Indiana Super Bowl and the Temple of Data

Posted by

A commentator writing in a recent edition of Marketing described data analysts as the new ‘high priests’ of the marketing world, able to conjure up otherwise unfathomable insights and measurements from the vast oceans of data on which the industry now floats.

Far be it from me to dispel such perceptions (!) but the reality, of course, is still somewhat different.

Having large volumes of real-time data at your fingertips does not necessarily lead to the generation of major insights about the likely needs or intentions of the decision-makers. This is especially true in the analysis of social media data, where much of the information shared may have no value at all.

For example, the June 2013 edition of the Journal of Digital and Direct Marketing Practice contained an article on the use of social media tracking data in helping to manage the Indiana Super Bowl. The purpose was to publicise the potential benefits of using such data in marketing analytics.

What was striking about the article, however, was not the specific results but the lack of structure used to derive insights from the data. In many respects a case study in how opportunities in data can be wasted if there is no pre-planning.

At the sports event data was collected from a small range of social media feeds (Facebook; Twitter; and some relevant blogs). The result was a collection of 71,063 posts. Interesting enough, but hardly a deluge (and not what one would call “Big Data” – looks more like “Tiny Data” to me – but I digress).

The data was analysed in real time every day using a number of automated (‘word cloud’) and manual processes throughout the period of the event, beginning in the weeks leading up to the event and the weeks that followed. Posts were classified as either positive, negative or neutral and within each of these topics were identified that were fuelling the positive and negative comments.

The authors of the study say that the organisers used these “insights” to derive real-time assessments of the efficacy of the marketing campaign for the event in addition to mitigating threats to safety and react to PR opportunities & weaknesses. No specific examples of actions taken in response to having this information were able to be given by the authors however. What they do describe is how the flow of information was centred on 3 key themes:

- Hospitality
- Accommodation
- The game itself

and within each of these on a range of sub-themes such as:

- The friendliness of the cityfolk
- The city’s entertainment facilities
- Views on hotel prices and standards of accommodation
- Parking facilities
- Over-crowding

and so on.

The derivation of these common themes was, of course, easier to do when all 71,000 posts were compiled. The question is, how much data is needed before you find something that is ‘out of the ordinary’ and a true insight that means you can change something? There are no examples quoted.

Our view of the results from this exercise, which is held to be a ‘good example’, is that what it could well be a memo issued by Monty Python’s “Ministry of the Bleedin’ Obvious” and that few, if any, useful pieces of information were obtained.

The only useful information being fed back to those policing the event was that the city and the area around the stadium were becoming over-crowded on match day. One assumes the police may have guessed that might happen. Moreover, the police commanders on the ground could presumably use their eyes to see that over-crowding was occurring – unless of course they did not believe it until they saw that it has been reported on Twitter (it used to be CNN but that was in the old days).

All-in-all we could not see any evidence that the analysis of the social media data added anything to either the execution of the marketing campaign or to the way the event was organised and run. That is not to say it could never have any benefits – just that this particular piece does not show any.

What it shows to us instead is a missed opportunity.

Judging the success of the marketing campaign could have been monitored more effectively if a set of target topics had been established at the outset – such as the five dimensions we use to evaluate marketing effectiveness. Analysing the social media feeds along these 5 dimensions would have given much more insightful outcomes – and ones that could then be benchmarked.

Similar sets of ‘target topics’ could have been set-up and activities measured against them. This is not rocket science – all you need are some people who know about organising the event.

So in this case we feel that the Holy Grail fell into the fissure and was just beyond reach. If only Indy had been there (or maybe the Monty Python team).

Post-recession Priorities for UK Consumers

Posted by

Writing in the September issue of Marketing magazine, Andrew Curry, a Director at the Futures Company, reports that the post-recession consumer will, for some years to come, have a very different set of priorities to those before it.

He points out a number of key changes:

- More cautious attitudes towards debt
- Social connectivity (and the support network it engenders)
- Greatly increased intolerance of “rewards for failure”
- Looking for value (not cheapness)
- Needing assurance
- Seeking “purchase decision process simplification”

In reality, though, almost all of these are not, in fact, new. They were just as present before the recession. How so? Because they are all manifestations of the five key dimensions that drive consumer preferences, namely:

Relevancy (right solution)
Identification (right image or brand identity)
Accessibility (perceived – and actual – obtainability)
Value (right cost-benefit trade-off)
Confidence (expected reliability that the brand promise will be delivered)

Most of the points Andrew identifies can be re-interpreted as examples of these five forces at play.

