Below are some best practices around TURF! Following these tips and tricks will help you achieve solid results when running TURF actions on Suzy!
- Think carefully about your audience and sample composition when running a TURF question! If you’d like to filter by a certain demographic, consider setting up quotas or an auto-assign to ensure there is an adequate sample when filtering TURF results. Suzy recommends a minimum of n=50 for any TURF filters/cuts.
- Consider pairing your TURF question with a MaxDiff or other relative metric. TURF is an absolute metric, measuring KPIs such as purchase intent or overall appeal. When using TURF, it may be difficult to observe measurable/significant differences in attribute performance. Running a relative metric such as a MaxDiff or Ranking question can help force discrimination between attributes and potentially inform the starting attribute for your TURF.
- Understand how attribute complexity and number of attributes being tested contribute to respondent fatigue and results reliability. If you are testing 30 claims with ~150 characters each, that is a lot for a respondent to process! If your attributes are complex, consider testing fewer of them, or reducing your list via a MaxDiff and only taking top performers into your TURF question. Conversely, if you are testing attributes that are short and digestible, such as flavors, testing 30 in a single TURF question might be an efficient and effective approach. Understanding the relationship between attribute complexity and number of attributes being tested can help you maximize your TURF questions while also maintaining respondent fatigue/results reliability.