Bringing TURF Back to Earth
Short on time? Check out the tl;dr at the bottom of this post.
One of the classical product challenges that we deal with in the research world is the idea of feature prioritization. Product planners and marketers typically start with a broad list of product features and capabilities and need to narrow that down into a small list of features to build or market to their target audiences. This is often tackled through a stack-rank exercise like MaxDiff or a Card Sort that produces a ranked list of the attributes from best to worst. But stopping there often misses the point. Another method that can be layered over the top of this to better optimize your top features is called TURF (Total Unduplicated Reach and Frequency).
TURF analysis may sound complicated, but the idea behind it is quite simple. It helps us answer the question, “Which items (features, offers, messages) allow us to use the fewest number of items to resonate with the most potential customers?” In other words, how do you get the most bang for your buck?
Below is a more practical example for how the TURF approach manifests in the real world:
Imagine you own a restaurant and have room on your menu for 3 side dishes but have a list of 10 options you are considering. You want to offer the 3 options that maximize the number of side dishes that you will sell. You start to answer this question by collecting data. So you decide to conduct a survey and ask your customers to rank the potential side dishes from Best to Worst. French fries are the clear winner followed by (2) tater tots, (3) onion rings, (4) mozzarella sticks and (5) Brussels sprouts. You might think the obvious answer is to offer your top 3 items. But if you look more closely at your data you will see something else.
In your data you see that the diners that love French fries also love tater tots. If you offer both French fries and tater tots, you may sell very few tater tots because there is so much overlap with the French fry lovers. In order to broaden your pool of potential side dish purchases, you need to find a side dish that is additive instead of duplicative. You may find a similar overlap with onion rings, given their similarity as a dish. The first side dish in the stack rank that is not duplicative to French fries is Brussel sprouts.
So, even though tater tots are the second most popular item, we probably do not want to offer them. And, even though Brussels Sprouts is the 5th most popular item, it will help us reach more diners than tater tots, onion rings or mozzarella sticks. A big win for our menu.
The results of this straight-forward TURF exercise have taken us a long way to optimizing our menu, but we may still be missing the mark. We are left with questions like:
Does “reach” really mean sales?
Are we maximizing sales at the expense of our bottom line?
Do the options we are offering align with our brand and our other offerings?
We can improve upon the initial TURF insights when we bring our TURF results back to earth using real-world constraints. We will explore how we do this in our next post.
TURF goes beyond stack ranking to maximize the reach of our offer by optimizing for duplicative features that reach the same people twice. However, it also leaves us wondering if we have the solution that maximizes profit.
For more information, please reach out to us at info@tldr-insights.com. We’re always happy to share our experience and help you think through challenging scenarios.
Rich Bojanowski
Partner at TL;DR insights
Rich has been tinkering with data and building models professionally for 20+ years. He has focused on translating the voice of the customer into insights for nearly 10. His technical acumen and hands-on experience with many different technologies make him an invaluable resource to his clients.
Contact Rich Bojanowski