Seeking a simplified purchase process is a way that consumers have of making brands more accessible. And the constraints on their ability to purchase imposed by perceptions about their future likelihood to be able to repay debt is also part of perceived accessibility in the form of “perceived affordability”.

Perceived value is a perpetual driver.

Increased intolerance of institutions that reward failure is one element of managing brand image and reputation.

And seeking reassurance is the Confidence dimension.

The social network change is slightly different and especially interesting, however. It underpins all 5 dimensions by providing both an information flow and a feedback loop for all of the others. Increased and more rapid connectivity and feedback makes the probable scale and speed of consumer response on each of them much greater than ever before. Meaning that the ‘penalty you will pay’ for being out of line is going to be much sharper, deeper and more painful than ever before.

That, I believe, is the real lesson to be learned.

Measuring the Marketing Benefit of Social Media

Posted by

In the July 2013 edition of Marketing Richard Medley of Nexus Communications highlighted the fact that Facebook ‘Likes’ were a debased currency. Marketers employing social media tactics just to maximise their number of ‘Likes’ were deluding themselves and the only value in the data generated was in proving that the campaign had happened, not that it had been successful or effective.

In fact, so much in a similar vein has been written on this topic it has become amusing. It was recently described as being as popular a topic as teenage sex. Everyone says they are doing it but how often and how well no one knows. So as practitioners in the field (of Measuring Marketing Effectiveness – not teenage sex) we thought we would inject some realism.

The problem of measuring the effectiveness of social media is like that of experiential marketing – it is easier to measure that it has happened than to measure what effect it has had. Sure there are some immediately measurable impacts, especially if there are sales links or incentives to participate, but that is not quite the same thing as measuring what impact it has had on underlying long term brand preference. Worse still, unlike advertising where bad ads are usually either ignored or just do not get noticed in the first place, bad social media and/or bad experiential can actually damage a brand’s reputation.

So we need a way to judge how comments are influencing underlying brand preference – and a way of doing it directly from the data so that we can do it in real time. The answer is to combine skills and expertise from traditional market research with knowledge of how to measure effective brand preference and data that shows how underlying brand preference affects sales. How can this be done?

At RedRoute we know that our Effective Net Preference (ENP) Score, based on brand perceptions across the five key dimensions that most influence likelihood to purchase, explains underlying sales trends more accurately than other metrics such as Net Promoter Score, Brand Equity measures and so on. This is because it includes perceptions not just about brand image and whether the product “does the job” but a holistic set of five key factors that influence actual ‘preferred choice’ behaviour. These factors are: Relevancy; Identification; Accessibility; Value; Confidence. For more information about these please visit our web site.

How do we know these work? Because we have predicted individual choice behaviour as tracked by loyalty cards and through customer sales programmes for 12 months ahead with very high degrees of accuracy, and not just for one or two people but for millions. In other words, “been there, done it, got the t-shirt”.

Into this framework has stepped social media data – and when you look you find you can classify social media comments using the same types of code frames as we use for verbatim comments collected via traditional market research. We have the same positive, negative and neutral spectrum. And we have the same five dimensions as mentioned above. And they give the same read on what is happening to underlying consumer preference. The only difference – it is more immediate and because of that, more valuable and more actionable.

Tracking social media comments enable you to see over time whether a particular social media campaign has moved the needle on any of the five key metrics – and whether positively or negatively. Amplification measures enable you to scale the likely impact onto overall effective brand preference. The only remaining step is to prove the relationship between Effective Preference and sales. That can be achieved via both market research surveys (which directly show the links between, for example, a brand’s ENP score and that brand’s share of wallet) or by modelling actual sales data using ENP as one of the explanatory drivers.

So this is the real way to measure the value of social media to long term brand building. And as our clients know, it works.

ENP is probably the most important brand metric that you have never heard of.

But master it and you will master the fortunes of your brand.

So you may well hear more about it in the future.

The Problem with Human Beings – from a Marketer’s Perspective

Posted by

Rory Sutherland, writing in the September 2013 Edition of the journal of the Marketing Society, Market Leader, highlights in his usual eloquent manner the potential tendency for people’s ‘threshold of indignation’ to be lowered if they believe an organisation is not using information they hold about them to improve their customer experience.

This is, of course, not a new phenomenon.

The classic problem with human beings is that they are capable of independent thought. That is why armed forces training is the way it is. Sometimes too much independent thought is not good for business, or rather, not good for controlling a business.

Traditional marketing didn’t have to worry too much about this. You created your message, put it out there, and people either responded or they didn’t. There was no expectation on either side that it would be any different.

