Should the Net Promoter Score be part of one’s KPI performance review?

Video Transcript

Prof John Dawes: Hello, my name is John Dawes from the Ehrenberg-Bass Institute, and the question this month is 

Should the Net Promoter Score be part of one’s KPI performance review?

The short answer is no. But there are quite a few points to make about Net Promoter Score (NPS) to properly understand why.

Firstly, let’s start with the basic idea that the NPS is a measure of willingness to recommend. Bain and Reichheld, and their consulting arm Satmetrix say it’s a leading indicator of growth. Thousands and thousands of companies around the world use the NPS based on this promise, and partly because all of the competitors in the industry use it as well. As well as that, the sales material for NPS is extremely well-produced and is very persuasive. It often shows examples linking a NPS to the growth of a company. 

But of course, those examples are the only ones where there is a link. We tend not to see the examples where there isn’t. High NPS are meant to indicate that your company will get lots of recommendation. But a well-known flaw with the NPS questioning method is that consumers misinterpret the question, how willing are you to recommend? That’s not the same as actually recommending. Because things like which brand of bank, or lawnmower, or piano, or telco to buy, will only come up in conversation if someone seems like they need a recommendation. So marketing people can massively overestimate the amount of recommendation that actually takes place from looking at or relying on metrics like the NPS. And we also know that recommendation is more likely to come from new customers to a business, not established highly loyal ones like Bain say. New customers – they’ve done something new – like bought a new car or a new phone – that’s something a little bit news worthy to tell friends about. And what that means is that growth can drive the NPS, not necessarily the other way round. 

In fact, we’ve been digging into published examples that purport to link NPS to revenue growth. As one example, some years ago, Bain showed revenue and NPS for Jet Blue in the US. It had a NPS of 62 and had revenue growth of 18%. But the revenue growth was for the years before the NPS. Another example is Verizon. It had a reported NPS of 50 and a growth of 12%. But again, the revenue figures were for the years before the NPS, not after it.

So I looked at what happened next. Well, Jet Blue did grow 19% the following year, I mean, it had been growing briskly for a number of years but in the following year from its NPS, it launched really cheap fares and opened up a lot of new routes across the US. So that’s another reason why it grew, not necessarily just because it had got a high NPS. And Verizon grew only about 5%, which was way lower than the growth it had achieved before its high NPS. So just in those two examples, we can see the NPS is not necessarily a strong leading indicator of growth, as is often claimed. 

Now, while the sellers of NPS say that it’s loyalty or a word-of-mouth measure, NPS for businesses correlate extremely highly with satisfaction scores. So, essentially it’s a quasi-satisfaction question that’s masquerading as a recommendation or word-of-mouth question. So if you want to measure customer satisfaction, measure it using proper satisfaction questions, rather than an opaque question about willingness to recommend. And if you want to measure recommendation, ask about actual recommendation, not intention to recommend. 

Now, we’re not saying getting a good NPS is a bad thing, it’s certainly not a bad thing. But the other thing about the NPS is it’s quite volatile. The process of creating the net score involves: you throw away the 7 and 8 scores out of 10, and then you subtract the 0-6 scores from the 9-10 scores, and that process dramatically lowers the information content in the result “net” score. So this means that NPS gyrate around a lot more than what normal mean average scores would if you just look at the average scores out of 10. And what it means is that it’s quite a poor measure to link to manager’s KPI’s, because it’s going to fluctuate around way too much. 

The final point is that businesses that use the NPS are embracing a mindset that’s wrong. They’re implicitly or explicitly tying their business and the managers of the business, to a belief that growth will come from recommendation. This is sold as the number one figure, the one figure that you need to grow. We know that it doesn’t. Recommendation is a good thing, but people just do not recommend as much as marketers think. Brands grow, businesses grow, by building mental availability and physical availability. Managers are fooling themselves if they think that it is by boosting their NPS, they will magically grow their business. The metrics that should form KPIs for managers should be ones like new customer acquisition, improvements to mental availability, or physical availability. Or doing things that help those factors, such as, how much reach we’re achieving for the dollars that we’ve got to spend, or the prominence of our distinctive brand assets in our advertising, which is making it more efficient and effective. 

The Ehrenberg-Bass Institute is starting a major project called The Net Promoter Challenge to assemble a large amount of evidence as to whether the NPS does or doesn’t link to future growth, and we’ll be telling our sponsors more about this initiative in the coming months.

Thanks for listening.

What is the role of outdoor media in my marketing mix? How does outdoor compare with other types of brand support? And what is the best practice for outdoor to maximise effectiveness?

Video Transcript

Danielle Talbot:  Hello, my name is Danielle Talbot and today I’ll be talking about outdoor advertising. We were recently asked:

What is the role of outdoor media in my marketing mix? How does outdoor compare with other types of brand support? And what is the best practice for outdoor to maximise effectiveness?

These are all really great questions!

In comparison to other mediums, outdoor has high visibility and wide local coverage. It provides greater opportunities for flexibility, relevance and dynamic real-time creative. Outdoor is also considered to be a cost-effective media supplement – supporting previous exposure to other mediums, and reaching harder to reach people, such as light TV viewers.

However, outdoor typically allows for only a brief exposure of a limited, visual message. This means brands don’t have the opportunity to share a detailed message or strengthen their auditory cues. In comparison to other more targeted media, outdoor does have greater wastage and has to compete for attention in a highly cluttered environment. It’s also difficult to measure outdoor, with concerns around metrics, transparency and accountability. But the good news is that the industry is advancing in this area.

Our research into the value of outdoor advertising is currently underway. But here are three key principles of best practice to prioritise when incorporating outdoor into your media mix:

1. Use big, recognisable branding with simple brand messaging and imagery. Remember that people who come across outdoor advertising are focusing on something else, such as walking or driving, so attention is generally quite low – around just six seconds. Use outdoor as a ‘reminder’ medium, extending messages conveyed through your other media.
2. Buy spaces with the most relevance, close to the point-of-purchase. Ideally, a combination of outdoor formats is recommended to reach different people in different ways. Make the campaign cohesive across these formats, but use differing formats creatively.
3. Finally, aim to achieve high reach by using a number of outdoor advertising spaces in different locations. For example, purchasing a single large format billboard in a prime, high traffic area in each city, with other smaller formats supporting this message in other areas. Negotiate greater discounts for longer-term commitment and discover what’s possible within your budget.

That’s it for this week! Don’t forget, if you have a question you would like answered, please contact your Sponsor Officer or ask via our website.

For more information:

BEST PRACTICE: Ehrenberg-Bass Institute Key Media Principles

Opportunity To See

A Guide to Continuous-Reach Advertising

How advertisers rate media buys: Developing rules of thumb to guide media selection

Light TV Viewers: young & desirable or just hard to reach?

Is there research to shed light on whether running advertising and price promotions at the same time is good practice? Or should we try to limit price promotion activities when we run an advertising campaign?

Video Transcript

Dr. Steven Dunn:   Hi, I’m Dr. Steven Dunn and this week I’ll be addressing a question relating to price promotions and advertising.

A useful starting point here can be first to discuss what we know about price promotions and advertising and their best practices. First, what do we know about price promotions? Well price promotions are a tactic for short-term sales activation. They’ll reliably give you a boost to your sales but only for the duration of the discount. Price promotions do not have a long term brand building effect. A brand’s promotion is typically bought by the people who already have the brand in their repertoire. Given the lack of positive long term effects from price promotions, we recommend evaluating these in terms of their profitability. Now most price promotions are not profitable, especially when we start taking into account the wide cannibalisation effects they have on the portfolio.

Next, what do we know about advertising? Well in contrast to price promotions, successful advertising produces both a short term response and a longer term brand building effect. Advertising working in the short term is a necessary condition for the advertising working in the long term – it’s not one or the other. The short term sales effects from advertising are more subtle compared to what we see from price promotions. Exposure to advertising is the role of nudging the propensities of the brands we thought of or noticed in the buying situation, essentially boosting its Mental Availability.

There are two key findings about the effects of advertising exposure on brand choice that we need to consider when thinking about scheduling our advertising. First, with regards to timing, an ad exposure will have a greater impact when it occurs closer to the purchase occasion. Second, regarding the number of exposures. The greatest impact is when a consumer goes from not seeing the ad, to seeing the ad. Subsequent exposures can still have an impact, but they follow diminishing returns. For these reasons, we recommend media plans that spread out advertising to be continuously on air throughout the year and maximise cumulative reach of category buyers.

