TURF Analysis

Description

Finds the combination of items (features, plans, content) that reaches the largest number of users at least once, rather than the individually most popular items.

Ranking items by popularity often picks items liked by the same people. TURF asks which portfolio of k items maximises unduplicated reach: the proportion of respondents for whom at least one item is appealing, with frequency as a secondary measure. Computation is combinatorial (exhaustive for small sets, greedy or optimisation for larger). Inputs can be survey acceptability, MaxDiff utilities with thresholds or behavioural usage. It is useful for deciding bundles, menu items, plan features or which content to commission, especially where breadth of appeal matters more than average preference.

Attributes

Loss type
Unmigrated
Effort
Days

Steps

  1. Collect per-respondent acceptability for each candidate item (binary or thresholded from MaxDiff utilities).
  2. Define the portfolio size k or a cost constraint.
  3. Compute reach and frequency for all combinations (or use greedy/optimised search for large sets).
  4. Examine the incremental reach curve to see diminishing returns as k grows.
  5. Check robustness with bootstrap resamples and segment-level reach.
  6. Choose the portfolio balancing reach, frequency, cost and strategic fit.

Sources

  • Miaoulis, Free & Parsons, TURF: A new planning approach for product line extensions (1990), Marketing Research
  • Orme, Getting Started with Conjoint Analysis (4th ed., 2019)
  • Feature Prioritization Methods (slide deck)

Also known as

  • Total Unduplicated Reach and Frequency
  • TURF

Categories

  • Analytics & Behaviour
  • Product Analytics & Strategy