Now, however, buyers expect you to know who they are if they have shared information about themselves with you. They don’t expect you to ignore it and, because they place a value on it, if you do ignore it then that says to them that you do not value their custom.

Fortunately these ‘feedback loops’ as they are termed in the world of customer data modelling are a normal part of the information set used to predict future customer behaviour.

If someone holds a loyalty card this will be flagged on their customer record and forms part of the set of data we call ‘circumstance’ information. When modelling customer behaviour this fact will make a difference in the modelling and combining this with other information, such as the facilities available at his local Starbucks store, will enable us to predict Rory’s rightful indignation that as a supposed valuable customer he would be more put out by the withdrawal of useful services such as WiFi (even more so in that case as the customer record would also show he was a user of the WiFi).

I can assure Rory that there are some companies out there (like Sainsbury’s, Carphone Warehouse, Homebase and others) who do take those types of considerations into account when planning business decisions.

They form part of an overall framework known as ABC Modelling where ABC stands for Attitudes, Behaviours and Circumstances and you can learn more about it on our analytics web site (

As a practitioner in modelling customer behaviour it is amusing to see many digital marketers suddenly realising that there is another whole dimension to consumer behaviour – the fact that people do not only have opinions but those opinions really do change and influence behaviour. If only all humans were like Pavlov’s dog it would be so much simpler (but, I would argue, a lot less interesting).

Greener Grass – Measuring Marketing Effectiveness in an Omnichannel World

Posted by

Do you know the expression: “The Grass is Always Greener on the Other Side”?

Well it’s funny to see it happening in marketing.

Phuong Nguyen, Head of Advertising at eBay, commented in the October 2013 edition of Marketing magazine that the “increasingly complex” purchasing journey – where search and purchase behaviour are so intertwined that no one knows where one starts and the other will end – was making it difficult for digital marketers to measure the effects of their activities.

Moreover, another digital marketer recently asked me how to justify the budget for digital when “it’s easy for traditional marketers to justify the effectiveness of their spend at building a brand but digital marketers cannot because they can’t reach enough people quickly enough”

Strange but true.

Meanwhile marketers using traditional media channels lament the fact that their digital counterparts can track the effectiveness of their activities in a manner they can only dream of.

Seems the grass really is always greener on the other side.

Unfortunately the problem with both these views is that neither is looking in the right direction. Both, rather ironically, start from the idea that marketing is “done” to consumers and that internal business metrics, if only configured correctly, could solve the riddle.

But that is an impossible dream.

No individual company will ever have a complete view of each and every consumer. To seek it is to seek the modern El Dorado. The 21st Century version of Fool’s Gold.

A more practical – and more consumer-focussed approach is needed to measuring marketing effectiveness. One that looks at the 21st Century World through the eyes of the consumer and judges marketing effectiveness as its success in delivering the right solution to my needs. Today, tomorrow and for as long into future as relevant. One of the best ever lines in financial marketing sums this more admirably by most: “I’d like a pension company that won’t retire before I do.”

Consumer purchasing is driven by more than short term offers, who has the best looking web site or great customer service. It is driven by all of these and more. And what’s more this ‘pull’ can be measured.

Almost all digital measurement is ‘transaction oriented’ and whilst some of those transactions can be used to infer likes, preferences, and other attitudes, the overwhelming use is simply as a more real time version of direct marketing. Offering deals and adverts to influence purchasing is hardly cutting-edge, no matter how much more sophisticatedly you can target them. It is simply information direction, and just one element of marketing effectiveness.

Getting someone to buy through offering them a deal is purely transitory, a bribe to get them to shop with you. Here today but gone tomorrow. To get them to use you again your ‘short term’ deal must become a permanent price cut, which works until the guy down the road offers an even better deal than you do.

Truely effective marketing is wider than this and needs a wider metric to judge it by. One that tells you how powerful your marketing is at influencing customer purchasing behaviour whether there’s a deal or not. One that measures the “pull” a brand or store has and explains why when you see something online you go to a specific store to check it out.

This “pull”, which we term “Effective Net Preference” covers the five key dimensions that influence customer purchasing behaviour:

- Relevancy (is this what I need?)
- Identification (are these the people I want to buy it from?)
- Accessibility (how much hassle is involved?)
- Value (is it a fair deal?)
- Confidence (what’s the risk I’ll be disappointed?)

If your marketing is getting you ticks in all these boxes with more consumers than your rivals then market share gains will follow and not just because you are the ‘cheapest on display’ this week.

