Getting the Most from Net Promoter Score (NPS)
The creator of NPS often points out that to get all the insight from NPS, one should ask an open-ended question after the rating question, asking “Why?” that rating was given.
I’ve had several decades of reading responses on open-end questions. They have occasionally given me a thunderstruck insight. But mostly, respondents avoid giving us much just as they avoided essay questions in school. Open-ended questions just feel like too much work.
Which means we need other measurements to supercharge the insights from NPS. But first, some history.
Brand Health Tracking with Net Promoter Score (NPS)
The Net Promoter Score was created by Fred Reichheld, a Liberal Arts Major who evidently had some disdain for the communication capability of classic statistics, because the NPS is a calculation that essentially delivers the same information as an average would. But getting marketers to take action when an 8.3 average rating drops to 7.9 can be tough, because it doesn’t seem as significant as it might well be. So, innovating on an aversion to statistics? I like it.
An NPS question sequence, consists of a rating on how much or little the consumer would recommend the product/service/organization on a rating scale of 10 down to 0; and an open-ended “why” follow-up. These questions are often incorporated into a Brand Health Tracking study, for good reasons: a) a recommendation rating has long been used* to assess the degree of customer loyalty and enthusiasm and b) since 2003 Mr. Reichheld and Bain & Co. have successfully promoted widespread use of NPS, this allows some comparability “apples to apples” across a wide array of industries, services, not-for-profit activities, etc. (*I’ve been asking about recommendation levels since the 1970s and I’m hardly alone.)
However, NPS is only asked among customers, users or beneficiaries of the product/service/organization, because they have the experience to form a “recommendation” rating. So NPS alone is insufficient as the one key measure for Brand Health Tracking, as BHT seeks to learn the brand’s reputation among prospects as well. Further, if one sets up NPS to only ask of current customers, one never learns what the Trier-Rejectors of one’s brand would rate, ’cause they’ve already left the database. All of that is easy enough to fix, so let’s supercharge NPS.
List of 9 Questions & Analyses to Supercharge NPS
The techniques listed here apply to NPS and any other “overall” rating about brand reputation. In brief:
- Defining Value perceptions
- Distinguishing Loyalty from Frequency of Purchase/Use
- Distinguishing Convertible Prospects from Uninterested Prospects
- Discovering which brands have a Leaky Bucket
- Utilizing Ad Triggers to assess ad reach and effectiveness
- Assessing a wide array of brand imagery attributes, benefits, claims, messages
- Determining whether and when competitors experience message wear-out
- Accounting for seasonality and/or competitors’ ad drop calendars
- Defining Message Clusters and their impact
Here’s a brief description of each, with a link to a more extensive article on most. BUT HERE’S SOMETHING YOU NEED TO KNOW: You may THINK your current Brand Health Tracking study isn’t providing these superchargers but often a good tracking study is capable of it. Let’s talk about whether our analyses can cost-effectively tap the latent power in your current tracking study.
Defining Value Perceptions
A rating of “good value for the money” alone will not cut it. You need to know what proportions of your customers and prospects view your brand as a Solid Value (think Walmart pricing) versus a Premium Value (think L’Oreal or TaylorMade golf clubs). And how many are in the downsides of Beneath Consideration or Beyond Consideration? Customers with a strong NPS rating who view your brand as a Premium Value might slip into Beyond Consideration in a era of recession, inflation or stagflation. Learn more about our ValuEdgeSM system here.
Distinguishing Loyalty from Frequency of Purchase/Use
Some proportion of your users will provide strong recommendation ratings on NPS and yet use a competitor more frequently, or simply not use your brand very often. If its a big proportion, learn why. Your brand may be in a niche use based on: the characteristics of an occasion; or the feature set you’ve been messaging; or your price. Any of these are addressable once you know the root cause. Ask Jameson Irish Whiskey about how often “Irish coffee” is served at a home with their bottle in the cabinet. (Hmm, there’s the Advent season and New Year’s and, ummm, not many other times.) Ask Arm & Hammer Baking Soda, who used to have one box in every home – for months and months and even years. One box. But then they sure fixed that starting in the 1980s, eh? Now nearly everyone knows there are all kinds of uses of A&H. Learn more about LIONCUTSM here.
Distinguishing Convertible Prospects from Uninterested Prospects
I’ve heard pundits say “Everyone who isn’t a customer is a prospect.” Yeah, but not every prospect is worth pursuing. Some prospects have heard or seen something about your brand that has them interested in a future trial. Learn what that teaser is for your Convertible Prospects. Learn more about LIONCUTSM here.
Discovering which brands have a Leaky Bucket
In your category, there may be a brand or two with Leaky Buckets: they have almost as many Trier-Rejectors as they do Loyal-Frequent users. Learn why. Consider using your learning to pitch your Management for more resource money for your plan to grab the lion’s cut of the customers the Leaky Bucket brand is shedding. Learn more about LIONCUTSM here.
