If you’re a marketer working with an analytics team, it’s likely you are providing input for and/or reviewing the output of different types of analysis. It’s critical that you communicate your requirements clearly, check for understanding, and request that all assumptions and business rules are clearly stated in all output. It’s easy to misinterpret and misuse data when definitions and rules are unclear. Here are 3 tips for communicating and clarifying requirements.
Communication #1: What are you counting, and why?
Can you give us a count of business contacts with and without email, by size of company and industry?
This seems to be a fairly straightforward request – tag each contact with a 1/0 flag indicating presence of email, and then count the number of contacts by company-size bands and industries.
|Company Size||Industry||# Contacts with Email||# Contacts without Email|
In the example above it appears that we have good representation of emails in both small and large companies – but do we? What if we ran the exact same profile, only we counted ‘# of Companies with at least 1 contact with an email’? Our output might look as follows:
|Company Size||Industry||# Companies with at least 1 Contact Email||# Companies with NO Contact Emails|
Although greatly exaggerated, we now see there’s a big difference for large companies, with all email contacts belonging to only two companies. If our intent is to implement an email campaign to reach out to our customers, it would be critical to know that we cannot reach a majority of our large customers. In this case, more discussion about the objective for the output should lead to more relevant results.
Communication #2: What do the labels mean? Are the labels clear enough? Are the definitions referenced on the output?
Can you give us count of companies by small, medium and large size?
Your first clarifying question should be to confirm the definition of size. We often see output for various teams within the same organization using different definitions. Individual teams may be clear about their assumptions and rules, but when data and analysis is shared more broadly, this can easily lead to miscommunication of data and wrong conclusions.
We observed a recent scenario in which the label appeared to be very clear, but the two outputs were vastly different:
|Number of Company Locations with > 500 Employees|
For Output A, the definition of >500 Employees was applied at the Account level, meaning every Starbucks Coffee location in the U.S. would fall into the >500 Employee band. For Output B, the definition was applied to each individual company location, meaning a majority of the Starbucks Coffee locations would fall into a much smaller employee band. In this particular case, two teams were comparing output, each convinced that the other made an error. Neither was wrong, but a reader of either analysis could have made an error in interpreting results.
Communication #3: What do we really need to know? How much is too much?
It’s tempting to ask for everything, just in case… Perhaps the greatest communication challenge is to focus your request at the onset of your project. We often see output that is chock full of interesting numbers and calculations, sometimes so deep that our marketing and business peers have trouble interpreting results. It’s not that they aren’t capable of doing so, but we are requiring too much effort on their part to distill the output and extract the nuggets.
What are some of the ways to reduce complexity?
- Decide early on how deep you need to go. If you are analyzing data by geography, do you really need to see county-level detail, or will state-level detail suffice?
- How many different calculations does your audience need to react to? If you are discussing market share, do you need to see market dollars, your business dollars, and percent share – or will two of the three data points suffice?
- Finally, even if you need to provide many different data points, find ways to highlight the most relevant points for your audience rather than making them sift through the data to do so themselves.
What steps can you take to avoid these and other miscommunications?
- Write it down. Keep track of business rules and definitions, in some cases keeping track of multiple definitions based on use and audience.
- Ask questions… lots of them. And then test for understanding.
- Foster good working relationships between marketers, strategists and your technical experts. Review objectives, requirements and output as a team, to make sure everyone is on the same page.
- Use your intuition. If something appears odd, go back and test assumptions before moving forward.
About the Author: