Went to an interesting talk on consumer marketing in digital age, with particular reference to use of personality data today.
These are the comments/notes I made…if I sound like I’m slating it, I probably am but I suspect my issue is I disagree with a few different groups in various ways, in particular aspects of personality psychology, and marketing. Anyway, point of that disclaimer is that I don’t want it to sound like I’m criticising the presenter – who did a good job, etcetc. – but the general approach is, I think, flawed.
The premise of the talk was that we’d moved from a Business->consumer communication model, to a business>consumer consumer>business peer>peer model. Specifically, that two things had revolutionised marketing – the internet, and social media.
Now first off, I think it’s pretty dubious to say some of that stuff didn’t happen before, and I think it’s not terribly helpful to segment history into pre-internet, post-internet. Off the top of my head, the supermarket would be another key thing which modified the consumer-business (and business-business) relation. Recognising that context is, I think important because it helps us to understand that one interpretation of what’s happened now is that it isn’t particularly a democratization of consumerism, but we’ve shifted from the supermarket being king, to the platform (facebook in particular).
Anyway, from that premise, the idea was that products (or, their marketing) have personality traits implied in them, which individuals are drawn to. An evolutionary reference to signalling theory was made here, the idea that we externalise our traits as advertisements. Now, another version of that goes, we externalise those things which advertise us – regardless of whether they’re our traits. There are risks to that sort of deception, but we can assume it happens – thus, the idea that we match our personality to that of the product is perhaps challenged. A more nuanced perspective would look at how particular personalities are drawn to particular marketing traits. Indeed, the example given – of coca-cola v. pepsi – which allegedly have rather different personalities in their respective marketing – indicated no significant differences between thte personalities of the purchasers.
That raised another interesting issue – who the hell ‘likes’ those pages on facebook, etc.? And for what reasons? A simple network analysis can tell you something about them, but you’d w ant to look at sentiment analysis to understand further reasons – including whether or not they follow the page in order to complain about it!
And I guess it’s that which I had a real problem with. I’m sometimes skeptical about how much business can use some analytics. However, businesses are doing sophisticated things looking at traversal trace-data to try and understand their consumers, and in some cases matching them up with ‘similar’ users – in the case of Amazon’s ‘collaborative filtering’ the outcome of this is the “customers who bought x…” and the “you might be interested in…” sections.
These things strike me as far more interesting than personality. And in particular, I’d want research showing that personality was stable over the internet, and that personality-product matching was important.
So the questions I’d ask probably are:
1) What is the relationship between consumer and marketing personality? i.e. can you match product personalities to consumers?
2) Is personality stable on the internet?
3) Does traversal data, or personality data better predict conversion rates (i.e. actual purchasing)?
I also think there’s a real worry about disenfranchising people, and placing too much reliance on these things – in a world where my Sainsbury’s purchased salad bag has a facebook page, it’s not entirely clear what those pages are for. Certainly, I find it hard to imagine what a marketer could do for the salad (although I can understand how ‘joining the conversation’ makes sense from their perspective in order to reply to criticism and support – but they aren’t about the kind of immutable domains personality claim to look at, they’re about sentiment and responsiveness).
Finally, I do wonder about personality as a construct. Obviously different people do different things in different ways with some commonalities across time, and between people, which might be drawn out. I’ve not thought about this extensively recently, but if I give an example: Imagine I ask you to go and whisk some eggs for me. You pull out your free range, pick 2, put them in your standard bowl, and take a fork (you don’t own a whisk), lean back on the counter and whisk. Another person pulls out their organics and takes 3 out into a larger bowl, using a whisk and sitting down. Now, no doubt there’s an interplay between environment and what you do, and you’ve partly setup the environment. But if we identify ‘egg whisking characteristics’ after the fact, and throughout time, ignoring nuanced changes, and looking for situations where we think these things are most likely to come out (all egg whisking situations probably) …well who cares? I’m not sure if personality is a useful construct in the sense that default behaviours which interplay with common environmental factors (and ways of perceiving those factors – which are also default) are just that -default, but so what? Particularly for advertising’s perspective, surely the whole point is to break those defaults, but also people do break them all the time – and I imagine particularly on the internet.