For Your Eyes Only - the three levels of customer behavior based data

I had a lot of good discussion based on last week's blog story about our digital leash. I discussed how classic 20th century marketing research data concepts like 'demographic' data is becoming useless when compared with far more powerful 21st century behavior based data, collected from digital channels. And I argued that mobile is by far the most powerful of those digital means, in capturing behavior based data.

DEMOGRAPHICS VS BEHAVIOR

To be clear, demographics were a great data collection purpose in the last century when marketing was crude, the targeting very weak and we used mass media broadcasting 'propaganda' to try to reach our target demographic audiences. That was the best we could do in an analog age of print, cinema, radio & TV. WIth digital media (starting with the 6th mass medium of the internet) we go early steps into capturing actual behavior based data, and with this, suddenly the ability to do far more powerful targting. I explained this two weeks ago in my Amazon book recommendation 'advertising' blog. TV or radio or newspapers can never ever hope to get the accuracy and power of what Amazon does routinely in making recommendations to us. The ads are so powerful, almost nobody thinks of Amazon book recommendations as 'ads' and nobody would want to skip them or avoid them. In fact we hope to see that in all aspects of our commercial life - why can't we have recommendations for x or y like we have for Amazon in books..

Demographics makes assumptions on our demographic data - and massive errors that come from this generalization and stereotyping. But Amazon is based on actual consumption (of us, and then mapped against the patterns of millions of Amazon consumers across the planet). Before we could measure individual use, it made sense to try to 'group' customers based on some mechanics like demograpics (and in business, by industry classification codes, what is essentially demographics for B2B use) and then psychographics - barely more than the same. Demographics would look at a 42 year old married mother two kids and make the assumption that she's interested in 'mother' and 'home' types of things, not interested in 'youth' things like partying and going out at night. Now, while her demographic profile honestly can be 'married' and she may technically be so - what if she's going through a divorce (which won't take effect for many months) and is suddenly experiencing the dating scene, looking for a new boy friend. She's behaving like a teenager or young single adult. If we use demographics, we make massive mistakes. If we measure behavior, we get the reality.|

MOBILE BRINGS THREE LEVELS 

On digital media, like the internet (and this includes of course any 'real internet' 6th mass medium consumption on smartphones ie the 7th mass media platform of mobile telecoms) - we can get to the first level of consumer behavior data - what we consume. On the internet this consumption info is almost always only limited to individual merchants (ie Amazon can't tell me what I like on the BBC for example, the Guardian can't give me recommendations on Disney etc). But like we see on Amazon, the recommendation engine is a huge leap forward in customer insights. Amazon would not (should not) bother to collect demographic data on its customers, because behavior is so much more poweful and actionable and accurate data. Obviously any mobile data service consumption as at least as good as data, as any internet consumption would be. On mobile it can be 'apps' like fashionable now on smartphones, or it can be the 'real internet' or it can be the 'mobile internet' (which is not the same as the 'real internet' on a phone) or it can be other digital services and solutions and content from multimedia content delivered via MMS messages to basic news headlines via SMS to music such as ringing tones, various ads etc.

CONSUMPTION DATA IS BASIC LEVEL

The first level (basic level if you will) of consumer behavior data in digital is the 'Consumption'. For most in marketing, this is at first glance 'the ultimate' data, isn't it. Thats the 'holy grail' isn't it. If we know Tomi consumes Pepsi and not Coke, this is VERY powerful marketing data, isn't it. Are you Mr Ahonen now suggesting something 'beyond' this, something... even better?. 

Yes I am. Looking at consumption, is like primitive humans looking at the sky and thinking the world ends after the clouds and the stars and sun and moon are the limit of the universe, that revolve around us (as it seemed from that vantage point). Obviously we now know there are stars beyond what the eye an see, and many of what look like stars, are in fact galaxies consisting of billions of further stars, countless of which have their own planets and moons etc... Yes, that is the degree of sophistication we are talking about. To focus on consumption is the earliest 'cave man' level of this marketing insight haha..

COMMUNICATION DATA RELATED TO CONSUMPTION IS FAR MORE POWERFUL

The second level of Customer Behavior market researh data comes with Communication. Previously with any communication technology and media consumption technology, there was no way to connect the two. We had our letters and postcards, our telegrams and (fixed landline) telephones. None of these delivered media content (to any significant degree to mass market consumers). And we had our books, magazines, newspapers, cinema, radio, recordings and TV. None of those had a communciation channel.

