The logic of most quantitative marketing research is relatively simple: find the number that will improve your business.

Which attribute will increase market share? Which media channel will have the most impact? These are the kind of questions market researchers try to answer.

You would think that marketers want to know how valid their measures are – and that market researchers try to improve the validity of their measures and models. But you’d be wrong. 

Consider the widely used customer satisfaction score. In a literature search going back over 30 years, I found the average correlation between how satisfied customers are, and what they actually do, is only about 0.2. In other words, 80% of what people do cannot be predicted by how satisfied they are.

It’s not much better for another widely used number: purchase intention. The average correlation between purchase intention and what people do is only about 0.27.

Then there is recommendation. Research shows the “net promoter” score has no correlation to business performance or market capitalisation.

Clearly these are not useful numbers. Why do these numbers perform so poorly and why does our industry insist on using them? The answer to the first question has to do with theory – the second, with the nature of institutions.

Straight line thinking and incomplete models

Market researchers know there’s a difference between what people want to do and what people actually do. What gets in their way is circumstances. For example, children may want Coca-Cola, but mothers won’t buy it. Men might want fast, expensive cars, but affordability stops them. A brand may be out of stock or government regulations interfering with consumer choice. Barriers are present in all markets in varying degrees.

Failure to measure barriers to consumption is an important reason why market research systems predict poorly. It is one of the most important ways market research models are incomplete.

The other is what I call “straight line” thinking or modelling. This happens when a market researcher attempts to predict, say, the sales of Coca-Cola directly from the market spend of Coca-Cola. Another example is predicting the brand image of Coca-Cola from the advertising awareness of Coca-Cola. I call this “straight line” thinking because the attempt to explain something comes straight from an independent variable, say, an aspect of Coca-Cola’s marketing, to a dependent variable, such as the consequences for sales. Straight line modelling ignores the fact that brand use has a much greater impact on brand image than brand sales. For example, what friends and associates say about a brand has a much greater impact than advertising. Communications by other brands also have an impact on one’s brand. Straight line modelling does not take such factors into account.

Measurements and models that predict

When joining Synovate, I wanted to clean up brand equity measurement by measuring only what makes a difference to the brand, and thereby create a more useful model. Synovate’s Brand Value Creator (BVC) does just that. It offers marketers a measurement system that more accurately reflects what’s happening to their brand, and puts tools into marketers’ hands that enable them to make informed decisions.

The key is the recognition of at least two critical metrics to understand what’s happening to your brand. First, you need “attitudinal equity”: a measure of how strong the brand is in the minds of people. Secondly, you need “barrier effects”: a measure of the extent to which barriers enhance or depress the chances of a brand being bought. This research must be at respondent level, otherwise you cannot do a proper driver analysis.

Attitudinal equity measures the strength of a person’s desire to consume each brand in a product category. Like many measures, it is correlated with “share of wallet”. It’s a number that makes a difference to brand sales. Increase your brand’s attitudinal equity and sales of your brand will go up. If your brand’s attitudinal equity falls, sales will go down.

No matter how good they are,  measures of attitudinal equity provide an incomplete picture. We also need to know how a person’s desire to use a certain brand may be distorted by market circumstances. That’s what the measure of barrier effects provides. It shows the extent to which barriers, such as cost, bad shelf space or poor store locations, might lead people to default brands.

By putting these two metrics together, Synovate’s Brand Value Creator improves marketers’ ability to identify the right ways to invest in a brand. The Brand Value Creator identifies why people are using what they’re using and predicts the likely future direction of their consumption.

Scenario planning and “what if” simulation

Synovate has developed a unique range of desktop modelling tools embedded in PowerPoint charts that make interactive modelling possible.

The attitudinal equity simulator allows research or brand managers to simulate the impact of brand image changes on attitudinal equity for both their brand and competitor brands. What makes the simulation possible is the unique way in which attitudinal equity scores are interlinked for all respondents and brands. By using the simulator, you can:

• Establish which brand attributes need to be changed and by how much in order to achieve specific improvements in equity and therefore sales.
• Establish the negative impact on competitor brand equity (and therefore, sales) of improvements in aspects of your brand image.
• Model the potentially negative impact on your brand equity if competitors manage to improve their brand image on specific attributes.
• Explore portfolio strategies that maximise the total business owned by your portfolio of brands and minimise cannibalisation across your own brands.

By combining market mix modelling with the attitudinal equity simulator, we can improve the cost estimates associated with achieving changes in image, equity and sales.

A second simulator combines equity simulation with the simulation of barrier impacts. By combining the two, the simulator takes brand decision support to total market management – a level far beyond attitude and communications management.

Although our simulators should not be used in a mechanical way, they do give brand managers a range of scenario planning tools which help to take the guesswork out of brand investment decisions.

Implementing the Brand Value Creator
The system is based on a normal brand strategy or tracking survey that covers awareness, usage, attribute associations and so forth. The first step to implementing the system is to establish which brands are relevant to each respondent. We then ask four questions about those brands:

1. Brand performance: how does the respondent rate each of their evoked brands relative to their needs in the category?
2. Brand involvement: how important is the role that each of the evoked brands plays in the respondent’s life?
3. Barrier attributes: a barrier-association question that establishes the barriers the respondent experiences in relation to each of the evoked brands.
4. Share of wallet: a question to measure the current share of a person’s spending that each brand in the category is getting.

Once the two key metrics have been established, desktop simulators in PowerPoint format then equip marketers to analyse the impact of various marketing scenarios and improve their decision-making.

Decision-making at the next level
I claimed at the start of this article that our industry appears to insist on using measures that fail to predict, even when many of us know that the measures concerned are weak. I also said that the reasons for this are institutional. In my experience, both our research executives and clients are far too hung up on “numbers” that they’ve been tracking for some years. They’re afraid of what might happen if they suddenly get a “new number”.

Synovate’s approach to brand and communications research aims to change all that. When we discover something that works better, we don’t hesitate to tell clients and to suggest that they change.

The Brand Value Creator is just the first step in this process of continual improvement. The system provides marketers with two powerful measures: attitudinal equity and barrier effects. When these numbers change, the real world changes. We hope that by doing this we make marketing research both more truthful and useful.

Attitudinal equity: a respondent level measure of the strength of each brand which establishes how each person prefers to distribute their spending across a spectrum of relevant brands.

Barrier effects: a respondent level measure of the extent to which market barriers, such as lack of in-home presence, poor distribution and so on, either improve or depress the chances of a brand being used.