We now arrive back at the original question of whether advertising the same time as price promotions is a good practice or whether it should be avoided. Now turning advertising on and off to coincide with price promotions essentially amounts to abandoning that ‘continuously on air’ strategy. Could this be justified? Well, we might want to coordinate these two activities if we saw much larger sales uplifts when they occur together, or if our advertising sales effects were less profitable when they occur with price promotions.

We’re conducting research in this space and we’ll be releasing our findings later in the year. But in the meantime, there’s still two things we want to be considering. First, even when advertising occurs at the same time as a price promotion, the advertising will have an impact on people who won’t be seeing the price promotion. Some people will be shopping at different channels or retailers where the brand isn’t on promotion in that week and other people might not be ready to purchase the category that week – they might come in and buy in 2, 3, 4 weeks time where that ad can still have a role in impacting their decisions. Second, turning advertising on and off rather than being continuously on air can lead to excess frequency and lower reach. This means that many advertising exposures will be less effective due to those diminishing returns that occur with multiple exposures.

Ultimately the key challenge is to manage advertising and price promotions separately for the best outcomes at the least cost. This means being continuously on air with advertising and treating price promotions as a cost of doing business with retailers in order to try and maintain or improve your Physical Availability. Aim to avoid deep, unprofitable price promotions regardless of whether or not they’re going to be coinciding with your advertising.

That’s it for this week and don’t forget to submit your own questions. Thanks!

For more information: 

Report 43: Price Promotions

Report 8: The Case Against Price-Related Promotions

Report 47: Cross-brand Cannibalization Kills the Profitability of Price Promotions

Report 66: Best Practice: Ehrenberg-Bass Institute Key Media Principles

How do you distinguish between frequency and penetration? And what time frame should we use to break down usage penetration?

Video Transcript

Alicia Barker:   Hi everyone, my name’s Alicia Barker and this week I’m going to be answering the question:

How do you distinguish between frequency and penetration? And what time frame should we use to break down usage penetration?

This is a really good question! To begin, let’s recap what penetration and purchase frequency are. So penetration is the percentage of buyers of the category who buy a given brand at least once in the specified time period. While average purchase frequency is how many times these buyers buy the brand on average.

Penetration can be used as a measure of how big the brand’s customer base is, while average purchase frequency is used as a measure of the loyalty of the brand’s customer base. When a brand successfully grows its market share, we see penetration increase more than loyalty which shows that growth requires improving the brands reach into the entire base of category buyers. These two metrics vary depending on the length of time window chosen to calculate penetration and purchase frequency.

So what is a good time period to use? Well, that depends on what information you want and what behaviours you want to encourage amongst your marketing team. Increasing the period of time that penetration is observed will capture more purchases by more people, while decreasing the period of time will capture fewer purchases by fewer people. If you use the long time window, like a year, this means that purchase frequency will be slightly higher. It can also have the disadvantage that the non-buyers and once-only buyers won’t be the largest groups. Lighter buyers will still vastly outnumber heavier buyers but all the brands will have a high penetration score. On the other hand, if your data analysis period is very short, then the non-buyers group will be very large as not enough time has elapsed for the occasional or light buyers to purchase. An average purchase frequency will be closer to 1, as few buyers have had time to repeat purchase and this is due to the skewed distribution of a brands customer base.

Our report on ultra-light buyers shows that for a typical consumer grocery brand, more than 80% of its buyers don’t buy the brand every year. So weekly penetration for example, is simply too short of time frame to account for infrequent buyers. Depending on how frequently the category is purchased, you risk mostly capturing purchases from medium to heavy buyers who actually make up the minority of your customer base. But then again, setting too long of a time frame to capture penetration will mean large brands, in particular, have high penetration scores and this can mistakingly give managers the impression that they can do very little to further improve their penetration.

We recommend quarterly penetration, although this can be shorter for categories that are frequently bought. This makes it easier for managers to see the path for growth and that not everyone buys the brand. And remember, in order to grow the brand, you need to nudge the propensity of all potential buyers – including non and light buyers.

Thank you for listening to this weeks Question of the Week. Make sure to submit any questions you may have for us to answer in future.

For more information: Report 73: Ultra-lights — The Unbearable Lightness of Buying

How do you identify and prioritise category entry points?

