Brands could soon predict their next customers online.

An interesting research project has been commissioned by Microsoft, to look at the connections between people on social networks and the hypothesis that the majority of your next friends on Facebook, Twitter etc.. can be predicted from a purely scientific standpoint. Jure Leskovec has received a grant from Microsoft to extend his research, already started, to examine the science of connections we make on social networks. The research project itself is fascinating, and early results prove the validity of the hypothesis. But this really starts to get interesting,when you look at the possibility of this being applied to brands, and the possibility to predict your customers of the future, before they even know they want you.

The research

Leskovec’s research is based on the idea that connections aren’t random and can largely be predicted by commonalities in behaviour such as commenting on a blog post, joining a group, or reading an article online. It is behaviour like this that Leskovec believes will lead to future connections between people on Facebook and this was proved by a project he ran at Facebook itself, looking at the connections between personal networks. His research extended into other areas of social network analysis, for example looking at the concept of six degrees of separation, and how it applies on Microsoft Messenger. The theory, he found, still rings true on the messenger platform, where the degree of separation is around 6.6 – meaning you’re only ever 6.6 steps between one person and another. This research comes among a growing number of projects commissioned to look at the science of social networks – including research into how location can help us predict who become our next friends. For now the research projects are largely focusing on connections between individuals, but what happens when we take the same hypotheses and apply them to brands?

Targeted marketing

Right now brands have more control than ever over the people they can reach on social networks, with the ability to target by demographics as well as personal interests and even content that’s shared by people. You can build up profiles of your different target markets and serve them increasingly relevant ads. But take this a step further – what if you could only market to those people that you knew were going to end up being your customers? The hypothesis put forward in the research above can easily be applied to brands, and in some ways already is.

By developing a more scientific approach to marketing (not that science has been lacking up to this point), there is a huge opportunity to look for commonalities between your existing customers, and apply those patterns to seek out new customers. If you’re able to access information about your customers such as patterns between the content they share, the groups they join, the sites they engage with, you can apply a formula to seek out other people whose patterns of behaviour mimc this and try and convert them into customers. What happens at that point of course, depends largely on the strength of your marketing, but the potential to get in front of exactly the right people who are likely to become your customers anyway shouldn’t be underestimated. It represents one of the most significant changes in targeted marketing that we’ve seen since the introduction of social advertising.

Who owns our decisions?

If the science of social media continues to evolve and be applied to brands, it brings up the question of who really owns our decisions? If the situation I’ve presented above is applied, it essentially takes out the element of choice by the user. Instead of allowing them to arrive more naturally and organically at the decision to try out your brand or product, you cut out this waiting time and get right in front of them straight away, pre-empting their decision to purchase. And of course, while you can’t physically force someone to purchase your product no matter how targeted your marketing is, the freedom of choice is largely removed. If this type of marketing continues, then you won’t be shown ads or marketing by other brands anyway, as they’ll have decided you don’t apply to them. Imagine if every single ad you saw – on your computer or on the street – was shown to you because you fell into the right demographic for the brand. How will you know what else is out there, or make educated decisions between one brand and the next?

For brands, this is largely a good thing. But there is a risk that this approach to advertising is misapplied. A customer that has chosen you naturally, after reviewing the competition and researching among friends, is going to be ultimately more valuable than one who uses you with little choice in the matter. It’s akin to having to go with one broadband provider because they’re the only one who serves your area. You’re likely to be annoyed that the element of choice has been removed. As consumers, we have to own our decisions, if any long term relationship between brand and consumer is to exist.

The science of online marketing

What’s interesting is that this research project comes amidst a string of other projects that want to examine further the connections between individuals in social media. And while this is valuable and necessary as social media becomes a vital form of communication between individuals, it’s likely that this type of research is being commissioned as brands or advertisers want to examine more deeply how to reach these individuals through social technologies and convert them into customers. Psychology has always played a huge, vital part in advertising as it affects the likelihood that you will make your ad relevant to someone and influence their decision. And now the same method is being applied to social media, though for some it may have been a while coming.

While for many brands the primary concern is how to measure social media, there is a much bigger opportunity in how you actually use social media to reach your customers. And we are only at the very start of this and are just beginning to find out what’s possible. The opportunities represented by advertising in social technologies are vast, because of the unique nature of social technology being completely trackable. The data is all there online, and what’s next is to examine how this data can be used to develop increasingly targeted and relevant advertising. If this can be developed to the point that customers can predict down to the individual person, who will be their next customer, then the power of marketing becomes absolute and we look at quite a different relationship between brand and consumer than we currently have. When social media first emerged, the power was in the hands of the consumers, but the very technology that facilitated this could potentially turn it around and give the brand power once again.

To show that this is much more than just a hypothesis, and that the power of brands is rising once again, I wanted to look at findings on how brands interacted with consumers online during the holiday shopping period last year. Beyond looking at how customers are being targeted for purchase decisions, it shows how they’re being targeted to change their opinion, publicly. The graph below shows how customers were targeted that had posted something negative about a brand, with surprising results :

In 34% of the cases where a customer that had posted a negative comment about the brand/retailer, they actually deleted their comment, with a further 33% even then posting a positive review. This filtered down into 18% who became loyal customers. Without the social technologies and methods of tracking that we now have, this 18% would likely have been completely non-existent and would have represented a loss in customers due to a negative brand experience. But niche targeting and individual contact by brands shows the power that they have to change public opinion and turn a negative reaction into something completely positive – with the benefit of this extending beyond the customer themselves and into the word of mouth generated by their new positive comment. And as the research project above shows, we could be about to see this extend even further, with the ability to reach new customers before they’ve even started talking about you, purely by analysing patterns between connections. It’s powerful stuff, and just a little bit scary…