be paralyzed by the daunting amount of data and try to do too much, or wait and worry about perfecting an unattainable resource environment.
Instead, take the opportunities available, and remember that the biggest key to your success is taking action and starting on the journey. Impact starts with activity!
Data image on home page via Shutterstock.