One of the central philosophies that Naren espouses is that all customers are not equal – and should not be treated equally. The profitable ones need to be pampered – and the non-profitable ones to be neglected. To start this process of iniquity, one needs to start with data.The data should help answer questions like – which customers are most valuable to me? Am I treating them well?

Kellogg School did a survey of 252 large corporates in the US – and found that leave data apart, they did not even know what to measure. So what is the relevant customer data? Here is a laundry list:

Return on Marketing Investment: Return on marketing investment (ROMI) is the contribution attributable to marketing (net of marketing spending), divided by the marketing ‘invested’ or risked. RoMI makes a lot of sense for an industry like airlines, where most costs are fixed. Additional revenue goes directly to the bottomline.

Customer Life time Value: Customer lifetime value (CLV) can also be defined as the dollar value of a customer relationship, based on the present value of the projected future cash flows from the customer relationship. Customer lifetime value is an important concept in that it encourages firms to shift their focus from quarterly profits to the long-term health of their customer relationships. Customer lifetime value is an important number because it represents an upper limit on spending to acquire new customers. This may require estimates of probabilities of retaining customers.

Net Promoter score: The Net Promoter Score, or NPS, is based on the fundamental perspective that every company’s customers can be divided into three categories: Promoters, Passives, and Detractors. By asking one simple question — How likely is it that you would recommend [your company] to a friend or colleague? — you can track these groups and get a clear measure of your company’s performance through your customers’ eyes. Customers respond on a 0-to-10 point rating scale and are categorized as follows:

Promoters  (score 9-10) are loyal enthusiasts who will keep buying and refer others, fueling growth.

Passives     (score 7-8) are satisfied but unenthusiastic customers who are vulnerable to competitive offerings.

Detractors       (score 0-6) are unhappy customers who can damage your brand and impede growth through negative word-of-mouth.

To calculate your company’s NPS, take the percentage of customers who are Promoters and subtract the percentage who are Detractors.

A balanced score card: should have less than 10 parameters, to be able to get managerial attention. Parameters can be –

Strategic measures – LTV, Brand, Test drive

Tactical – Sales, Lead conversion

Operational – Take rate, cost to serve, RoMI, ad spend

Segmentation and Positioning

One of the oft neglected fields in marketing is segmentation – deciding who your target audience is. Knowing that the next question should be what are the attributes common to your segment?

All customers are not equal. The best way to separate them out is based on Life time value. If we plot profit vs life time value on the XY axis, strategies to be followed are:

1. Hi LTV, Lo profit – Manage costs

2. Lo LTV, Hi profit – Maintain relationships

3. Lo LTV, Lo profit – Let go or manage service levels or migrate

4. Hi LTV, Hi profit – Grow and retain

Positioning: Take your sales material. If you replace your name with your competitor’s, will the customer notice? 

Pilots and Surveys

When a professor at Kellogg’s was once asked about the errors inherent in sampling he replied. Why should doctors be satisfied with a 5 cc blood sample for your diagnosis. Why not drain your entire blood, so that they can be sure? Sampling theory dictates that samples even as small as 30 can give you a fairly reliable idea of how the population is going to behave. This is of relevance in evaluating a lot of marketing campaign alternatives before deciding which one will be required for a national level roll out. Examples where such pilots are useful could be to decide what creatives work better. Another more popular usage is for keyword selection in Google Ad word campaigns. By playing around with keywords you can find out which ones are working out better.

For example a company that pays for keyword ‘performance management’ could find that there are very few clicks and a lot of riff-raff with a campaign run around these words. The issue over here is the very general term – what kind of performance management are you looking at – the range extends from Viagra to HR! Keywords were changed – and traffic improved, but Google Analytics informed you that the bounce rate was high. The possible issue over here was content which was not compelling. So the page was redone – and then the company was finally ready for a national roll out. Such a strategy also makes it easier for a business case to be made to top management for a campaign.

Brand Building Nuggets

• Insights come from qualitative research (read conversation), Validation from quantitative.
• How does one build a brand? By continuously checking for customer happiness – and then trying to cross – sell or up – sell. Engaging with media of all sorts, social included, is important because these guys are the Narad Munis. They basically are info brokers – spreading good practices across the customers and industry.
• Marketing performers believe in NIHITO – ‘Nothing Important Happens In The Office.’ So attending seminars and trade fairs should be de-rigueur for the IT marketing guy. When you meet prospects at such places, engage them in conversations and listen. Ask them – ‘What keeps you up at night? What are one or two important things that your CEO wants to do? What would you like me (the vendor) to do?’
• One of good ways of measuring Brand health is Unaided Recall. Ask your prospects (not your customers) – ‘When I say entrance exams, which brands come to mind?’

