INSIGHT on INSIGHT: Guidelines for Creating Analysis Groups
While the amount and type of analysis are key factors that drive the overall cost of research, most projects should still try to do just a little more analysis than what they know is needed. It is wasteful to do too much analysis (that either produces nothing of value or has little or no influence on decisions), but it can be even more costly to not do enough analysis.
This is similar to the principle of always studying just a little bit more for a test than what was needed. Too little time studying will be evident by getting a lower grade while just a little too much time can ensure a top score while not having risked too much opportunity cost from investing an extra few hours.
Advanced planning is critical to arriving at the right amount of analysis. And, unfortunately, there is no standard approach or simple formula to know both how much and what type of analysis is just right.
LET THE PROJECT GUIDE YOU
Looking to the project’s objectives to provide initial guidance is obvious. But better plans ensure less obvious analysis is included that could produce really surprising insight that gets beyond the core objectives.
This begins by deciding the right slices of data to pull out of your research. A typical internet survey can produce literally hundreds of possible independent variables to use when tabulating the data. Because it is unreasonable to run (let alone review) every possible report, smart decisions need to be made to ensure the right groups are included.
Consider the following ten areas or approaches to initially guide what groups you include, either because they are necessary to deliver the project’s objectives or they seem likely to reveal other useful insight.
MUTUAL EXCLUSIVITIY: Are the analysis groups you’re using discrete, where a respondent can only fit into one group? Or are results blurred because groups overlap? Groups with numeric definitions are generally easy to separate and cleanly compare (like ages or incomes). However, other results (like shoppers of different retailers) can be difficult to interpret when individual respondents fit multiple groups (such as when a high percentage of respondents shop Walmart and Target and Walgreens).
DEMOGRAPHICS: What demographics could drive different attitudes or behaviors? Some of the common demographics we use to cut data include gender, age, income, geographic region, population density, household size and age or presence of children.
VALUE: What groups represent more or less value to your business? Consider large size v. small size buyers, frequent v. infrequent buyers, high volume v. low volume users, brand name versus private label buyers. Could today’s low-value buyers become tomorrow’s high-value buyers?
STORE SHOPPERS: Retailers want to understand your product in the context of there shoppers. However, showing information about competitive retailers (or what one retailer’s shoppers do at another retailer) can motivate more aggressive support.
STORE BUYERS OR NON-BUYERS: People walking into a retailer, but not buying your category could represent one of the easiest sources of incremental sales.
BRAND BUYERS OR NON-BUYERS: Understanding the differences between those that currently buy your brand and those that don’t can provide very fundamental perspective regarding who your prime prospects are or could be. Unfortunately, it is typically cost-prohibitive to get a sufficient number of buyers for brands beyond those with dominant market share. At the same time, the value of these groups can be limited by how discrete they are. In categories with high switching across brands, the data can be too blurred to provide meaningful or useful information.
KEY PSYCHOGRAPHICS: Are there critical attitudes that influence interest in your product? Does your audience believe they deserve to occasionally splurge on themselves? Do they believe you get what you pay for is true in your category? Do they derive social status from the things they own?
NECESSARY OR BENEFICIAL HABITS: Would it be beneficial to understand people that make at least five home-cooked meals each week and regularly buy fresh ingredients? Are they people that clip coupons and try to always buy on deal? Are they weekend warriors that like to do DIY projects? Do they stock up on your category or only buy once they run out? Some habits may be critical to making your value proposition relevant or they could reveal the opportunity to better align your offering with their existing habits.
CATEGORY MATURITY: Are you trying to understand people that are new to the category (and still exploring and learning and establishing habits) or those that have used the category for years (and might have loyalties that are hard to change, no matter how outdated they may be)?
TERMINATED RESPONDENTS: Most research disqualifies some segment of respondents. Sometimes, analysis of data from terminates can reveal new opportunities that go beyond the focus of the research. This may include identifying additional unmet needs (for different products to address) or understanding why certain people are not prospective buyers of your product (and what it would take to make them more interested).
This is not an exhaustive list, but hopefully puts you on the right path to identify the appropriate analysis groups to include in your project.