That changed with the 6th mass medium of the internet, which started life not as a media channel, but as a communication channel (originally with email). The internet was communication first, and media later. And now that the internet has figured out a way to combine the two - we get 'social media' ie Twitter, Facebook, YouTube etc - and we suddenly get 'inherent' ways to combine communication with media consumption. What do I mean? I mean that when we combine consumption with communication, we get recommendations, we send links to interesting media pages - embedded into our emails. And we offer comment threads in media pages.

If you think 'social media' or 'digital communities' or 'user-generated content' or 'web 2.0" is 'big' so far - we have barely scratched the surface! Five years ago YouTube and Facebook were nowhere. Twitter is essentially a two year-old phenomenon. CNN's i-Report is 3 years old, etc. We have barely started on this journey of the 'convergence' of consumption and communication.

So to customer data. What do gain from the communication 'dimension' to our consumption? We gain all that we associate with 'viral' marketing. We can find out who is talking about our brand, good or bad. Who sends links to our site, who is generating buzz about us. It does not 'preclude' or 'invalidate' the consumption data, but the 'communication' is far more powerful for anyone in marketing research. One person bought our brand, did not talk about it. Ok. Another bought our brand and talked about it. Better. A third person never bought our brand but likes it and talks about it - wow. This kind of real customer insight can NEVER be captured in 'consumption' data - because this customer has not (yet) bought our product but clearly loves it (or hates it..)

Where can we do this, capture, this, utilize this? This is 'cutting edge' market research data, and for the most part we can find it in separate discussion channels like Twitter. Facebook, blogs etc. So while both exist on the internet - we buy a book on Amazon, we then mention it on Twitter. Mostly the two are 'co-existing' on the internet but not part of the same consumption or communciation. It can be - we can perhaps gather it to our own branded website, perhaps a 'fan club' etc. We have paid services on social networking sites like the massively multiplayer farming games that are the big rage on Facebook for exmaple. Any smart brand in digital, and those consultants and experts and agencies advising the brands - will want this 'dialog' with the consumers. It is central to 'engagement marketing' which moves the advertising/market communications beyond the mass market 'propaganda' preaching of the past century.

But also understand, that the 6th mass medium internet is incapable as a platform to deliver this type of insights across all services and all consumers. We can have the dialog and capture it, but only like Amazon and Facebook, we can do it within one brand or one channel like Twitter or YouTube at a time. Not across all users and all services. On mobile we have for the first time the opportunity to capture all consumption and communciation data, potentially (when currently collected by the operator/carrier, subject to permissions etc..).

THIRD DIMENSION IS CONTEXT

This is part of what hit me when we discussed my blog from last week. We are now witnessing the birth of yet a deeper level of insight. Understand the power of this. Amazon recommendation engine is inherently more powerful than TV or radio ads in selling books (or DVDs, CDs etc). Consumption data is more powerful than demographics or psychographics. Then we have communication of consumption - far more powerful, because now we see true influence (potentially, when well data-mined and with good data). Now we have something even more powerful. Again I am not suggesting to abandon consumption data or communication data (but I am saying throw away demographic data, obviously, as obsolete, not just useless, but counter-productive data!)

The third dimension is the 'Context' of consumption. That was at the heart of the examples I talked about last week. When we will gain the ability to collect accurately the 'social context' of our consumption - and various other 'context' elements such as time, location, proximity - we get FAR more powerful tools. Please don't mis-undestand me. I do not mean 'location-based advertsing - which is a very very bad and discredited idea. I mean like in my blog, what of the patterns of where I am, in relation to others. How I move. But also time, proximity, and yes, communication and consumption.

Here is a perfect example, valid today. Your mission is to find real - loyal - viewers of American Idol. Find actual contact info for as many of them as is humanly and technically possible today, with any means that exist.

You could set up a Facebook page. It won't get anywhere near all loyal viewers. You'd be lucky to get 10%. You could measure those who voted for American Idol - yes this is far better, perhaps a third of the audience. You could try various 'traditional' market research tools, interviews, the Nielsen ratings system etc. Nowhere near the chance to find the total audience. No way. And trying to all 'every household' in market research surveys would be astronomically expensive. You'd end up with only thousands of respondents and then like in decades gone past, you'd have to make statistical estimates on this 'sample' data. No, I mean real actual viewers of American Idol, just like Amazon has actual buyers of my books..

