The decision by Bundeskartellamt, Germany’s competition authority, to rule against Facebook’s plans to merge user data from Messenger, WhatsApp and Instagram is hugely significant and perhaps the first step towards the separation of data from the digital monopolies.
I have been working on customer insight data for more than 30 years. Although in many ways my work has remained the same, the big change is the arrival of digital analytics – using data that arises from purely digital interactions to help you make your marketing decisions.
This data takes two main forms – one arising from the world of mobile telecommunications and one from browsing on the web. These two are, of course, converging as the use of mobile apps begins to dominate the conventional use of web browsing on computers.
As a result of my work in this area, I have decided to update my book on Consumer Insight, which is now badly dated – though still a useful guide to the world of database marketing for the market researchers for whom it was intended (it was published in Kogan Page’s series of books for the UK’s Market Research Society).
I’m working on this project with Tom Breur, editor of the Journal of Marketing Analytics (I’m on the editorial advisory board of the journal). His day job is VP Data Analytics at Cengage Learning, Boston, MA – Cengage is one of the largest educational publishers in the world. Tom’s expertise covers statistics, data mining and analytics, model building, MIS systems, market research and data driven decision support. Tom is from the Netherlands, one of the great sources of expertise and good practice in digital and direct marketing. He was previously director of Business Intelligence at ING Card, one of Europe’s leading credit card suppliers and before that director of Commercial Research at ING, one of the Netherlands’ leading and most innovative banks. Tom is a clinical and consumer psychologist by training, so he and I (an economist by training) will be able to cook up a nice brew of concepts and practice. Our target date for publication is early 2015.
We have started work on the ‘deep structure’ of our book. We had already decided not to create a hostage to fortune by covering the details of the techniques used to create marketing insight, whether from big customer databases, transactions records or web usage. This would have meant that the book would be out of date within a year. Our focus is on the capabilities that companies need.
We started with a listing of capabilities, from how insight management is planned through to how it is delivered and used. This led us to discuss why, when and where companies need different capabilities. We soon realised that if you ask how strong a particular capability should be and how it should be used, the answer is “It depends!” That answer is not very helpful. What does it depend upon? The answer: “Lots of complex, interacting factors!” That also isn’t very helpful. We know we must find a way to express the complexity of the world – of customers, markets, competitors, products, channels, technology, economics, social trends, regulatory and legal constraints and environmental pressures. We reviewed several ways of thinking about this and have produced the ideas listed below. We’d love to know what you think of them. Remember, though, that we are focusing on marketing insight (including sales and service). There are many other types of insight a business needs, eg. financial, operational, HR. Although many of these interact in some way with marketing insight, it’s the latter we are focused on.
First, here are a few general points on customer insight we came up with:
- Insight is not just or even primarily about averages, but also about distributions and about the detection of the early signs of change by identifying small variations in factors that are already known about, or the emergence of new factors.
- Insight is competitive – the key requirement is to be better than your peer competitors in mature markets, and to use insight to identify and find ways to defend against new competitors, entering with new ways of doing business and new value propositions, who may see the market in different ways.
- Insight can support strategic, operational or tactical decisions – strategic ones that involve significant changes in marketing, eg. capturing big opportunities or warding off severe threats, operational – eg. finding ways to improve how existing policies are delivered, or tactical – eg. optimising within significant delivery methods.
- The difference between the kind of insight derived from analysing massive databases of numbers or facts (common in classic CRM as practised by financial services firms, telcos, media, airlines) and optimising delivery by purely digital fine-tuning, eg. web design or search or use of affiliates or changing attribution models, looks large, but these two types of insight are converging as tools emerge for bringing them together.
- Big data – we must accept the term – is changing all forms of insight, as are the tools for developing insight from big data. Both data and tools are developing rapidly, helping in both on and off-line worlds
- The building of insight capabilities can be interpreted as a competitive game, in which firms can outbid each other for data, skills and other resources. It is definitely not a zero-sum game.
- Perfect insight is impossible to achieve – it is more productive to see the battle for improving insight as a battle to partially disperse what Clausewitz called ‘the fog of war’.
Now for the ideas, which we express below as a series of propositions.
The customer insight capabilities that a company needs depend mainly on:
- The stage of evolution of the business and its size – from being a new (probably small) business in a market niche, to a global giant
- Its competitive situation from being a sole supplier for a brilliant and differentiated (and possibly protected by patent or trade mark) concept, through being one of a group of similarly sized companies competing head to head with broadly similar products, to being a company defending itself from attackers with different propositions, or one of the attackers
- The rate of change in its markets, and the social, economic and other determinants of that change, and how that is reflected in the behaviour of buyers
- How digital a market is, and how digital techniques are affecting how buyers and suppliers behave
- Whose needs you are focusing on – marketing directors, brand/product managers, channel managers, pricing managers, marketing communications managers, sales managers, or customer service managers, or of course insight/market research managers. It also depends where you sit in the insight supply/buy chain – the buyer, the user, the supplier of data, of insight diagnostics, etc.
There is one other area we need to explore, and that’s the impact of web analytics on conventional or as we call it ‘classic’ database marketing, where analytics has reached very advanced levels, moving into areas such as complex and constantly changing real-time segmentation, highly customised next-best actions. The customer databases being managed run into tens of millions of named and known customers, with data about them being collected tens or even hundreds of times a day.
Web analytics is different. The name of the customer may not be known (though it often is), but the history of the device is, provided cookies have been enabled. The location of the device may also be known. Analytics in its most advanced form is seen in the activities of affiliates, who can help serve and convert to leads advertisements that are very specifically targeted. As data protection authorities become more extreme in their views, as in the EU, might it be better to throw away the idea that we need a long history of identified customers to do our marketing? Isn’t it simpler and better to capitalise on what we can learn about the customer at the moment of contact, even if we can’t at first know their name? If they are interested enough and move far enough down the sales funnel, we can pick up the customer data we need.
We’d like to hear if you have ideas about these thoughts – all contributions will be gratefully acknowledged. We’re particularly interested in good examples of the integration of classic CRM data with true digital data to produce good insights, including data which cannot clearly be categorised as one or the other. However, we’re also interested in the replacement of classic CRM data by web/app data.
In all the above, I haven’t covered topics such as what we can learn from social media, how the new analytics helps with every aspect of marketing, such as product and channel management, advertising and PR. We’ll be covering them all.
Please register below to unlock this article.
An email will be sent to you with your membership details.