If you listen to marketing community buzz, you may get the impression that big data is the magic key that unlocks the industry's greatest mystery -- understanding our customers. If only it were that easy. The truth is that, even with better access to data, interpreting customer data is still challenging for many marketers.
Since data is only valuable when we know how to use it, the brands that extract the most value are the ones that prioritize testing over theorizing and results over opinions. The most successful organizations take a pragmatic approach to collecting customer data, and know how to turn information into action.
The test you ran for a population of 1,000 users may deliver completely different results with 10,000 users -- or 1,000,000. There are many ways to estimate the right scale for your test, but here's an easy formula for a population of any size that gives you a 95 percent confidence level with standard deviation of 0.5 and a plus or minus 5 percent margin of error:
sample = population / (1 + (population*0.0025))
For example, if we were looking at a population of 10,000, we'd aim for a sample of 385 individuals:
sample = 10,000 / (1 + (10,000*0.0025)) = 384.6
If your sample volume is too low and you can't easily reach the scale you need, just extend the amount of time allocated to testing. Yes, it will take longer, but the improvements in accuracy will be invaluable.
Get buy-in from everyone at the C-level
Many C-suite leaders see the strategic value of marketing data, but some are more open to investing in it than others. How much value they assign to marketing data depends on their previous experience and current priorities. To get the whole C-suite excited about your programs, look at data from their perspective. Show them how marketing data can support the programs that matter to them and drive the business goals they care about most.
It's all about market advantage
To get out in front of your competition, start early and don't be afraid to make mistakes. Allocate a testing budget so that experimenting won't affect your bottom line, then test and test and test again. By building testing into your budget, you'll enable your team to avoid risk aversion and cut through the red-tape that might hold up innovation and experimentation. Scour industry journals for trends and emerging platforms to fuel your experiments. Hold informational meetings to get new perspectives from other stakeholders. Be flexible, and iterate early and often.
Data has a place in every decision
... but that place isn't always the center. Data can tell you a great deal about past performance and it can help you project the results of future experiments, but it can't tell you anything about the test you haven't thought of yet. The best organizations excel at turning to their creative human insights to propose tests and computers to collect and analyze the results.
You need the right tool for the job
As the saying goes: When all you have is a hammer, every problem starts to look like a nail. You should view your data partners like any tool -- pick high-quality vendors who share your goals and leave you in complete control of your data. Look for flexibility and transparency, and be open to new and interesting ways to apply any technology you invest in.
Understand the user interactions behind your "data"
It's easy to fall into the habit of looking at your data as a monolithic block -- easy, but not helpful. Click-through conversions carry a different value than view-through conversions. Customers who have purchased in the past have different behaviors than new customers. Find ways to separate your data pools so that you can pull the most relevant insights from each dataset.
Information? Or action?
Because so much data is available, it's easy to fall into the trap of collecting and analyzing, collecting and analyzing. But without action, all the data in the world won't deliver results. The organizations that seize the testing possibilities in every new input -- and see the potential for data collection in every new idea -- will find themselves on top in this era of big data.
On Twitter? Follow iMedia at @iMediaTweet.