ONLY MOBILE GETS CONTEXTUAL DATA

But what if you took actual SMS text messaging traffic generated during the broadcasts of the show? 78% of the owners of mobile phones worldwide are active users of SMS text messaging. Even in America its past 2/3 of the cellphone ownership. And 91% of mobile phone owners keep the phone within arm's reach 24 hours a day, so we have our phone always with us, when we watch TV. Many - not all but many - who watch a given TV show will like to share that experience with close friends. This is a growing global trend. And when do we discuss our fave show - it used to be we called our friends after the show. Now its more and more likely we send SMS text messages to our friends during the show. Any kind of TV, whether its a soap opera or the big game of tonight or something on Letterman or the news. The sending of SMS during TV shows is even more so with reality TV shows that feature TV voting, like American Idol. If we measured from the phone networks, every consumer who has a texting 'pattern' to send (and receive) within a given 'circle of friends' (ie a digital community) messages during the TV show, we find a distinct group. When we extend this to see who then communicates with others, and do this nationally, we'll find millions who are part of this group.

I am not suggesting this will get us 'all' viewers of any TV show. I am suggesting it gets us the biggest actual personal identification of an audience. In the case of reality TV, roughly speaking we'd probably find half the total audience this way. FAR FAR FAR more than by any other means available.

Note that we did not need to 'read' the text messages for 'content' (like say Gmail does). We didn't need individual members of our audience capture, to 'consume' the content (ie vote on the show). We 'caught them' based on their context of consumption - who they shared with, when mapped against time and communication. We didn't need location information. Now we get easily over half of the actual total viewers in any country, who truly are loyal passionate fans of the Idols show. Whether you voted for the show or not...

Undrestand how powerful this is. On Amazon, you have to be online to be 'captured'. We get you from watching TV my way. On Amazon you have to consume 'their' content - buy a book or look at a book page. On mobile we capture - television viewers. Yes, we can capture also mobile content consumers (who used a phone for Twitter or Facebook or YouTube or downloads a ringing tone or goes to the Weather channel's mollie page). Amazon and the internet cannot get us audiences on other media - only mobile can do that! We can use mobile to capture an audience on TV (and if done cleverly, audiences on radio, the cinema, newsappers, magazines etc)

This is super-power 'bleeding edge' marketing research. This is the 'future' of marketing insights. This is where the big money and big commercial opportunities will be made this decade. Here is the real power where we need to deploy and develop in data mining. NOT to seek demographic data. That is pointless. Its as useless, obsolete info, as is the study into the efficiency of carrier pidgeons as communications tools today haha..

WINDOW IS NOT OPEN FOREVER

Now, the data today resides only in mobile phone networks (we can't get this data from fixed landline networks because the fixed landline phone is not personal - the whole family uses it - or more likely than not, the whole family today all refuse to answer the ringing of the home landline phone as they know the call is not for them personally haha). Same is true of the internet, because of firewalls, shared computers, deleted cookies etc, the sixth medium internet cannot deliver the personal data of consumption, communication and context - on a national level obviously - as we only can get from mobile phone 7th mass medium consumption.

So, today - 2010 - the handset maker (Nokia, Samsung, Apple, Blackberry) does not have this ability. The internet brands (Google, Yahoo, Amazon, Facebook) do not have this ability. The media houses (BBC, CNN, Sisney etc) do not have this ability. The banks and credit cards and loyalty card schemes do not have this. The only entity that can gather the full picture of Consumption. Communication and Context today - is the mobile operator, and in advanced markets like Scandinavia and the Baltics, Japan, Singapore etc, the operators/carriers already are deploying the technology to do this deep level of data mining. But they have almost no tools or skills or abilities to monetize this opportunity. Why? Well, a big part of that is because mobile is so far ahead of the curve that most top-trained 'Ph D' level brand name marketing professionals and gurus still talk about the supposed value of demographics and psychographics and have not been brought up to the speed of how much more we can get out of mobile.

WILL EVOLVE AND IMPROVE FAST

But marketing is a global competence and adapting and growing fast. The giants like Google and Apple and Nokia want to get into this space. The big mobile operators like Vodafone, China Mobile, Telefonica etc are slow to capitalize on it but are trying to. And the various ad agencies from the Ogilvys and Saatchis and McCanns are hungrily studying this space. Today that data collection is at the mobile operator exclusively, through essentially its billing system, but that will not last. There are too many 'outsiders' who are seeking alternate ways to this info. The smartphone revolution is a big step into that direction. Soon we will get at least part of the data off the phone, bypassing the operator billing data. 

So one, this is a blog to wake up the marketing, advertising, media, social networking and market research companies, to understand how much more data we'll be able to get to, when we abandon obsolete demographics, start with Consumption, and add Communication and Context.