ENP measures the effectiveness of your marketing programme through the eyes of the only person who counts, the buyer. It is not based on what the retailer or manufacturer’s view of the outcomes but the buyers view of what all your marketing efforts means to them.

And it is not idealistic. The measure can be derived at brand level or for individual customers. It can be turned into propensity scores for targeting and predicting future spend behaviour. By measuring your strength on each of the five underlying dimensions it can be used to improve your offer in ways that increase preference without reducing price. It can be used to ensure you are becoming more buyer-centric and valued by them for more than just convenience. And it can be used to track the preference for your brand over time, the “willingness to deal” that keeps your sales strong whilst others may falter.

Best of all, however, it is not confined to a single channel, a single media, or a single marketing lever. We are using it, for example, to convert social media dialogue into a real time measure of consumer brand preference. All social media comments about a brand, product, service or company can be classified under one of the five key headings listed above and ranked on their level of positive endorsement or detraction. This enables real time tracking of strengthening or weakening consumer preference and knowledge of the sources of those movements. Simplifying the complexity that Nguyen finds challenging.

Moreover, using an endorsement scale instead of just ‘positive, negative, neutral’, makes it possible to produce ‘Sentiment Indices’ on each key driver so you can see whether you are staying in tune with buyer needs – much more quickly than via traditional research. So you know all the time whether the consumer thinks your brand meets their needs and is delivering on its promise.

ENP is probably the most important brand metric you have never heard of. Master it to guide your business and you’ll place your future destiny in your own hands. Remain locked into the mind sets displayed by the current era of traditional and digital marketers alike and you will be buffeted by the winds at every turn.

To find out more about ENP just visit our web site.

The Democratisation of Personal Data

Posted by

The October issue of Management Today has an article by Simon Caulkin that highlights the fact that ‘Big Brother’ technology is now so pervasive, and the ability to link it together so close to being complete, that the debate about public versus private is about to explode.

As we highlighted in an earlier blog, for data to be of real value it must provide a complete picture of the person. Their attitudes, their behaviours, their circumstances. That pulling such data together enables highly predictive models to be produced is not new news. We have been compiling and using such models ever since we started analysing supermarket loyalty card data.

The issue that Caulkin raises, however, is that doing it with the permission of the individual is one thing. Doing it without that explicit permission is quite another. And, more than that, people at large do not even realise what is possible.

In Orwell’s book 1984 he paints the picture of a society where cameras record everything and armies of people watch and codify all that is happening. This is no longer science fiction but science fact. Amazon has a global army of such paid volunteers (called Turkers) to process information. The mobile phone companies use them as well – ever wondered how some of the Siri-type functions on your Smartphone work? There are real people reading and responding to your requests.

Sounds far-fetched? It isn’t.

What’s more, the technology to make ever more and ever faster connections grows all the time.

Amusingly, new companies are now springing up allowing you to control your online personality, yet more faceless institutions to help you manage the original set of faceless institutions. So how does that work then?

Obviously the threats of data misuse are very real and very big. But trying to control this at the point of use is like legislating the way the wind will blow. And giving people tools to control what can and can’t be kept is not what people want.

What people want is for undemocratic and exploitative ways of using the data to be illegal and the penalties for misuse to be draconian. And to have the right to opt out. That won’t be easy but ultimately it is the only workable solution.

This debate needs to take place and sooner rather than later.

If you agree – let Schezzer know (or do you want to opt out?). :-)

Doubling the Profits from CRM

Posted by

Using Market Research to Optimise Customer Database Strategies

In 1898 the British Navy found that for every 1000 rounds fired by its battleships only about 17 ever hit the target. So Admiral Percy Scott studied the techniques employed by his gunners and discovered the most accurate were those who were better at allowing for the pitch and roll of the ship. Using this knowledge, he designed an automated aiming mechanism that took their hit rate to over 50%.

Direct marketing practice needs a similar leap of thinking and making better use of research into customer understanding is key.

Retailers have developed loyalty scheme programmes as a potentially groundbreaking weapon in the battle for customers. 25% of all marketing spend is now on direct marketing and this is increasing. Unfortunately, because modern CRM systems mean consumers can be bombarded with communications, many are now being used to do just that. Consequently there has been an explosion in the number of messages being sent to consumers with no corresponding increase in response per communication. In fact, one could make the case that response per message is now lower than when the industry was simpler!

More retailers are therefore recognising that the depth of their customer understanding is something that they need to use to differentiate their DM programme from the competition: everything else is technology-driven and hence can be copied by competitors. This makes it the best (or only) way of achieving sustainable competitive advantage. Their customer communications can then be targeted and tailored to address individuals’ emotional and personal needs and hence add value.

