Ratings-Based TURF

The primary objective of a TURF analysis is to identify the set, or combination, of items that appeal to the widest audience. TURF assumes that the item line-up should be developed to provide at least one item for as many consumers as possible, commonly referred to as maximizing consumer reach. 

The results of a TURF provide you with two metrics: 

Reach:

Percent of the audience that is “reached” by at least one product in the set or line. What constitutes reach must be defined by the researcher for the analysis to be conducted. Most often, this is a Top Box or Top 2 Box rating on a 5-point purchase intent scale (e.g., definitely would buy, definitely or probably would buy). Also referred to as Unduplicated Reach.

Maximizing Reach is the most common objective in a TURF analysis, making this the key output.

Frequency:

Average number of items in a set that reach an individual. Frequency should not be interpreted as purchase frequency or volumetric purchase interest. In contrast to Reach, which removes duplication among consumers for a set of products, Frequency measures any overlap that may exist. Increases in frequency indicate an increase in the average number of viable “back up” options consumers may have at the shelf should their preferred product be unavailable. 

Although it is a standard output of TURF, Frequency as a metric is typically considered as tertiary behind Reach and raw item scores (e.g., Top Box, Top 2 Box).

The most common set-up employed for a TURF study is a ratings-based design. Often in product line research, this takes the form of a 5-point purchase intent question asked and answered for every product by every respondent. This article will specifically focus on ratings-based TURF.

For a successful TURF result, it is critical that every item to be included in the TURF analysis be evaluated by every respondent. The analysis needs to know whether or not each item, independent from other items, "reaches" each respondent.

Input:

• Since all respondents need to evaluate all products for a successful TURF analysis, keep the number of products to test reasonable enough for respondents to evaluate in one survey.
• If only asking purchase intent for each product, we recommend including no more than 30 products maximum. If measuring multiple metrics per product, or including a large question set outside of ratings for a TURF analysis, reduce the number of items.
• Other scales can be used, such as an appeal-oriented scale, but keep in mind the design used will be the basis of interpretation for the analysis. If your goal is to report a line-up with the highest reach of consumers where reach implies purchases, then a purchase-oriented scale should be used.
• As with any concept test, be sure to randomize the order of items being evaluated to reduce order bias.

Output:

• With a ratings-based design, reach is typically defined as a Top Box or Top 2 Box response, possibly more depending on the number of scale points (e.g., Top 3 Box on 10-pt scale). Using a stricter definition of reach is more conservative and will result in lower overall reach and frequency values, and vice versa for more lax reach definitions.
• While Top 2 Box (purchase intent) is a popular metric to report in any concept test, using Top Box as reach criteria for TURF analysis has many benefits. Unlike conjoint and other advanced methodologies, TURF does not simulate choice or predict interest in an item in the context of other options. Using Top Box as reach ensures the inclusion of only the most appealing products, and is likely to have purchase intent that translates to real-world purchase. Top Box also provides a lower starting point to grow incremental reach as products are added to a lineup.

Now that you are familiar with the methodology of a Ratings-Based TURF, let's explore how to build one in the aytm platform.

We recommend programming with a Smart Loop (video), which is a feature that allows you to quickly and easily set up your concept test using automation. Smart Loop streamlines the programming of your survey and will set you up for success when analyzing your results; by using the Smart Loop method, you will be able to use our one-click presets to create Virtual Questions for analysis.

In this example, we have 20 new product concepts and will ask respondents about their purchase intent along a 5-point scale. To follow along, open a new draft survey or visit the Editor page of an existing survey on the aytm platform.

1. Add a TURF-compatible question type (sliders, progressive matrix, checkbox, star/smiley) to your survey, hover over the question, and click Convert into a Smart Loop.
2. Input all the product concepts in the Smart Loop table. You can add text, images, or videos in the columns in the Smart Loop tables. The first row are the named variables that will be used in the automated aytm logic.
3. When you have programmed all your variables, the next step is to create your question set. The variable names from the Smart Loop table are used to customize each run. In this example, one variable, [name], to pipe each concept's product name and [image:x] to add the corresponding concept image. (the ":x" notation in the logic statement will expand the image to full size).
4. Preview your survey. The Smart Loop will create a separate run of the set of questions for each item/concept/product. 
5. Continue building your survey. Watch out for the length of your survey - keep in mind the number of questions already reserved by the Smart Loop!

Helpful Hint: Remember, in a TURF, every respondent must evaluate every concept, so be sure the Smart Loop is set to "show all runs/respondent."

1. Export a TURF simulator from within your survey results on the Export tab from the left menu. 
2. Select your question (most likely the purchase intent question), and click download.

Once you have the pre-formatted spreadsheet open, you'll see two tabs of contents: TURF Simulator and All Combinations.

The TURF Simulator tab provides an interface for exploring combinations of items, one set at a time, and returns:

• Number of products (items) included
• Reach - total % of "reached" respondents
• Frequency - the average number of times a "reached" respondent is "reached"
• Reach x Frequency - the average number of times a respondent is "reached" of the selected combination
  • Although this is not a standard TURF metric, it has been used as a tiebreaker or additional data point in weighing possible line-ups.

This tab also provides a column to input a cost for each item and an overall budget input to represent effort associated with a set of items against the pool of resources available.

Note: These values are simple calculations of the total cost of the set of items at face value; they are not tied back to reach or a percentage in an attempt to understand revenue returns if a set of items were to be introduced to the market.

The All Combinations tab holds all possible set combinations for the items being tested with each combination’s output metrics and underlying data that powers the simulator tab.

If you are unfamiliar with TURF or need custom analysis beyond what our built-in export supports, our research services team offers custom quotes for TURF simulators and expert analysis.

Key features of a custom simulator include:

• No limits on set sizes based on the number of items tested
• Support for data filters to explore multiple subgroups within one file
• Support for multiple Reach definitions (toggled?) via a dropdown (e.g., Top and Top 2 Box)
• Preview for “next best” addition, by Reach and Frequency, to existing simulated set
• Template for building a chart of Reach incrementality, often called elbow or whale’s tail

To speak with our team about your next TURF project, open a support ticket or start a live chat.