Video Transcript 

David Thorpe:  Hello, my name is David Thorpe and I have, what I hope, will be a helpful answer to the question:

How do you identify and prioritise category entry points?

A quick recap of what a category entry point is: Now as marketers for our respective brands, how could we increase our brands Mental Availability? And what is the likelihood that buyers think of our brand when approaching the category we compete in. One way is to communicate category entry points alongside your branding in advertising. So category entry points or CEPs are the thoughts that people have when they transition to becoming a category buyer.

For example, have you applied for a personal loan for a car or appliance in the past couple of months or need one in the near future? If not, you’re probably not really thinking about financial service providers and their loan products that much. But when it is time for you to enter the financial services category, you’ll have thoughts that determine which financial service providers in the market you’ll investigate further.

Let’s say we were marketing a brand in the financial services category, we might guess that a category entry points in the financial services is “need to quickly access some money to buy a new appliance”. By advertising the brand with creative, showing people easily and quickly accessing the money they need, the chances are increased that for this category entry points, our financial services brand will be one of the brands out of all the competitors, to come to the buyers mind.

As marketers, we can influence the probability our brand will be thought of when someone enters our category, but only if we have built our brands links to the thoughts buyers are likely to have when they need to buy from our category. But how would we know which CEPs are actually worth investing in? We first identify then prioritise the category entry points. So to identify CEPs it could be a useful investment to conduct an initial survey to understand how category buyers interact with our category, under the framework of the W’s:

• Why
• When
• Where
• While
• With or for whom
• With what
• and hoW feeling

We can then look for CEPs embedded in these experiences. From this first survey, we will have a range of possible CEPs that we’ll prioritise in a second stage and work out which are good options for our brand. If you want to be confident you have a comprehensive set of CEPs, check your CEP list against the W’s framework, it can highlight areas you might be neglecting.

Next we want to prioritise CEPs by considering three dimensions:

• The category buyer
• The brand
• Competitors

For the category buyer dimension, we want to know how frequently CEPs are encountered – the more common, the better. For the brand dimension, we want to discover for each CEP, does our brand currently have an advantage mentally? How extensive is this mental advantage? From this, we can gage how much should be invested in defending or protecting this CEP. Or maybe, should we consider building advantages in new CEPs?

For the dimension of competitors, we look at where competitor brands have mental advantages for each CEP. Fewer competitors with mental advantages for a CEP, the better. The prioritisation of CEPs is a balance between these three aspects. From here, we can use this list of priority CEPs to create a messaging strategy that will increase Mental Availability for your brand, which is the path to brand growth.

What is the evidence that a new launch has to be superior or better than other brands to be successful?

Video Transcript

Kirsten Victory:   Hi everyone, my name is Kirsten Victory and today I’m going to be talking about a question that we were recently asked, which is:

What is the evidence that a new launch has to be superior or better than other brands to be successful?

Well the short answer is no. We haven’t seen enough evidence that more innovative or superior new launches are more successful than those that aren’t. I know we’ve all heard claims that innovative new launches are more successful, but in reality we’ve all bought something new that wasn’t that much different to what we’ve bought before – and that’s if we buy something new on a shopping trip at all.

Institute research has shown that around 9 in 10 consumers don’t see the brands they regularly buy as being that much more different or unique than their competitors. So being different or unique isn’t likely to be all that important to new launches either. We also see very few new products, innovative or not, fail in their first year. So this ‘innovation’ or ‘difference’ doesn’t seem to be the key here. For more information I recommend checking out Report 17 and Report 90 on our website.

So I can hear you asking, “If being different or innovative isn’t all that important, how do I have a successful new launch?”. Well, like for established brands, investments into Mental and Physical Availability are key – not only in their introduction but also in their continued support in the market. We are currently developing some evidence-based guidelines around which Mental and Physical Availability investments are key to new products success, which we will share with our Sponsors soon.

Well that’s it for this week’s question! If you do have a question, make sure you ask us via our website.

For more information: 

• Report 17: Perceptions of Differentiation? Do users see their brands as different?
• Report 90: How often do new products fail?

What is the impact on sales and brand perception from on-pack changes, both in terms of short-term sales and long-term brand impact? What is reasonable to expect?

