It's bad form for a book like this to draw too extensively on a single source. And worse than bad form, it's potentially bad business, too – you can use a small handful of short quotes free of charge, but as your level of borrowing starts creeping up from dozens to scores to many hundreds of words the question of payment becomes more and more likely to arise.
Still, we simply can't write a chapter about the hugely important, ever-increasing and totally transformative effects of the irresistible advance of big data without drawing right up to the limit of propriety on the book that first got us all thinking about the subject – Don Peppers' and Martha Rogers' The One-to-One Future, first published way back in 1993.
Re-reading the book today, it seems more than ever one of the very most extraordinary examples of prophecy since, well, John the Baptist or even Nostradamus, and even George Gendron is starting to appear a lot less ridiculous than he did at the time.
We have no idea who Gendron is, or was, but the one thing we know is that he's the chap who turned up on the front cover of the US edition of The One-To-One Future saying: ‘This is not the book of the year or the book of the decade. It's one of the two or three most important books ever written’. At the time, this caused a bit of a stir among fans of Tolstoy, Shakespeare and Proust. But all these years later, it looks as if Gendron may have had a point. The One-to-One Future really is a fairly important book.
In essence, it argues that technology is bringing about a massive change in the whole nature of marketing – that we're moving out of an era of mass marketing in which the main aim is to sell products or services to as many customers as possible, and into an era of one-to-one marketing in which the main aim is to sell as many products and services as possible to each individual customer. On this basis, the authors say, the key measure of success must move on from share of market to share of customer.
What's truly remarkable is that the book attributes these developments to the advance of technology – even though almost all the authors' technological predictions never happened, and they missed all the important things that did.
Here are the big things to look forward to in interactive communication, as they saw them in the early 1990s:
- By the end of the decade, many major magazines will offer subscribers not only personalized advertising, but personalised editorial content as well.
- Some newspapers may also offer personalization.
- Fax machines, already found in 30% of Japanese homes, will be found in more than 50% of US households.
- By the end of the decade, airplane seatbacks will come not just with telephones, but with interactive video screens as well, connected by satellite to programming providers and catalog merchandisers.
- Microwave ovens and VCRs will respond to your spoken instructions.
- Nintendo sets will be used for homework, connecting televisions by phones to databases that provide encyclopedias, textbooks and news.
Like most cloudy crystal balls, what makes it funny is not so much that it's all completely wrong, but rather that it's all just fairly wrong. By the end of that decade, the few remaining fax machines were gathering dust in remote corners of offices. Airplane seatbacks did have video screens, but they were never used for browsing catalogs. And kids were using technology to do their homework, but the authors' enthusiasm for Nintendo machines blinded them to the potential of computers.
In fact, writing at the start of the Internet era, the authors make the occasional rather oblique reference to computers, e-mail and websites, but in an oddly tentative way as if they're really not too sure what these things are or where they're going. And of course they have nothing to say about any form of social media – their top tip for dialogue with customers is something called interactive radio.
Their tin ear for technology provides plenty of amusement, but when we stop sniggering we realize it makes the substance of their prophecies all the more impressive. Their big ideas about relationships between companies and their customers were right in almost every detail. Here, chosen not quite at random from different chapters of the book, are eight of the things they have to say about the ‘one-to-one future’.
They were wrong about the technology, but right about almost everything else. (We say ‘almost’ because, arguably, they made one other mistake that's commonly made by futurologists. They expected the future to arrive a bit more quickly than it has. The time horizon of Peppers' and Rogers' book was, by implication at least, the turning of the millennium some seven years or so after it was written. Although we still believe in the one-to-one marketing revolution they predicted, we can't deny that large chunks of it are still on the to-do list today. And that's not just in the world of financial services, which is routinely lagging behind when it comes to revolutions. It's in most of the rest of the consumer marketing economy, too.)
To test the truth of this, let's review a small selection of the inbound marketing messages received by one of your authors on the day of writing this. (The author in question is in fact Lucian – as a dedicated online shopper, he has provided the greater quantity of personal data to allow for more targeted one-to-one communications. He apologises in advance if any of his favoured retailers have a negative effect on your perceptions of his personal brand.)