The potential benefits are huge and often run to tens of millions of pounds: a combination of additional sales through improved response rates and/or reduced costs through more effective use of the CRM budget

“Why?” Research

The key to leveraging customer databases is not just analysing how different groups of customers behave but understanding why they behave in the way they do and therefore how to influence their behaviour.

This depth of understanding can only be achieved through combining transactional data with research data. Models of how attitudes and circumstances motivate customers to behave can then be developed to identify the best customers, when, where and how to talk to them and what to talk about.

For the last 5 years, RedRoute has successfully applied a structured framework (CVM) to do this. A database containing information on behaviours, attitudes and circumstances for a large sample of customers is compiled to both segment customers and to develop motivational models of behaviour. By revealing which behaviours most closely correspond (given each customer’s circumstances) to particular attitudes, their likely response to new propositions and their future needs can be predicted more accurately.

RedRoute has proved that consumer purchase behaviour can be predicted using 5 common-sense motivations:

- The Relevancy of the Product or Service on Offer
- Identification with the Brand or Company providing the service or product
- The Accessibility of the Brand or Product
- Its perceived Value for Money
- Consumers’ Confidence they will be satisfied with what they will get

The information on attitudes and circumstances comes from a survey. Having obtained a respondent’s explicit consent, their market research data is either directly cross-matched with their behaviour and characteristics from the client’s database or data fusion is used to “marry” respondents from survey research with surrogate donors from the database. Often a mixture of both approaches is used.

As well as developing motivational models of behaviour, the measures of customer behaviour that are most predictive of differing customer attitudes can be identified. This enables us to project attitudinal data accurately across the entire database without having to survey to every customer.

Supermarket Savercard

The principle of using research data to understand customer motivations was used by RedRoute to develop a segmentation of a leading supermarket’s Savercard database. The effective use of market research has played an important part in the five-year turnaround programme started four years ago to rejuvenate the supermarket group. It is perhaps no surprise then that the company was quick to recognise the value of linking research data to their card database to understand their customers much better.

The supermarket launched its Savercard customer loyalty programme nationally in the in 2004. Currently, there are some four million customers signed up for the card and more than 60% of sales are tracked by it. The resulting database provides the company with a basis to analyse shopping behaviour, fine-tune the product mix in individual stores and refine marketing initiatives.

RedRoute interviewed 2,000 shoppers at exit interviews across 32 stores. These interviews used a bespoke questionnaire to capture data across all the key dimensions and potential motivational drivers: shoppers’ reasons for choosing where they shop, attitudes to grocery shopping, attitudes to food and life, level of satisfaction with The supermarket, mix of shopping missions at The supermarket, etc. It also captured a range of information about the shopper’s circumstances that were not already held on the Savercard database (eg other store cards held and demographic details). The interviews also included a number of non-Savercard holders to provide comparisons and benchmarking for the Savercard sample.

For each individual interviewed, the responses to this survey were then linked to that individual’s transactional data on the Savercard database. Using our CVM framework, RedRoute was then able to measure and validate relationships between key attitudinal characteristics and shopping behaviours (with particular regard to the types of products purchased). In doing so, behavioural measures were determined that could be used to estimate these key attitudinal characteristics, not just for customers in the research sample, but across all four million customers on the database! Hence a customer segmentation was created for the entire database which reflected the relationships between behaviours and attitudes.

The Head of Consumer Insights for the group, recognised the value of this and commented that: “The result is a customer segmentation that we can monitor and track, together with a much greater understanding of our customer base. This is a tremendously powerful base from which to develop business applications”.

Beyond Segments

The segmentation that this approach provides is a true motivational segmentation: ie not just attitudinal or behavioural. However, the segmentation is only the starting point.

Models can be developed which, for each individual customer on the database, score the relative importance of each motivational driver. Communications can then be targeted and tailored to groups of individual customers based on their potential receptiveness to a proposition and couched in terms they find most motivating.

Often some highly predictive attitudinal or circumstance variables cannot be proxied using data already on the database. In this case the client can put mechanisms in place to obtain this information continuously and cheaply and do so just for those customers for whom the information will make a difference to effectiveness. This saves money by avoiding the need to run regular large surveys.

Other initiatives, such as gathering potential leads from contact databases, web-site referrals and so on, can be analysed and profiled using the same techniques. Their potential can be better judged and, more importantly, the investment needed in CRM capabilities to exploit them can be justified. One client, with a direct sales force and call centre operation, used the model to better predict what type of proposition would be most motivating and whether a prospect was likely to convert. This helped prioritise staff time on the best opportunities and increased conversion rates by 38%.