Video Transcript

William Caruso:  Hello my name is William Caruso. Recently we were asked the question:

What is the impact on sales and brand perceptions from on-pack changes, both in terms of short-term sales and long-term brand impact? What’s reasonable to expect?

This is a really important question. The packaging space is ongoing area of research that I have been looking at for the past two years and still have some way to go. My research is looking at reasons for redesigns along with what helps changes be successful in terms of both sales and perceived success.

First, let’s discuss what we do know about packaging, a quick guideline for redesigns. A statement I like to remind everyone of is “packaging must reach, not teach”. Let’s take an example of a consumer buying a product with a limited edition design. These short-term redesigns often result in increased sales as they are linked with seasonality effects or increased Physical Availability in the store. This is why it looks so appealing to change. Something to consider is small changes at these times – do you want to run the risk of not making a sale because you don’t have a festive pack? However, no change should be so great that it ruins the brand’s identity.

On the other hand, a long-term change such as a refresh, where a brand has made the decision to redesign their packaging with the idea to help them increase sales or fix a perceptual issue is much harder to evaluate. Knowing your brand’s Distinctive Assets in these situations is a must and what you should never change is vital.

From our research to date we can see that brands have slightly different packaging architectures, knowing what elements on pack represent your brand, those which are linked to the category and then those that are creative devices, is a must. More information on the relative strength of these elements can be found in chapter five of Building Distinctive Brand Assets. However, to get the most accurate results we do recommend testing these elements which we also have a research solution for.

In summary, changes are often dangerous. They can increase the number of buyers who fail to notice you, but they can be done right. For example, a large chocolate brand recently made a global change that got everybody talking. From a branding perspective, they maintained their strong Distinctive Assets and the newer pack design has all the ingredients for success.

If you want to contact me in a year’s time to discuss what further knowledge has been developed in the redesign space, I will be happy to chat. For now, all we can do is hypothesise or look at one off-case studies.

Insights into loyalty (or not) within commodity categories – how do we move consumers from light to medium to heavy users?

Video Transcript

Dr Zachary Anesbury:   Welcome, my name is Dr Zac Anesbury, I’m the inaugural Gerald Goodhardt Postdoctoral Research Fellow, and Senior Marketing Scientist at the Ehrenberg-Bass Institute. Recently we were asked if we could provide some insights into brand loyalty within commodity type products. More specifically, you’re looking to know how can we move consumers from being light brand buyers towards being medium or heavy brand buyers, and overall what that would mean for your marketing mix elements?

Now I have some great news for you – there is indeed behavioural loyalty within these types of categories. For example, we see consumers repeat buying a single brand and we also see consumers having polygamous loyalty to a repertoire of brands, within a category like bread.

In relation to light, medium, and heavy brand buyers – one of the most robust findings of marketing science is the distribution of these buyers – we know that this patter always occurs. There are lots and lots of light buyers, there are fewer medium buyers, and there is even fewer heavy buyers. We refer to this as Ehrenberg’s Law of Buying Frequencies.

We know that the relative numbers of light, medium and heavies is actually dependent on the overall rate in which the brand is bought. If the average purchase frequency of the brand, for example is five occasions per year, then the numbers of people buying it once, twice, three times and so on, is very predictable based on that rate.

Now for you, that actually means it is extremely difficult to selectively influence consumers to move from light towards being medium, or medium towards being heavy. If it were possible, the way it would occur is by moving that average rate from five to six occasions. You would see a slight increase of the distribution towards slightly more lights being mediums, slightly more mediums being heavies. But we know the brands buying rates within all categories is very, very stable, so it makes it quite difficult to actually influence.

Alternatively, we recommend re-framing that issue in terms of driving penetration in smaller periods of time. If your penetration is quite large in say, an annual period of time, then you may wish to increase that penetration through maybe six months or a quarter. You can nudge the propensities of consumers so they become from non to light buyers in those smaller periods of time. We do that by pulling on two levers of brand growth in order to drive that penetration. These are:

• Mental availability: linking your category to the needs and purchase situations that it involves.
• Physical availability: ensuring that you have offerings that are suitable for every prominent usage occasion, and that your brands are widely available and easily buyable.

So there you have it, the answer to the most recent Question of the Week. Don’t forget, if you have a question, please send it through to your Sponsor Officer and one of our excellent Senior Marketing Scientists will be able to provide you with an answer.