Actually (Lucian adds), even more finally, on the home page of my browser here's another ad, just served this minute, for that Future of Digital Banking conference, which I would start finding a bit creepy if it wasn't for the fact that I'm in the business and know, at least roughly, how these things work.2
Apart from saying that it wasn't terribly interesting, you might raise three objections to this list. You might say:
On the other hand, the sheer quantity of this one-notch-up-from-spam material is a problem, even if only because anything genuinely well targeted and personalized is likely to be drowned out amidst all that noise. And most, if not all of the senders could have cleaned up their data with remarkable ease at any time over the years: the Audi dealer could and indeed should have known that their records are two cars out of date, Coutts should have been able to see that none of their investment e-mails has ever been opened, and after 12 years Brightwells might have been wise to check whether Lucian still has any interest at all in four-wheel drive vehicles, or in driving, or indeed is even still alive.
But if the problem among digital marketing organizations generally is that they're continuing to pump out far too much irrelevant, untargeted and unpersonalised stuff, the principal problem in financial services is still the opposite. Returning to Lucian's inbox for one last time, his last 100 marketing messages include just three from financial services firms – the Coutts market review mentioned above, an extremely boring Terms & Conditions update from Virgin Money and a personal loan offer from PayPal. This may well suggest that he spends more of his online time exploring the world of four-wheel-drive vehicles and planning trips to watch football matches than dealing with financial services, and this is certainly the case. But still, he is an active and current customer of at least 20 financial services providers of one sort or another, and has signed up as a prospect (even if only for work reasons) with at least twice as many more. Really your authors should know better than to base so much of this chapter's content on a research sample of one, but even if the one in question is very atypical indeed a clear point still comes through.
At least, on the upside, this suggests that at present financial services firms need to worry less than many about the danger of perceived creepiness – or, worse than creepiness, thoroughly disturbing intrusiveness – which is starting to reach alarming levels in other sectors. The problem is partly that consumers don't understand how marketers are able to target them so precisely, and so feel uncomfortable when, having bought, say, a pair of reading glasses or a case of Chilean wine on the Internet they're then pursued by ads for reading glasses and/or Chilean wine on almost every website they visit over the following month. Bu actually it's worse than that: even when they do have some sense of how their data is captured and used, many really don't like it very much.
The evidence suggests this is largely a generational issue, and that older people reflect their generally higher levels of discomfort with the digital world in their higher levels of anxiety about personal privacy. But, that said, research indicates overall levels of anxiety so high that younger people can't be immune. A survey among a nationally representative sample of over 2,000 people tells us that 71% don't feel comfortable sharing geo-location data, and 68% feel uncomfortable sharing data from their social media profile, while 45% say they would stop dealing with an organization altogether if they found it was using their personal data in a way they didn't feel comfortable with, and 57% say they don't trust brands to use their data responsibly. On this last point, just over half say they have received communications from firms that they believe have in fact misused their data.
The same research, incidentally, confirms two other points. Only 8% of respondents said that they understood where and how organizations use their personal data, while 31% admitted that they didn't have a clue. And among the separate sample of marketing people, the findings indicated that consumers are currently right to be suspicious. 41% of marketers say they don't understand the laws around using consumers' personal data, and only 35% say their own organizations are transparent in the ways that the collect data. And only 40% say their organizations have training in place to ensure that the current data protection rules are adhered to.
All in all, these aren't hugely auspicious findings for the two hugely important data-related legislative changes taking place at the time of writing. GDPR, imposing new constraints and controls on the use of customers' personal data, and PSD2 obliging banks to make their customers' data available to customers, and through them to third parties, could both result in significant and very positive change. But, although it's very early days, at the moment it does rather look as if industry ignorance of the former, and customers' security fears about the latter, will act as considerable obstacles to progress.
Data protection and security issues are certainly one of the items on the list of difficulties in bringing Peppers' and Rogers' vision to life, but from the fact that 25 years after the publication of their book there's still such a long way to go, you can tell that it's by no means the only one.