Whether the customer data is derived from on-line or off-line transactions, or even if it only contains quite basic information about customers, the same approach can be used.

Step-Changing Profits

Naturally a company will only wish to enhance their targeting in this way if the rewards are worthwhile. To make the communications more relevant and to target customers who are more likely to respond, the model places greater emphasis on Relevance and on targeting people with greater Brand Identification. They are less price sensitive and have potential for increased spending. More value-driven messages will be directed at price-sensitive customers who are most likely to both switch and repeat. The result is more sales-effective and more profitable marketing campaigns.

The immediate benefit, typically a step-change in response of more than 50%, generates additional profits that often exceed the required project investment by a factor of 500! Overall returns from the CRM marketing budget can be doubled and evidence provided to justify continued investment in database marketing that Finance Directors can easily accept and support.

Targeting ABC

In summary, using market research in developing one-to-one targeting models is both feasible and, if done correctly, highly productive. It requires consideration of the five key drivers of response and using the data to model the relationships between attitudes, behaviours and circumstances. In so doing it provides a new “ABC” of targeting effectiveness.

This approach tells companies how to mine their data: what to look for, what extra data to collect and what themes and messages would be most beneficial in driving response rates.

Nowadays, RedRoute often analyses cross-matched survey and customer data for clients to understand both the “What?” and “Why?” of consumer behaviour. The learnings directly impact on the client’s database marketing strategies. Such analyses will only increase in the future as more and more digital communications are used for customer contact and tracking.

For the benefit of all concerned, and for long term consumer trust, what is needed is to replace what some commentators have termed “Carpet Bombing” with a situation where we know how each communication we send fits into the life of the consumer, what they are going to think when they get it, and where the expected response is 98% rather than 2%. The benefit of taking this more considered approach will be a step-change in targeting accuracy, and an improvement in effectiveness that even Admiral Scott would be proud of.

Learning to Love Low Profit Customers

Posted by

For many years now the principle of Customer Value Management has proposed that companies should invest behind those customers who offer the highest return and ignore those with low profit potential.

The latest manifestation of this seemingly rational approach is the recent proposals from CRM gurus Don Peppers and Martha Rogers that companies should be using “Return on Customer” rather than “Return on Investment” to measure the effectiveness of their activities and allocate resources.

There is, however, a somewhat different perspective that is now gaining interest amongst a number of leading lights, based on observations of who is winning and losing in the marketplace.

Rather than ignoring low profit customers, or worse still, seeking to evict them, many people now believe that this group hold the keys to creating a successful and healthy business. “How so?” you well may ask.

The thinking behind this approach begins by recognising that individual customer profitability is the product of both fixed and variable cost elements. The fixed costs, such as running a branch network, are shared equally across all customers (or should be) and additional costs are then attracted to each account depending on the types of transactions they have. Those using the branch will cost more per transaction than those who use the equivalent telephone banking service.

If customers are managed purely on the basis of their profitability however, those using the branch will be discriminated against. A number may leave or the bank may decide to close a number of branches if they deem that the level of business it supports is insufficiently worthwhile.

If customer numbers are reduced however, this does of course mean that the fixed costs are now spread across fewer people, making even the higher profit customers less profitable than they were. Moreover, if branches are closed or other cost reduction measures introduced, the satisfaction levels of the customers who remain will also tend to fall. This increases customer defection rates and reduces loyalty.

Meanwhile, the disenfranchised customers seek alternative providers who are willing to provide services at lower cost, such as supermarkets. The fixed costs of their branch networks are already covered – probably by exactly the same people as were previously customers of the banks. They are therefore able to introduce such services and incur only the variable costs associated with doing so. Whilst they are unlikely to ever be able to provide the specialist financial services required by the really high value financial service customers their business gains from serving a mass market of lower revenue but relatively easy-to-service customers.

Therefore the Customer Value Management approach leads to a smaller but more profitable business in the short term but with a possibly shrinking pool of loyal customers and an increasing pool of dissatisfied ones. This is not a recipe for lasting business successful. Little wonder then that we now see on TV companies like Nationwide advertising a single loan rate for all customers, RBSNatWest putting back into its business many of the costs others had stripped out, and even Barclays reversing its branch closure policy and re-engineering its customer centricity.