Net Marketing Nuggets

Take an example of the clichéd ‘Half of marketing dollars are wasted. I don’t know which half.’ Some of the internet marketing companies are even insinuating new meanings ‘Half of marketing dollars are wasted. I know which half. TV’

• Falling storage costs have started making big data more ubiquitous. In big data, transactional data is most important. On a relative scale, social data is less useful.
• Marketing qualified leads – are generated through stuff like Webinars, videos and white papers.
• For webinars, have reminders 1 week, 1 day and 1 hour before the start. If 1 week reminder is not working out well, change the email subject and try.
• Take rate – number of offers made, to those taken. What should be take rate for email offers? 10% + is the industry norm.
• Test drives are great customer acquisition tools. A test drive is a pre-test which is offered prior to purchase. This works best for ‘experience’ goods.
• Bounce rate could also be a result of navigation issues.
• 1 impression on Youtube is equivalent to 36 page views. Also the views take on an upward spiral, as most viewers decide to view based on the views already made.
• In the US data.com provides lots of ways to segment the data that they have. You can get free access to data.com by bartering information of your own.
• Excel is a simple tool that you can use for data analysis at a small company level.
• An interesting marketing tool. You get a postcard which says – ‘Hey Sumit Kulkarni, We have been following your progress on our course. And we have a great offer just for you – please check it out on www.hitbullseye.com/sumitkulkarni’ Wouldn’t Sumit be pleased to at least check out this offer?

Developing the metrics of a campaign.

Incremental Revenue desired will decide number of customers to be got. (This requires knowing average deal size per customer). Knowing the lead conversion rate, this will translate into the number of leads required. Ideally, this should be broken up into monthly targets. This can also be split channel-wise. The metrics like deal size, lead conversion rate should not be taken as fixed. Benchmark with industry to find out how you are performing. A company which has a low conversion rate nees to train its sales people – maybe they require better call scripts, competitive decks, displacement stories and deal support. If the raw lead inflow is less, then the typical cause is low brand awareness. Marketing underperformers spend proportionately more on lead generation because they are targeting everybody and hence their marketing dollars are spread more thinly.

Case Study: Carnival Cruises

Carnival has more than 23% of the US market for cruises. Against the industry norm of 30%, Carnival has 50% of its customers who are first timers. This implies two things – a higher spend on lead generation (repeat customers have lower acquisition costs) and lower revenues – because a repeat customer on a cruise has an average spend which is $ 3500 Vs $ 3,000 for a first timer. Repeat rates dropped even more rapidly after the second cruise. Perception seemed to be – This is not fun.

Data analysis pointed out that most repeat customers are empty-nesters. Makes sense because it is difficult for parents to keep their kids occupied for 5 days on a floating island.. So Carnival decided to offer special packages to get back the empty nesters. Destinations like the Bahamas, which were preferred by their target, were launched more often. Special decks were reserved as Couples-only No-Kids areas, so that our empty nesters could enjoy their solitudes without interruptions. Targetted offers were made. These high value customers were pampered further by ensuring Priority Lines when it came to queues inside the ship. As a result even though the number of passengers carried remained the same, the profitability went north resulting from a higher mix of repeat passengers.

Comments on the article by my friend, Manoj Thomas, Prof. of Mktg at Cornell Univerity

There are some very interesting points in Mr. Patil’s talk.  I suspect that he would have been a good speaker because he definitely knows how to coherently organize his message points.

But there are some overgeneralizations and somewhat superficial insights about the use of data in decision making. For example, “Kellogg School did a survey of 252 large corporates in the US – and found that leave data apart, they did not even know what to measure.” That’s an overstatement. Companies that care about data know what to measure and how to model to optimize their marketing mix. In fact, companies such as American Express have been using CRM procedures for a long time and they regularly use predictive modeling to target customers. CRM has always been about using transactional data to predict behavior and most B-schools have been teaching CRM for over 10 years now. P&G, Unilever, Nestle, J&J etc. have been using scanner panel data for decades now. However, there is a tendency in the popular press and amongst practitioners, particularly those from Silicon Valley, to present data or ‘big data’ as the new revolution in marketing. When P&G, Unilever and all other clients use 100s of thousands of transactional data for marketing ROI computations, were they not using big data? The difference is that now big data is more ubiquitous. With the advent of online transactions every online company has access to browsing, clicking and transaction data. All you need is a free Google Analytics account. This ubiquity has made big data the new fad.  Now Silicon Valley people are talking about the same big data applications that the traditional marketing companies were employing for long.  They are talking about data, particularly big data, as if it is a radically new concept. I don’t mind this misrepresentation if it leads to more scientific decision making in companies.

There is one concern, though. Like all fads, this ‘big data’ fad could divert attention from more fundamental issues at hand. Specifically, the usefulness of transaction and browsing data is a bit overrated. Don’t misunderstand me; they are useful. And as Mr. Patil said, transactional data are more useful than social data in predicting behavior. But people do not understand the one big limitation of such data: they are historic in nature. They are about how consumers behaved in the past. Historic data are useful for fine tuning your marketing mix elements and in segmenting and targeting. But such data cannot inform you about new breakthroughs innovations. It cannot guide you to the next big idea. Was Steve Jobs’ idea of a more functional and better designed portable music player based on historic data? Was Facebook based on historic data? Was the idea of online retailing pioneered by Amazon based on historic data? Break through consumer insights can come only from experimentation not from looking at historic data. Experimentation allows us to test new ideas. Most people do not understand the power of experimentation. Some do. And those who do are gainfully employing experimentation to test powerful consumer insights. Here is one such company specializing in online experimentation: https://whichtestwon.com/. This website allows you to compare your predictions about online consumer behaviors with actual experimental results. Such testing (sometime referred to as A/B testing) is forward looking. Using such an approach, one can test novel ideas that are completely new to the world.