TURF analysis, short for Total Unduplicated Reach and Frequency analysis, is a statistical technique used in market research to find the combination of products, features, or messages that will appeal to the largest possible audience. It solves a specific problem: when you can only offer a limited number of options, which ones should you pick to attract the most people?
The Core Problem TURF Solves
Imagine you run an ice cream shop and you’ve developed five flavors, but you only have room to sell four at a time. Your instinct might be to pick the four most popular flavors individually. But that approach has a blind spot: your top flavors might all appeal to the same customers. If chocolate brownie fans also love chocolate cookie explosion, offering both doesn’t bring in anyone new.
TURF analysis catches this overlap. In a real example using survey data from 100 consumers, a lineup of the four individually most popular flavors (chocolate brownie, chocolate cookie explosion, extreme vanilla, and coffee crunch) reached 80 out of 100 consumers. But swapping chocolate cookie explosion for the less popular mango madness, which attracted a different slice of customers, pushed the reach to 90 out of 100. Fewer people liked mango on its own, but it brought in buyers that the other flavors missed.
This is the central insight of TURF: the best combination isn’t always made up of the individually best items. It’s the set that covers the most ground with the least redundancy.
How Reach and Frequency Work Together
TURF measures two things. Reach is the percentage of consumers who would choose at least one option from a given set. If you offer four flavors and 90 out of 100 people would buy at least one of them, your reach is 90%. Frequency is the average number of options each consumer would pick from that set, revealing how intensely people engage with your offerings.
These two metrics tell different stories. High reach with low frequency means your lineup has broad but shallow appeal: you’re attracting lots of people, but each person only connects with one item. High frequency with moderate reach means you have deep appeal within a narrower group. The ideal scenario is high reach and high frequency, where many people are interested and each of them likes multiple options. TURF helps you see these tradeoffs clearly so you can decide what matters most for your goals.
What Makes It “Unduplicated”
The “unduplicated” part is what separates TURF from simply counting how many people like each item. When calculating reach for a combination, each person is counted only once, regardless of how many items in the set they’d choose. If someone would buy both chocolate brownie and extreme vanilla, they add one to the reach count, not two. This prevents you from overestimating how many unique customers a product lineup will attract.
Mathematically, the process works by marking each survey respondent as “reached” (a 1) or “not reached” (a 0) for a given combination of items. A respondent counts as reached if they selected any item in that combination. The reach is then the proportion of respondents marked as reached out of the total sample. The analysis runs this calculation for every possible combination of items at a given size, then ranks them to find the winner.
How the Data Is Collected
TURF analysis starts with a survey. Respondents see a list of options and indicate which ones they’d buy, use, or find appealing. This is typically a simple checkbox-style question: “Which of the following flavors would you be interested in purchasing? Select all that apply.” Image-based selection questions work too, which is helpful when you’re testing visual concepts like packaging designs.
The responses get converted into a grid where each row is a person and each column is an item. A 1 means that person selected that item, and a 0 means they didn’t. From this binary grid, the algorithm tests every possible combination of items and calculates the unduplicated reach for each one.
Some researchers feed TURF analysis with data from a MaxDiff exercise instead of a simple checkbox question. In a MaxDiff survey, respondents repeatedly choose their most and least preferred options from rotating subsets, producing a more precise preference ranking. This can sharpen the TURF output, especially when the list of options is long and you want to account for how strongly people feel about each choice, not just whether they’d consider it.
TURF vs. MaxDiff Analysis
MaxDiff and TURF are complementary but answer different questions. MaxDiff ranks features or products by preference, telling you which single item people like most and which they like least. It’s powerful for understanding relative appeal, but it doesn’t directly tell you which combination of items will reach the broadest audience. Two top-ranked items might appeal to the exact same people.
TURF picks up where MaxDiff leaves off. It evaluates combinations rather than individual items, identifying the set that maximizes overall reach while avoiding redundancy. In practice, many research teams use both: MaxDiff to understand individual preference intensity, then TURF to select the optimal portfolio.
Common Business Applications
Product line optimization is the most common use case. A beverage company deciding which five flavors to launch from a list of twelve candidates, a snack brand choosing which varieties to stock in a convenience store with limited shelf space, or a cosmetics company selecting a starter set of shades can all use TURF to make that call with data rather than intuition.
The technique also applies to communications and media planning. If you’re running an ad campaign and can only afford to place ads on three channels, TURF can identify which three channels will reach the largest unduplicated audience. The same logic works for messaging: if you have six possible taglines or selling points but only room for three on your packaging, TURF reveals which combination resonates with the widest range of consumers.
TURF is particularly useful when resources are constrained. Offering too many choices can actually reduce sales, a phenomenon known as choice overload. The analysis helps you pare down to a focused set that still captures the maximum number of potential buyers.
Running a TURF Analysis Step by Step
The process follows a predictable sequence. First, define your target market. Consider demographics, buying behaviors, and the specific needs you’re trying to address. This determines who you survey and ensures your results reflect the audience you actually want to reach.
Next, design and field your survey. Present respondents with your full list of options and collect their selections. Sample size matters here: you need enough respondents for the combinations to be statistically meaningful, especially if you plan to segment results by subgroups.
Once data is in, the TURF algorithm evaluates every possible combination at each portfolio size. If you’re choosing 4 items from 12, it tests all 495 possible four-item combinations and ranks them by reach. The output typically shows how reach increases as you add each incremental item, making it easy to see the point of diminishing returns. Maybe going from three items to four adds 12 percentage points of reach, but going from four to five only adds 3. That helps you decide whether the extra item is worth the cost.
Finally, you act on the findings. The optimal combination feeds into product development, marketing strategy, or resource allocation, depending on what you were testing.
Software Options
Several platforms handle TURF analysis. Displayr and quantilope both offer built-in TURF capabilities aimed at market research teams. SightX provides an AI-powered research platform that includes TURF among its tools. Survey platforms like Alchemer can generate TURF reports directly from checkbox question data within their standard reporting. For researchers comfortable with code, R and Python both support TURF calculations through straightforward scripts that loop through item combinations and compute reach for each one.
The computational demands scale with the number of items and the combination size you’re evaluating. Testing all possible 5-item combinations from a list of 30 means checking over 142,000 combinations, which is trivial for modern software but worth knowing if you’re attempting it manually in a spreadsheet.