Friday 17 May 2013

The embodied cognition of Tesco's gendered toys

 

Tesco got in trouble on the internet last week for having toy chemistry sets labelled as being for boys, not girls in their online store. There's a lot of noise about how inappropriate all this gender labelling is (and rightly so - it's everywhere and it's awful). Lots of potential customers are being very annoyed all over Twitter: so why does Tesco do this? Why is this sort of thing so very common? Oddly, I think an embodied task analysis (using our 4 questions which we describe in our Frontiers paper) might shed some light on this question, while also perhaps serving as an accessible example of what it is we're up to with our work.

The thing that got me thinking was this: a Tesco spokesperson said on Twitter that “Toy signage is currently based on research and how our customers tell us they like to shop in our stores”. No one believed them, with replies ranging from 'what kind of research' to the less useful but more common 'I would never tell you to do this so what the hell?'. I think they are telling the truth, and I think their behaviour can be readily explained using our four key questions. More importantly, I think it reveals how we can change Tesco's behaviour, and the answer means knowing about the extended system from which that behaviour emerges.

1. What is the task to be solved?
Tesco, like all large corporations, is in the business of making money. Unlike humans, who try to do all sorts of things like speak and catch balls, this is basically the only thing Tesco are ever trying to achieve. This is the task they are trying to solve, and this means that any behaviours they exhibit must be interpreted with respect to this task goal. 
  • Task: maximise profits from children's toys
2. What resources do they have available to solve the task?
In order to make money, Tesco offers a range of goods and services to people and does their best to make sure people get these from them, rather than a competitor. Which goods and services they offer is a function of what they think they can sell, and they therefore require information about what people want to buy. There is also flex in how Tesco go about the act of selling their products, and so they need information about the factors that affect consumer decisions (pricing, presentation, etc). They then try to sell things to people and see how they go, and fine tune their selling as a function of that outcome.
  • Resource 1: things to sell
  • Resource 2: information about what people buy
  • Resource 3: information about what affects what people buy
  • Resource 4: the ability to engage in the transaction
3. How might they assemble the resources to solve the task?
The important thing to note is that Tesco only has a limited amount of information. Like people, it only 'perceives' information that is relevant to it's activity. It 'perceives' how many of each item type people buy and at what price, and via it's online store it also 'perceives' what people search for and whether people buy any of the results they serve up.

Tesco claim that their toys are labelled with gender because research tells them this is what people want. They wouldn't tell me what kind of research, but my guess is that it's mostly based on search terms used on their website. People casting around for ideas will type in 'toys for boys' or 'for girls' and then buy the results they think fit that search the best. 

From the first person perspective of Tesco, all it perceives is which search results successfully sell someone a toy. That's it. It doesn't perceive your indecision, or the middle-class angst you feel in buying a chemistry set even though its labelled 'for boys'. So my hypothesised system to explain Tesco's gender labelling behaviour is simple: 
  1. Offer toys for sale online.
  2. Note that search terms for toys often specify a gender. This tells Tesco people want access to information about this to facilitate their purchasing.
  3. Note which toys are bought after people search for toys 'for boys' or 'for girls'. Label appropriately.
  4. Sell toys.
  5. PROFIT.
  6. Monitor for any changes.
4. Does Tesco, in fact, do this?
I'm not able to answer this for sure; as I mentioned, they wouldn't tell me what kind of research they were relying on. Even if the research is focus group based, though, the basic form of the system remains the same, with just minor differences due to the different dynamics of that different source of information (e.g. focus groups would only enable periodic updates, so the system would be sluggish to respond). I'm happy to bet that there really is research of some kind, though - Tesco needs information in order to maximise profits as much as people need information in order to achieve their task goals. 

So what have we gained?
The embodied analysis reveals a few important facts. Tesco's behaviour, like ours, is a function of the flow of information it perceives and what that information tells it about whether it's achieving it's task goals. Therefore, to understand its behaviour, and, more importantly, to figure out how to change that behaviour, it is critical to understand the dynamics of that information flow, i.e. what is and isn't included and how that information plays out over time.

Tesco can perceive that people include gender information in their searches for toys. Tesco can also perceive that when they shape their online store to include this information, it affects consumer behaviour; specifically, it makes people more likely to buy what they find in the search. (This must be the case, or else Tesco would have no drive to either make or maintain the change to using gender information in their labelling. Remember, Tesco's task is simply to make money.) Tesco cannot perceive that you were buying the toy 'under protest', nor can they perceive you not ever shopping there - the absence of information cannot shape behaviour (see: friction).

The first thing we learn, then, is that Tesco is not evil. They are simply responding to the flow of information in which their behaviour is embedded. The behaviour 'gender labelling' is an emergent property of Tesco performing the task they are in the environment they have information about. It is a property of the extended system, and not an internal trait of Tesco.

The second thing we learn is how to change this behaviour. The only way any system can alter it's behaviour is by either trying to achieve a different goal (becoming a different task specific device) or by being exposed to a different flow of information that makes the old behaviour no longer stable. Tesco can't change task: making money is the only reason it exists. The information therefore has to change, and because the information is generated by us, the consumer, we have to change what we're doing. We have to tell Tesco, in a format it can understand, that gendered labelling is no longer going to improve profits. So stop searching online for toys by gender. Stop buying those toys full stop. By itself, however, this absence of information can't tell Tesco what to do differently, but it can help prevent the active maintenance of the gender labelling behaviour.  