But secondly, this is a warning blog to the operators/carriers. Don't fall asleep on this issue. You have to move NOW - because this ability is not going to remain exclusively yours. The market 'window of opportunity' will not remain open forever. It will close. You have to set up your ability to educate the marketing industries in your markets, provide accurate, actionable info into your customer base, and help grow this market - which will be measured in hundreds of billions by the end of the decade - or else you will be bypassed. 

I have more bubbling in my head about this topic - I will also develop a few concrete examples for our readers to make more concrete these examples. This may be a bit too 'theoretical' as a blog for now, but I want to get this thought out there, to add to the discussion from last week. Stay tuned, and please do leave comments if this triggers thoughts.

Comments

[@SannaJanonkorva]

Hei Tomi,

Nice blog entry - as much as I'd like to, I can't quite agree especially with the 3rd level "context" fully though.

Still lot of the context and communication takes place outside the mobile sphere (think about chatting on your computer while e.g. watching TV - not everyone is necessarily using the SMS channel only) so you cannot capture all the necessary data via the mobile operator channel.

Nor do I believe that time & and the fact that someone has been communicating in that very moment necessary provides the missing link to particular content in another channel - maybe there is a high probability for that but no absolute certainty (statistics please proove me wrong).

NOnetheless, I believe that there are more and more possibilities for mobile data mining and clever MNO's will improve their capabilities in it - by investing into xtract products (if that's what you suggest? :)

But unfortunately its not quite that simple I am afraid. Consumer privacy groups will certainly have their say too.

In any case it will be interesting to see to what this develops into. But I guess it will still be a long shot.

[Vasco Duarte]

Great post. Really mind opening. What you suggest about the power to reach consumers is very important, but the MNO's need to create the opportunity for consumption not just use the data.

Take this example, if I am browsing for music in a carrier's store I am open to "suggestions"/ads. But I don't want to get the same suggestion via MMS when I'm watching the game in the evening.

Amazon's recommendations work because they are in the context of a search for product that they try to convert into intention to buy.

What are carriers doing today to create this "opportunity for consumption"? I don't see them doing much...

[Tomi T Ahonen]

Hi Sanna and Vasco

Thank you for the comments, will respond to both individually

Sanna - good comments as always and we mostly agree. To start with, like you say, there are privacy concerns - and I am adamant about it, that all mobile advertising has to be permission-based. If the mobile operators were wise, they would totally eliminate any spam banner ads, spam SMS etc and make sure only permission-based ads transmit in their networks (as the major industry bodies have recommended literally for a decade now). But yeah, I totally feel it has to be opt-in, which addresses most of the privacy issues. If its used for snooping, that is hideously bad (as we've recently seen for example with mobile operators where management used phone records to snoop about their management teams - the hideous mess it made in Finland and now a little while ago in Germany - that led to the press etc covering the scandals... idiots!)

You make a great point about context happening also on other media, in particular on the PC/internet. Yes, theoretically we might be able to capture that info from computers - except, that even if you and I happen to have a personal PC that nobody else uses, that is not the norm on the internet. In fact most PCs are shared or only semi-personal. Almost all work-related PCs are owned by the employer and they have IT departments that manage them, install firewalls, read emails etc. And of home PCs, a good part of those are shared among family members. So right off the top, even though in many cases a PC may be (or may seem) personal, it is not. Whereas the mobile phone is not shared (except in the very poor areas like parts of Africa where a family gets one phone handset, which itself then has multiple accounts on it for hte husband, wife, kids etc)

So if that context is to be 'commercialized' and 'monetized' and we want to do it beyond one service provider (like Amazon) then mobile is the only way. Its not easy and few companies are yet able to do this - but some are - and very few have figured out how to commercialize the opportunity (so far), but this to me is the obvious direction where we will go. We already get consumption info for all digital consumption (by vendor) and will start to get more out of that through various means. We already collect near-perfect communication data on mobile networks today (but mostly not yet connecting those dots). Context is logically the next step.

But I agree with you, its not reality today and it will take time, and not all will even figure out how to do it haha.. And yes, Xtract is a good example of companies who have commercial products in this space and already a proven track record to capture such info as the user's profile in how they communicate, what is their level of influence etc. That is all around social context... Obviously Xtract is no longer alone in that space and we can see companies like say Google angling very strongly into this space as well as many of the Scandinavian/Nordic mobile operators, some handset vendors etc..