In fact, we're in no doubt that of all the challenges posed by the new financial services marketing, moving to a truly data-driven approach is right up there among the very greatest, not just for marketing people, but for their organizations as a whole. It may well be that the available benefits, for firms and customers alike, are the greatest, too: let's hope so, because otherwise it simply wouldn't be worth the effort.
Like many – perhaps even most – of the marketing advances described in this book, this one presents very different challenges for large, complex, long-established organizations, and for small, simple, young ones. And like many – perhaps even most – of those marketing advances, they're much easier for the latter firms to tackle.
If yours is a small, simple, young financial services firm, then arguably you should only have one significant problem in adopting a data-driven approach: you don't have much data yet. It is of course possible to acquire any amount of third-party data, and it's also possible to set about building a significant prospect database (or data lake, as it seems we now like to call it) of your own as an urgent marketing objective in its own right. But, nevertheless, you're still aiming to overcome a disadvantage compared to large, long-established firms with thousands or even millions of customers.
The US market for online investment services provides a case in point. Digital startups like Betterment and WealthFront worked extremely hard from a standing start to build their customer bases, and their efforts in attracting assets under management (AUM) of a billion dollars or more within just five years of launch were widely admired. As the size of the market for this kind of service became clearer, established giants in retail asset management like Vanguard and Charles Schwab launched similar services. With stronger brands and bigger budgets, but above all with data including millions of existing customers and prospects, they overtook the startups' five-year AUM numbers within a week or two.
Schwab and Vanguard were unusual in that they were large, successful firms with big customer and prospect bases but, as relatively young and simple businesses, they were less-than-averagely affected by many of the legacy and complexity problems that afflict so many other firms. Bearing in mind that most of the biggest firms are the result of numbers of mergers and acquisitions, and that mergers and acquisitions invariably create horrendous IT and data integration problems, it comes as no big surprise to hear that some of the biggest are still running up to a dozen separate data platforms, and several are unable to tell how many customers they actually have or what combinations of products they currently hold. For firms in this sort of situation to embrace the one-to-one data-driven future is almost unimaginably hard. Some data experts believe that in this situation, we're moving inescapably toward a two-speed future where smaller, younger businesses with less legacy and newer computers have a sustainable agility advantage, at least until they themselves become bigger and older and carry out some mergers and acquisitions of their own.
Still, in the end, even the firms faced with the biggest, hardest and most complex challenges can only tackle them in the same way as any other – one step at a time. On this basis, we held a workshop with a group of consultants, all with extensive current financial services experience, to identify the biggest issues that the industry is currently tackling. The main 10 points that emerged from this session – in no particular order – were as follows:
As well as these generic, industry-wide issues, there are specific points relevant to specific industry sectors. Among the most thought-provoking are those that relate to insurance, where it's arguable that the rapid increase in individual customer-level data effectively signals the end of insurance as a concept: by its very nature, insurance depends on the principle of pooling risk, in situations where the overall level of risk across the pool is quantifiable but not the individual level of risk run by each individual within it. When we have data that defines each individual's level of risk with ever-increasing levels of accuracy, it's difficult to see how risk-pooling as a principle can survive. At the time of writing, the industry is still maintaining moratoria in a couple of specific areas, not using genetic data in the pricing of life assurance and not making full use of flood risk data in home insurance. But it's difficult to believe that either of these policies will last for long – and meanwhile other sources of data are having an effect on life assurance underwriting (with, for example, the highly successful firm Vitality offering discounts for people maintaining high levels of fitness), and on motor insurance, particularly with telematics, the in-car ‘black boxes’ that allow individuals' driving styles and behaviours to be reflected (or controlled in the case of 10 p.m. curfews for young drivers), for better or worse, in their premiums. The impact of ever-increasing and ever-improving data on the concept of insurance underwriting is a subject beyond the scope of a book about financial services marketing, thank goodness, but it's very difficult to see how insurance as we know it can last much longer under these growing pressures.
We've come a considerable distance in this chapter without yet mentioning either Artificial Intelligence (AI) or its close relative Machine Learning (ML). Given the truly vast amounts of money being invested in developing these capabilities, and the much vaster amounts that will be required to implement them across the industry, this feels like a major oversight.