But most importantly, you must tell Tesco (and others) what you do want (gender neutral labelling) in a format it can understand so they they have a behaviour to switch into. Use Twitter, sign petitions, critique them when they do it wrong (and hey, look - that new flow of information is already have an effect) but, most importantly, reward them when they do it right by buying their neutrally labelled toys more than their gender labelled ones. Otherwise they'll simply never notice they have to change anything. 

References
Wilson, A. D., & Golonka, S. (2013). Embodied Cognition is Not What you Think it is Frontiers in Psychology, 4 DOI: 10.3389/fpsyg.2013.00058

11 comments:

  1. I always wonder if some of this behaviour is so innate that to go against it would be like trying to swim upstream. My wife says it this way, "pickling a brain in testorone has to have some debilitating consequences"

    ReplyDelete
    Replies
    1. Pretty much none of the gender stuff is innate. Read Cordelia Fine's book 'Delusions of Gender' for an excellent review and critique of the 'evidence'. It's not that males and females are identical, but most of the specific psychological differences people talk about are non-existent.

      Delete
  2. Where is the science here? Maybe Tesco knows its market better than you do. Where is the evidence that Tesco is doing something harmful?

    ReplyDelete
  3. Please define "science", "evidence" and "harmful"

    At least one of those is in the eye of the beholder, isn't it? So, your point doesn't really hold water.

    ReplyDelete
  4. This is really interesting. This is a very plausible explanation of Tesco's labelling policy - though I see even less justification for applying this to real-world shelves where you can visually scan much more easily. Confronted with a jillion toys on a website I suppose some people will search for 'toys for girls/boys' as a conventional distinction, so it's circular. Consumer behaviour is both cause and result. But this behaviour has been shaped and reinforced by marketing, as it's got noticeably worse in the last few decades.

    'We've always done it this way, everyone's used to it' isn't a good reason to carry on. Cf cigarette marketing, seatbelts in cars, slavery, women not having the vote etc.

    People may also search this way because they know NOTHING else about that child other than its gender and need some way of reducing the choice available. Too many birthday parties? Too many toys?

    As you say, this 'research' excludes so much, including failing to find what you want, or stomping off in a huff to shop elsewhere.

    Anonymous: Harmful: My children are given toys that don't match their interests, and are told their toy choices are 'wrong' both implicitly in shops, and explicitly by their peers policing the boundaries they see reinforced in marketing. Yes, I think this is harmful. Where do these people get off telling children which toys they are supposed to like?

    ReplyDelete
    Replies
    1. Confronted with a jillion toys on a website I suppose some people will search for 'toys for girls/boys' as a conventional distinction, so it's circular. Consumer behaviour is both cause and result. But this behaviour has been shaped and reinforced by marketing, as it's got noticeably worse in the last few decades.
      Indeed! One of the hallmarks of this kind of dynamical, embodied cognitive analysis is that kind of circular causation. Information flows both ways, so the relevant 'causal' factors live in the Tesco-Consumer system.

      As you say, this 'research' excludes so much, including failing to find what you want, or stomping off in a huff to shop elsewhere.
      Agreed, but my suggestion is that if this is all the information you have access to, then gendered labelling is a pretty rational solution to the task 'make money'. Information about these other factors is much harder to get into this kind of system, although not impossible with Twitter campaigns, etc.

      Delete
  5. I'm willing to believe that people search by gender - but this is LEARNED behaviour, because we know that the shop categorises by gender - same way we know to use quotes and Boolean operators on Google etc. Also, without data on the frequency of gender-based searches against other categories e.g. "toy cars" I am not sure how you can come to the conclusion that gender-based searches are driving the marketing... I would *expect* (note, I don't have any more data than you do about Tesco's search queries) that gender-neutral searches (e.g. "kitchen sets") are more common than very vague "boys toys" type searches, as most buyers have some sort of idea of what they're looking for, and the more wide gender-based querying would come later, as a result of not finding a suitable kitchen set that you want to buy.
    DA

    ReplyDelete
    Replies
    1. I'm willing to believe that people search by gender - but this is LEARNED behaviour, because we know that the shop categorises by gender
      Of course it's learned, but the learning is mutual (see my above reply about the circular nature of what's going on). Tesco learns from us, we learn from Tesco, and we both change our behaviour as a result until we settle into a stable relationship that satisfies all the relevant constraints. (This is roughly the thesis of Andy Clark's books 'Being There' (the role of the environment in this sort of system) and 'Natural Born Cyborgs' (the role of the person)).

      Also, without data on the frequency of gender-based searches against other categories e.g. "toy cars" I am not sure how you can come to the conclusion that gender-based searches are driving the marketing.
      It's an educated guess. I'm assuming Tesco aren't evil but that their behaviour is in response to some information relevant to profit maximisation. Search engine data is just one obvious source.

      That said, I heard from a company on Twitter (Tootsa MacGinty) that search data on their site typically includes gender information and that this hurts them because they are trying to sell all their toys as unisex. They've added 'for boys' and 'for girls' to everything to catch the searches. This makes the gender search uninformative, potentially shaping consumer behaviour, but also making their site produce a less refined search than other sites, a potential cost again.

      So gendered searching does seem to be a thing. Another victory for theoretically driven speculation :)

      Delete
    2. Thanks for the reply. I know a bit about HCI (I'm not an expert though) and I get the feeling that a generic category search (e.g. gender) would come AFTER a more specific search in typical user behaviour... however I may just be projecting MY search habits onto the general populace... I am off to see if I can find any articles investigating this - would you like me to share links if I find any?
      DA

      Delete
    3. Sure - I'm always interested to know if my guesses can be backed up with data :)

      Delete
  6. As most readers of this blog know, I'm writing a series of four books on Standard Operating Procedures for Small Technology Consultants.click to explore

    ReplyDelete