There is also one other big gain. By learning how to service low profit customers more effectively, and doing so in a way that does not alienate either them or the vast majority of other, more profitable customers, the lessons can be used to make all of your customers more profitable. In the UK, for example, one leading high street mobile phone retailer believes that every customer is valuable – even the unprofitable ones. Their approach is to try to understand why the customer is unprofitable and to seek ways to reduce the costs of servicing such customers. By looking for new, lower cost suppliers, re-bundling activities, understanding their motivations and hence what their needs are and so forth, they find ways to make those customers profitable. Not by changing what the customer is doing but by changing the way they service them.

These new, lower cost, way of working are then re-applied across the whole customer base suddenly making the initial, seemingly low return investment into a massive win. Finding out how to raise the profitability of the lowest 20% by £5 per customer can easily end up increasing the profitability of every customer by £5.

So the next time someone puts up a four box grid with the labels “Defend”, “Develop”, “Switch” and “Ignore” and argues that your low profit customers should be left to wither on the vine, just remind them that those customers are exactly the ones who should be getting the highest attention, not the lowest. Then simply re-label their chart for them using the word “Manage” instead of “Ignore”.

Turning Nectar into Honey

Posted by

Many FMCG manufacturers are still missing out on the opportunities afforded by retailer loyalty card data. That’s the view of Steve Messenger, RedRoute’s CAO.

“Marketers do not seem to appreciate that this data gives manufacturers the chance to know exactly how loyal each of their individual consumers within that retail channel are to their brands and to their companies, not just in aggregate but person by person” he says. “But instead of using this to actively manage their level of loyalty to both the channel and the brand almost all are simply using it to deliver money-off coupons. This is a travesty.”

It is well known that the supermarkets have for sometime been targeting offers according to the consumer’s past purchases and attempting to get them to “trade-up” or try new products by offering bigger discounts to switchers. Tesco even offers to redeem other supermarket’s vouchers at some of its highly competitive stores. All this does, however, is simply encourage consumers to shop around.

According to Messenger, “Instead of rewarding them for their disloyalty by sending them money-off coupons, FMCG manufacturers could dramatically increase the returns from their marketing spend by rewarding customers for their brand and channel loyalty over time, which they can do as the data allows the retailers to track spending person-by-person.” This would work, he says, by retailers and manufacturers rewarding people for their loyalty to the manufacturer’s brands, and the channel they purchase them from, retrospectively once a year or once every six months. “It’s a clear win-win.” says Messenger. “And it is the most effective way to offer differential value to consumers as you can reward loyal customers with added value items that cost you less than a price discount whilst only offering price discounts to those that are looking for them. It works for the retailer as well because they can reduce the reward costs for their loyal customers whilst still incentivising potential store switchers.”

Such cooperation is already happening in some places but Messenger is surprised it is not commonplace. “The manufacturers and retailers have much more to learn about how to work together to get the best out of these new data sources” he concludes.

RedRoute has been combining market research with customer transactional data for many years and have uncovered many insights from this work. “We now know that there are five key dimensions to providing effective customer loyalty management. Relevancy, as popularised by Clive Humby, is one but brand identification, accessibility, value for money and expected satisfaction are just as important” says Messenger.

By using these dimensions we have been able to show clients how to actively manage consumer loyalty at the one-to-one level, not just give away coupons, and have shown that the result is far more profitable both for the manufacturer and the retailer.

Getting Emotional

Posted by

Ask anyone in DM about using attitudinal information in customer targeting and you’ll probably get one of two responses. Either “it’s a waste of time” or “what you need to do is to project attitudinal segments onto your database”. Unfortunately neither of these answers is correct. The first is mistaken and the second is misguided.

Currently, agencies that do recommend to clients that they should include attitudinal flagging within their database targeting usually suggest one of three methods. Either to buy-in flags from a commercially available attitudinal or psychographic database, like Wegener’s Real Motivations, to create one’s own attitudinal segmentation and project that onto the database, or to contact as many people on the database as you can to gather information directly from them. This is then simply appended.

Despite this range of options, many in the DM industry remain unconvinced and we believe they are right to be so. The current approaches are all akin to using a fork to spread butter – they are just not up to the job. What each of these approaches are missing is a fundamental understanding of the key “Perceptual Drivers” that people use to compare the attractiveness of the various options open to them. These are the manifestation of attitudes as they apply to a particular company’s offer.

Based on RedRoute’s work in this area for a wide range of companies including supermarkets, DIY stores, mobile phone stores, petrol stations, utility switching web-sites and others, we find that these Perceptual Drivers can be grouped under 5 key headings. Knowing what these headings are can then ensure that you ask the right questions to find out exactly what you need to know in order to optimise both the tone and the content of your one-to-one customer marketing activities. Moreover, our clients have found that using the approach increases the effectiveness of their direct marketing activity by 50% compared with traditional targeting approaches.

