Editor's note: James A. Rohde is research supervisor at MARC USA, a Pittsburgh research company. He can be reached at jrohde@marcusa.com or at 412-562-1193. This article appeared in the October 14, 2013, edition of Quirk's e-newsletter.
Brand identities are fluid; they always have been. So it's no surprise that brand equity tracking studies are so common, if not always popular. Internally selling the study is not such a chore given that most people are comfortable with the idea that the brand is important enough to be measured.
The setup, however, can become an ordeal - so much so that the planning stages can derail the whole study before it's even started. (See my article "How to get brand equity studies off the ground" from Quirk's Marketing Research Review, December 2010.)
That said, when these studies are up and going, the pain involved in the setup has a nasty habit of paralyzing the study so that it never evolves to match the changing environment. Around this time, brand tracking becomes unpopular. After five or 10 iterations, you're left with a study that is expensive but ignored since it is no longer relevant to the current market or brand environment.
Think more broadly
To keep brand tracking studies alive and healthy, the idea of "changing" the study needs to be temporarily set aside to let us think more broadly about how to let the study evolve. Once a study has been fielded even once, making any alteration requires a specified methodology that is aimed at mitigating the risk to results.
Changes to any tracking study can mutate the outcome, making the result incomparable to previous waves. Whenever we see a change in any tracking study that has been altered, the immediate question becomes: What changed? Did the market register a change or did the results change because the study changed?
By having a set process that allows these questions to be answered, the study can evolve without losing everything that was gained in previous iterations.
When the study was first starting, you hopefully had a clear set of objectives. If not, look back to the last wave and think critically about what was actually accomplished. At this point we need to clearly define each of the following aspects of the study:
What is the breadth of focus in the study?
- Your brand
- Competitive set
- Total industry
Who is the target?
- Your customers
- Non-customers
- Competitive set
- Industy
- Custom segments
How are brands assessed?
- Brand strengths
- Competitive strengths
- Industry strengths
To bring a current study up to date, start with how the study is positioned right now. While each of these components is an independent consideration, the implications to changing any one of these aspects will impact the other two.
Reflects the current marketplace
In this part of the process, the goal is to determine what needs to change to ensure your brand tracking is measuring something that reflects the current marketplace. Finding inspiration for what the study has yet to achieve is probably the easiest of all the steps. With very little probing, the departments that surround market research will be all too happy to provide a list of things that brand tracking has not delivered.
Take heart and don't let anybody stop with "It's not actionable."
For example, let's say that we are starting with a study that was designed to measure our success in defending our brand against our primary competitors. So we have been measuring brands within our competitive set, among our own brand's customers, against our own brands strengths.
What is the breadth of focus in the study?
- Your brand
- Competitive set
- Total industry
Who is the target?
-
Your customers
-
Non-customers
-
Competitive set
-
Industry
-
-
Custom segments
How are brands assessed?
-
Brand strengths
-
Competitive strengths
-
Industry strengths
I think we can also assume that one of our primary issues is that the study has not been seen as actionable (a common complaint when dealing with an established study). Specifically, we have been only confirming what our brand does well within our own customer base. As a result, we have not seen how we have progressed in expanding our brand.
Two problems
On the surface, this is easy: We just have to start sampling non-customers along with our own customers. However, that raises two problems:
- "Non-customers" is a broad group of people.
- Changing the sample without any additional context mutates the results of the study.
We are not in a position to address reliably adding to our sample until we are very clear on whom we are adding. In a perfect world, there would be a target group in mind that we could sample against. That said, nobody is going to confirm the next acquisition target without some kind of feasibility confirmation and the best vehicle for getting that confirmation is the study we're working on now.
So we need to find a way to define non-customers based on only the things we know about them. In this instance, these are people who are not shopping our brand. This leaves the question: Do they shop our industry or not? Are we selling Android to iPhone users or smartphones to seniors?
For the sake of the example, we are looking at targeting the customers of our competitors. In this instance, all we do is loosen the screening criteria to include anybody who says they shop our competitors even when they do not shop with us. The simplicity of this change is also its strength since we no longer have to make a decision that we may not have the information to make.
We do not want to make assumptions about age, income, gender, etc., if we don't have to. By letting the basic attributes of our sample surface naturally based on their behavior, we allow ourselves the opportunity to see differences that would go away if we forced all respondents to look the way our customers look.
Assess the brands
This leaves us with determining how to assess the brands. Now that we have non-customers, simply rating our own brand strengths is not going to help us understand why some consumers prefer other brands over our own. This means we have to make sure we address competitive strengths in addition to our own brand strengths.
At this point, we know what dimensions of study are going change:
Breadth of focus in the study: stays the same
-
Your brand
-
Competitive set
-
Total industry
Target of the study: increases to include non-customers
-
Your customers
-
Non-customers
-
Competitive set
-
Industry
-
-
Custom segments
Brands assessment: increases to include competitive strengths along with our own
-
Brand strengths
-
Competitive strengths
-
Industry strengths
New brands, new attributes
Assessing new brands means new attributes. Of every step so far, this is definitely the most difficult. As we add attributes to measure the brands, we will likely need to remove attributes to make room. This is where research must have a strong methodology in place to determine what attributes are best suited for removal.
Having a methodology in place to determine what can be removed does not mean that we'll get our way. However, we will be in a position to speak factually about the actual insight being gained or lost as these replacements are made.
Before making any rash judgments, research your research.
- Take your most recent data set and run a linear regression against your dependent variable (usually overall satisfaction).
- Rank order your standardized beta scores so you know which attributes have been most strongly contributing to your brand performance.
- Attributes that fall to the bottom can be added as candidates for removal but don't cut them quite yet!
Next, still using your most recent data:
- Run a factor analysis on all your attributes.
- If you have over five factors, take a look at low increases in your explained variance to determine if there are entire groups of attributes that are not adding enough value.
- On the other hand, if you have fewer than five factors, look at the individual attributes and determine which are duplicating insight.
- Attributes that provide the least amount of explained variance are candidates for removal but again, don't cut them yet!
At this point you should have two lists of attributes that could be suited for removal when additional attributes need to be added. This is where some level of art enters the process. The two analyses are probably not going end in the same lists, so understanding how the attributes resonate with your internal audience, along with the intricacies of how the attributes interact with each other in the data, will play a role in how your final list is created.
Again, just because you will now have a reasoned point of view does not mean that you can control the changes. However, even if you are ignored completely, this work will give you the context needed to understand how attribute changes could impact the next iteration of your study.
Know everything
After we have confirmed the final list of attributes, it is easy to get carried away with the relative ease of implementation. A word of caution: Know everything about what you are changing and what is staying the same.
There is no such thing as a small change when it comes to any tracking study but this is especially true within brand equity. Keep in mind, everything discussed in this article was the result of adding non-customers to our sample in addition to the customers we had been speaking to before. Having a solid grasp on what will be tracked from the prior wave to the current one is going to be the difference between successful evolutions vs. failed mutations.
If the direction of the study changes too much, evolving a brand equity study will not always be possible. But when the option is there, holding on to all the value created by past research will further add to the study's ability to positively impact the business.