Vasco - very good point and we agree. Yes, that moment of when to make the recommendation is very important, and ideally it should not be 'automatically guessed' by the operator, but should actually be asked from the consumer - when do you want the ads. Or a little reminder that 'now I have something for you' a bit like the envelope flashing when we have received an SMS etc. Yes, we have to be very careful that we won't bombard consumers with ads at the wrong time and in the wrong context. Good point.

Thank you both for writing

Tomi Ahonen :-)

[Sheldon Nesdale]

A bit wordy to be honest Tomi. Next time please use more bolding and bullet points. :P

[Tomi T Ahonen]

Hi Sheldon

This is a complex industry. Often a superficial treatment is of no value to readers who expect to listen to a global guru who has written 9 books.

This is a blog, we have no advertising, I have no interest whatsoever in attracting you to this blog, Sheldon, or keeping you here, except my altruistic passion as per my motto 'in a connected age, sharing info is power'

If you can't handle this type of treatment of new topics, I have 'edited' text in books for those who want it in better formats. My regular readers know that what I blog about now will be in my books next year. So if you don't need to catch the latest thinking and are happy with 'very well organized' but somewhat out of date thinking, go ahead and buy my books.

If you feel this was not of value to you, please do not come back.

If you can appreciate it, that this is only a hobby for me, we get no income from this blog and all thoughts here are early drafts into next books, then like some who visit this site religiously, will put up with the fact that sometimes it is very late at night, I write out of a passion to share, and my style is long and winding.

We have readers of very varying degrees of competence and knowledge. Sometimes I explain things that to some readers seem basic - but for example Americans and Europeans have very different understanding of this space - yet most comments on this blog tend to suggest my readers appreciate the blogs. Perhaps you are out of your league here Sheldon.

We are often selected as the best blog of the week by our peers (the weekly changing hosts of the Carnival of the Mobilists). Like this week, my blog posting about Fools Gold was again voted best blog of the week. And it was typically a long Tomi Ahonen blog. What I do when I write, is to be inclusive, even when repetitive (even when that means I waste more of MY time doing so) to make sure our readers get the full story, not just jargon for the insider crowd.

I am a windbag, yes. You can also read other authors who cite me. There are 70 books that do so. That may be better suited for you. some of them have bullet points and pretty pictures. But they tend to cite my books from a couple of years ago, so your info is getting a bit tired by then.

But this is my passion. I apprecite it that you did not like the length of this blog. Too bad. This is my hobby and my blog. I will not go back to put it into bullet points or edits or bolding. And I dare you to show me one blog who covered this same very abstract, very complex, very bleeding-edge topic with less words prior to 8 March 2010.

Go away

Tomi Ahonen :-)

[Polly]

We have readers of very varying degrees of competence and knowledge. Sometimes I explain things that to some readers seem basic - but for example Americans and Europeans have very different understanding of this space - yet most comments on this blog tend to suggest my readers appreciate the blogs. Perhaps you are out of your league here Sheldon.

[white iphone 4]

Recently i am looking fot the white iphone 4 available conversion kit. If you have any recommendation, plz tell me. I will be very grateful!

[karen millen outlet]

Peace of mind from the minute you buy.

[karemillen]

The new brand of clothes are nothing like the old one. Right now, almost everything is done in different countries and then imported to the United States to be sold to consumers.

[oil mill]

can't again so capricious, so naive, so childish.
Learn seriously, the earnest of the people, serious thing right.
Learn to be patient, the shut up just shut up, this silence is silent.
Life of the.

[iphone cases]

e patient, the shut up just shut up, this silence is silent.
Life of the.

[Tablet PC]

Happy to see your blog as it is just what I’ve looking for and excited to read all the posts. I am looking forward to another great article from you.myefox

[christian louboutin]

est of the people, serious thing right.
Learn to be patient, the shut up just shut up, this silence is silent.
Life of the.

[best android apps]

Did you just get a new Android phone? Maybe a new HTC Incredible or Droid Evo? You've gotten the screens looking good, the font options just the way you want them...but now you need a few Android apps to make it even better? And free apps are always better than ones you have to pay for, right?

[karen millen dress]

If you live to be a hundred, I want to live to be a hundred minus one day so I never have to live without you

[cherry]

Wow, this was a really great post.

[クレジットカード現金化]

ありがとうございます

[ercai]

Did you just get a new Android phone?

[high heel sandal]

Haha

[Pandora Canada]

It’s a Very helpful article for me. Actually, I am fond of reading online punjabi news. Thanks for writing such a complete ..And,I wantn’t to miss them.
Thank you for sharing..