To understand these 5 headings and how they work, consider the DIY market as an example. The first heading is “Relevance” and in a very simple sense there will be a clearly different level of relevance for someone who is renovating their whole house compared to someone who just needs a new light bulb. Understanding the level of interest and need someone has for the category is the very first step.

The second heading is “Identification” and by this we mean how well does the customer relate to your company or brand. If they have never heard of you they will have few opinions other than what you are presenting now but in a well-known market like DIY retail, consumers already have opinions on the major players such as B&Q, Focus, Wickes and Homebase based on those company’s previous marketing activities and the stores they have seen. Most men, for example, would expect B&Q to stock a much wider range of the more sophisticated DIY products than Homebase whilst most women will feel that Homebase is a more “shopper-friendly” environment. Clearly these perceptions will impact upon their likelihood to respond to the differing marketing communications from each company.

“Accessibility” is the third, and sometimes most important, dimension. There are typically several aspects to be considered. The first is physical accessibility. If you live in London, you are unlikely to travel to Scotland just to visit a particular brand of DIY store (unless it is the only store in the country that stocks what you need!). The second is financial. In the DIY market, your disposable income will dictate to a large extent the amount that you can spend on your house and, just as crucially, whether you do it yourself or pay someone to do it for you. This will affect your spending at DIY stores. Then there is your skill level. Not everyone feels confident in doing DIY jobs so this can add another barrier. Understanding each of the key attitudinal and physical constraints that may affect someone’s ability to take advantage of your offer is vital if you are to be able to use attitudes to make a worthwhile difference to your targeting.

The fourth consideration is “Value”, namely, their attitudes to life; to making sure they get good value for money every time they shop; to marketing; to offers and discounts and so on. In the DIY market, for example, some people may well possess the need, be happy with the company and have all the skills and ability they would need to do the job. The problem for them is that life is just too short. Doing DIY would get in the way of doing other more exciting things. For others, the need to “get a good deal” would make them readily switch supplier based on price alone.

The final area we consider is “Confidence” (or “Trust”). Ticking all the boxes above may still not be enough to get a response to your offer if past experience leaves the consumer with a doubt about the level of reliability of the service they may get. In the case of DIY, if past experience shows that half the bits were missing from the last piece of flat-pack furniture you bought from the store, you will have doubts about them next time. These are personal experiences but if you know your company had a problem with a previous product then the likelihood is that any customer that bought that product from you will be less likely to buy again. Including this in your CRM strategy and reflecting it in your communications can pay handsome dividends in terms of subsequent customer re-purchase and loyalty. People can be very forgiving but these days they expect that you will remember what happened to them last time.

Using this simple check-list to make sure that you have thought through all the sorts of information you may want to know about the customer will take you a long way towards defining the questions you need to put on your survey. Once these have been gathered, a Perceptual Drivers analysis will help distil them into the key criteria customers are using to compare your offer with their needs. Example drivers for the DIY market are shown below and the differences in the importance placed on each of these factors by differing customer segments is clearly visible (a value of zero on the scale means the score equals the average amongst all customers, less than zero means less than average and more than zero means greater than average).

This chart shows, for example, that those with a high interest in DIY are less concerned about simplicity, they actually get pleasure from doing the tasks themselves, whilst those with a low interest have the opposite opinion. On product quality, this is a vital “esteem” factor for those with high interest, but is not at all concerning for those with a general level of interest. They feel that provided the product does the job, any brand of paint will do. For those with low interest, product quality is more of concern as they wish to have the reassurance of using, for example, well-branded products. They do not feel they are expert enough in this sector to “trust” using cheap or unknown ones.

Such analyses make it much easier to think how to tailor the tone and content of differing communications to the needs of differing segments in the customer database. Moreover, the level of loyalty of each segment is directly related to how well you are able to match your offer at delivering on each of these key Perceptual Drivers according to the importance they place on them. For example, compromising on the Garden Range would be a significant problem for the High Activity group shown here and would produce a significant reduction in their loyalty. By contrast, stocking only those products suited for the most ardent DIYers would be a major problem for the Low Activity customer group.

Flagging customer attitudes on the database itself also becomes much easier. Correlations between the Perceptual Needs shown above and individual customer behaviours enables customers to be scored according to the importance they are likely to place on each key driver. This enables targeting models to be developed which include such variables for customer selection. The net result, as mentioned above, is to produce much more effective marketing campaigns, with higher response rates.