Utilizing Ad Triggers to assess ad reach and effectiveness
When we obtain a decent measurement of advertising impact within a Brand Health Tracking study, the useful power in the study is greatly enhanced. Ad Triggers tend to align better with media spending and social media mention levels than other kinds of “ad awareness” measures. Learn more about Ad Triggers here.
Assessing a Wide Array of brand imagery attributes, benefits, claims, messages
Measuring more attributes, benefits and claims is better. You are more likely to be able to include your marketing messages on the list AND messages of your competitors for comparison. You get closer to a 3600 view, which means you’re less likely to be unhappily surprised. So think of that as more “width.”
Unfortunately, asking a respondent to carefully consider their “depth” of feeling about each message on a rating scale is one of the slower processes in a questionnaire. That works against a broad width of messages, as interview time is money and longer length works against respondent participation and their concentration on your questions.
Many researchers will say you have two choices, but you have at least two more. They’ll say you can measure the attributes, benefits and claims that your brand’s category messages to consumers with either a) a rating scale of 5 or 10 points for each item OR b) select “as many as apply” (or “on/off”) for each item. The first option narrows how many messages you can list per respondent (Boo!). The second allows more messages for width (Yay!) but has a hard time distinguishing among brands that compete closely (less depth, Boo!). Speedy rating times enables more messages rated, so there’s our twin goals.
Let us give you two options we’ve found work out better:
a) Use a 3-point scale of Superior, Competitive, Not Competitive that is almost as fast as the on/off for the respondent but gives you a bit more “depth” or,
b) Even better to get both depth and width, use our Imagery SliceSM system that’s faster than a multi-point system and provides a “share” output for each respondent for each message. Width and depth delivered, so more insights delivered.
Determining whether and when competitors experience message wear-out
You’ve heard the complaints from consumers and the advice from marketing pundits and of course they conflict with one another. The consumer complaint is “Ugh! If I hear that commercial one more time I’m going to throw up.” The pundit advice is that without substantial repeat of your message, you can be sure that most of your audience hasn’t truly heard it.
For the marketer’s planning purposes, the pundit advice is more relevant than the consumer complaint. It is very difficult to get a linkage between brand and message to “stick” in the consumer mind. So we repeat messages as often as possible.
Yet this leaves the brand open to the problem of message wear-out. There’s a point where the message has done all it can and now has ceased to motivate more trial nor more loyalty. A wear-out analysis can detect this point.
Wear-out for a major competitor can signal a major opportunity for your brand. When the competitor is shouting a message that’s no longer working, even if their share of voice as measured by spending is greater than your brand’s, their message effectiveness is poor. That’s a good time to ramp up your messaging frequency. (You’ve used our Message Optimization system so you know you have a more effective message, right? No? {sigh} Let’s talk.)
Learn more about how detecting wear-out helps Emerging Brands here.
Accounting for seasonality, competitors’ ad drop calendars
Reading some advice on a website lately: “Conduct your Brand Health Tracking study at least once a year.” Well, the “at least” phrasing saves them but let’s get real. If your brand’s category experiences any kind of seasonal shifts OR if competitors drop their major ad spending OR promotional events at times of year somewhat different than yours, what you’d measure in January could be quite different from what you measure in July.
Patience is a virtue. I know its tempting to spend all your budget the moment you have approval so you can get your answers right now. But over time you’ll be better served by respecting seasonality.
Figure out the maximum number of interviews you can afford in the year, and then spread them out. So rather than 1,300 interviews in January, conduct 100 every four weeks and aggregate into quarterly and annual read-outs.
Defining Message Clusters and their Impact
There’s a purpose to aggregating across two or three years’ worth of Brand Health Tracking also. By creating a mega-database of 3,000 or so respondents, it makes sophisticated analyses of patterns in message ratings possible. We may learn that the number one and two brands have been hammering at the same message theme. Meanwhile, there are two other major themes. Could be time for you to move forcefully to “own” one or both of those untapped themes.
Unlock Your Study’s Potential
We think this bears repeating. Most of the questions we need for these superchargers ARE asked in most well-designed Brand Health Tracking studies. Yet the data is not always analyzed to tap that potential. Talk to us about this. We play nice with other researchers. We’re interested in adding value, not stealing business.
Yes, there are a few measurement/analysis structures unique to us: our ValuEdgeSM, LIONCUTSM and Imagery SliceSM systems for example. But even these are variants of good survey research practice, not mind-blowingly different “probes in the brain” measurements from what other good survey researchers do. So we often find ways to approximate our desired results using questions your tracking study has been asking.
Reach out to us and let’s see what we can accomplish together. (Hint: Fill out the contact form in the column to the right as a first step 😉