Senior Marketing Scientists Peilin Phua and Dr Giang Trinh answer: To my view, what has been the most influential work by Professor Gerald Goodhardt?

Video Transcript

Peilin Phua:   Hi everyone, I’m Peilin Phua. Today we will be talking about an underrated but extremely useful analysis for marketers – Conditional Trend Analysis, developed by Professor Gerald Goodhardt and Professor Andrew Ehrenberg.

What is Conditional Trend Analysis? 

I imagine many of you have heard of the Negative Binomial Distribution, or NBD, and how brand buying follows this pattern, that many of a brand’s buyers are light, few buyers are heavy, but the majority of consumers do not buy the brand. But the NBD can do so much more than just showing the distribution of brand buyers. It can predict future purchases using the conditional expectations of the model. And this is where Conditional Trend Analysis, or CTA, comes in handy. CTA is essentially an extension of the NBD model, that can be used to predict the future purchases of the non, light and heavy buyers or any individual buyer based on their past purchasing behaviour.

How is CTA useful for marketers?

CTA can help marketers to benchmark and track buying behaviour over time. Practitioners can use CTA to project sales growth of different buyer groups. It can also predict sales of new brand buyers as well as the churn rate of current brand buyers.

For example, suppose a brand manager has recently run a promotion and there is an increase in sales. The brand manager will be interested to know, whether the majority of the increased sales came from people who had never purchased the brand or if the majority are from the brand’s heavier buyers? If it was the first scenario, then that is great news! Because the brand had acquired new buyers! But if it was the second scenario, then it is not ideal because this means that the recent sales increase from the promotion has taken a bite from the brand’s future sales.

This sounds interesting, how can I learn more?

Well, the original paper was published in Journal of Marketing Research in 1967 and since then many authors such as Don Morrison, David Schmittlein, Richard Colombo, Peter Faders and Bruce Hardie have published studies that are based on the original idea of CTA. Dr Giang Trinh, a Senior Marketing Scientist at the Institute and I specialise in this area. We are currently investigating the implications of advertising cuts on brand purchasing using CTA. If you’re interested to learn more, check out the resources under this video! That’s all for today. Remember to keep sending in your questions and see you again soon.

For more information on understanding Dirichlet type markets click here.

Professor Gerald Goodhardt and Professor Byron Sharp in conversation

As a bonus to our Question of the Week video, we also share two videos of Professor Gerald Goodhardt and Professor Byron Sharp in conversation discussing myth breaking and the Duplication of Purchase Law. Watch the videos here.

Video Transcript

Dr Lara Stocchi:  Hi I’m Lara Stocchi, Senior Lecturer in marketing at the Ehrenberg-Bass Institute. The Question of the Week that I’ll be discussing is Which of Gerald Goodhardt’s papers is the most influential and why?

So 1984 happens to be the year I was born but it’s also the year that Gerald Goodhardt, Andrew Ehrenberg and Christopher Chatfield published the paper “The Dirichlet” in the Journal of the Royal Statistical Society. When I first read this paper in 2007 I clicked, because finally I could read something that makes sense and explained what consumers do. 

To this very day, the paper provides you with three essential explanations to understand buyer behaviour:

1. It describes well established facts about buying behaviour and it provides clarification of how this translates into law-like expectations. 
2. The paper also offers a detailed explanation of the mathematical assumptions or statistical distributions that can mimic these facts. 
3. It also provides a discussion of the result in managerial implications.

I would suggest that this article still serves very well for nowadays audiences — why? Well, it also offers a very clear and simple description of the key measures that are worth monitoring and how to derive them from panel data. It then uses these measures to provide a step-by-step guide on how to fit the model and evaluate its results. In particular, the description of the step-by-step approach clarifies that results can be interpreted from the shape of the statistical distributions and not from the values of the fitted parameters. This is quite refreshing especially in comparison to modern and varying complex marketing analytics models, which do depend upon the estimates of the parameters.

Lastly, the manuscript introduced two important empirical marketing laws, Double Jeopardy and Duplication of Purchase, essentially setting the foundations to what we now know as marketing science. So for these reasons the article from 1984 by Gerald Goodhardt and colleagues is the most influential research paper out of his work. 

Hopefully that addressed this weeks question, please get in touch with your new questions for next week and thank you.

For more information about Goodhardt’s paper click here.