In one sense that's certainly true. In another scale-and-complexity point underestimated by Peppers and Rogers, the delivery challenges of data-driven marketing are so immense that it would be quite impossible to achieve the scale required if they ran on human brain power alone. At every point in the customer journey – but perhaps most of all at the points where the customer actually wants to interact directly with the organisation – computers with the ability to learn from their interactions and deliver an ever-more personalised service are the only possible solution. We have now reached a point where, in well-designed processes, chatbots can now do an acceptable job of maintaining the necessary dialogue with the customer (although there's little doubt that for some while to come, a bail-out to a real person will be necessary when things get complicated). As marketers, your authors aren't too embarrassed to admit that we haven't the faintest idea how the technology works, but we're happy to recognise that it works increasingly well – and that for the data-driven future ever to arrive, it's essential that it does so.
Up to a point, this kind of technological ignorance is acceptable, and indeed inevitable. It's not possible for everyone to know everything. But there are dangers that emerge when subject areas become so complex that only a very few specialists really claim to understand them – probably the all-time best-ever example being the area of derivative-based complex financial products during the 2008 crash, some of which were said to be so complicated that actually no-one understood them, or what could happen when they went wrong. At the time of writing, the data-driven marketing equivalent is the area of programmatic advertising, where it seems that hardly anyone involved at any point in the process understood that the way the system currently works, the appearance of ads for blue-chip grocery brands, charities and government publicity campaigns on sites featuring jihadist bomb-making videos is an inevitable consequence.
It would be a brave person who would predict how the current concerns in this area will develop in the future, but it seems clear that the biggest reason for the current crisis is that many buyers of digital advertising (a) haven't really understood what they've been buying, and (b) haven't really understood the metrics used by the vendors to give an account of its performance. This obviously creates the potential for some very nasty surprises when the truth belatedly dawns, whether it's the discovery that big blue-chip organizations have been buying space from racist, sexist and even terrorist organizations, or the discovery that video viewing figures involve a degree of statistical manipulation that makes them effectively meaningless.
As the world of digital media becomes ever more complex and labyrinthine, it's difficult to see how we can escape more problems like these. A few large and sophisticated companies may be able to afford to hire specialists with responsibility for monitoring actual performance, but most with have to rely on advice from third-party consultants, agencies and advisers. This, however, in itself raises a pair of potentially awkward questions: first whether these third parties fully understand what's going on, and second, even if so, whether the buyers of their services can rely on their complete integrity.
There's one more big point to make in this chapter. Every now and then a technology comes along that really does wipe out whatever went before it. We all know about the classic business-school case history about the effect of the transcontinental US railroad on the previous era's wagon trains. And although it took a while, within a hundred years or so the printing press knocked the bottom out of the medieval manuscript-illustrating business. But on the whole, this isn't what happens. On the whole, new technologies find a new place alongside the previous versions. Cinema didn't wipe out theatre. Photography didn't mean the end of painting. Electric music co-exists with orchestral. People still wrote letters after they acquired telephones, and still make phone calls after they've acquired computers. The balance between the old and the new can vary, and is hard to predict: sometimes the new becomes dominant, sometimes the old proves amazingly resilient. (Sometimes, also, the old evolves to make room for the new. There are, for example, as many horses in the UK today as there were in the heyday or indeed hayday of horse-drawn transport, but they're not pulling carriages, buses or taxis.)
This lengthy preamble is by way of an introduction to the point that data-driven marketing seems most unlikely to replace the previous, more broadly targeted kind. In part, this is certainly because it's so fearsomely difficult to do well, but it's for other reasons, too:
All of this, of course, only adds to the pressure on marketers. It isn't just a question of learning a large number of new skills: it's also a question of retaining most if not all of the existing skills, and then finding a way to bring both sets together.
On the upside, the prize for doing so successfully is a very valuable one. It may well be that it's through the gradual emergence of Peppers' and Rogers' one-to-one future that marketers finally achieve a new level of authority in their businesses.
3.137.164.24