The attention economy

“In an information-rich world, the wealth of information means a dearth of somethiing else; a scarcity of whatever it is that information consumes… the attention of its recipients.” — Herbert Simon.

The news media was tightly interwrapped with the attention economy, often using fear - e.g. lurid headlines - to draw people in. The history of TV, radio and newspapers tarnished by this to some extent - so we shouldn’t be surprised when we see this with social media.

Deluge of information from social media, and it’s challenging to deal with. There’s no way to “consume everything”, so we have to have coping mechanisms. Some people opt out entirely. But most of our tools are designed to make people feel guilty for all the things they haven’t read.

One thing’s clear: the amount of information isn’t going to decline any time soon. So what this means is you’re constantly struggling to catch people’s attention. Becomes psychological warfare as attention-seekers seek to use any and all emotions to pull you in. Which is where we get back to fear. Because fear works so effectively, it becomes more and more useful as a tool for getting attention. Result, a cyclical dynamic which amps up and up.

What we’re seeing with social media is more complex than we saw with TV, because instead of being broadcast we see messages within our friends groups - how they spread things, how they phrase it.

Radical transparency

An idea core to the tech crowd - notion that putting everything out in the open will make people more honest. Idea is to make corporations and government actions to do this to be more transparent.

But the reality’s much messier. A lot of how radical transparency plays out every day is about making the voices of ordinary people exposed - c.f. rhetoric around real names (nymwars), or around Anonymous. But still the idea that exposure will make people more honest.

The idea of “outing” isn’t new, and has been fraught for the LGBT community for a long time. (Should people be outed against their will to make it visible that there are gay people out there.)

Very hard morally to ask how benefits and costs balance out, but this is the question we have to ask whenever we put anything online. Advocates of radical transparency would say that yes, openness always has greater benefits in the long run - but boyd argues this is unresolved and not that simple.

In boyd’s work with young people, she sees that most of the meanness and cruelty isn’t from anonymous actors, but from people they know.

Also new technologies emerging to expose people in new ways - e.g. photos taken during protests (or riots, depending on how you define), then crowdsourced request for the public to identify who was in the photos.

Foucault, panopticon - people can be controlled when they feel they are being surveilled. These tools are part of that dynamic. Radical transparency is part of this because it pushes people out in a way that that surveillance is internalised. Those who are marginalised, those with the least power often suffer the most in these kinds of exposure.

The ideal of progress.

A lot of radical transparency follows a linear narrative of progress - that exposing people will increase openness.

But isn’t true. Example of queer space being much more inclusive prior to 1930s prohibition (than in 1970s/80s), because people weren’t having to hide in the closet. “Fear squelched that.”

These paths aren’t linear, they ebb and flow. Social progress isn’t linear.

Puts herself on the line - geek, third wave feminist, socially liberal, not particuarly religious but respects people’s beliefs. Acknowledges that all this shapes how she thinks about progress. e.g. deeply committed to the notion of tolerance, but it’s often claimed that it’s a neutral notion - everyone’s committed to tolerance. But it’s not. Our perceptions of where we should be tolerant our shaped by our beliefs and upbringing.

So why do we assume that just because people go online and are exposed to more people, everything will be good? But just because you have the opportunity to meet strangers and value their perspectives, doesn’t mean you necessarily do. The internet makes it possible to see people, but it exposes us to people who don’t necessarily share our values.

e.g. Cyberbullying. Data shows really clearly that cyberbullying didn’t increase the overall volumes of bullying. Bullying continues to occur in greater frequency & intensity at school…. But we believe cyberbullying is worse, because it’s more visible. A kid coming home moody isn’t clear, but a post on Facebook is more visible.

Fear flows across networks. Result is that parents are being constantly being bombarded by messages of fears, which generates constant anxiety.

“The mean world syndrome” - notion within TV that if you were exposed to the idea that the world was a mean and terrible place through media, you’d become convinced that the world was a mean and terrible place.

boyd is very pro online culture - it’s about makign the world a better place. But the internet is now mainstream and there hasn’t been an enlightenment. Nonetheless, we keep pulling back to utopian and dystopian rhetorics - our hopes remain very high, unrealistic.

e.g. that the Arab Spring was narrated as the “social media revolution” - utopian hopes for what it can do. NB that voices we heard about events in Egypt wasn’t in fact everyone, hence much headscratching as elections showed a different picture of what people wanted.

Eli Pariser, filter bubble.

However thinks Manuel Castells communicates this better. Critique of Arab Spring - when technology disrupts things, it produces a new scramble for power. Argues for four ways we can understand power within netowrks.

1. Networking power - inclusion, exclusion, e.g. being invited to a party

2. Network power - setting up the rules for who’s included/not

3. Networked power - the people who control the informational power, e.g. people behind SXSW who create the rules

4. Network-making power - the people who actually make the system, controlling what information flows and what we can see. These people become powerful even if they’re not the most powerful people within a hierarchical organisation.

In geek culture we have thought we were making the networks, but actually traditional forms of power are taking control.

e.g. Kony2012.
What Invisible Children has done over many years is set in motion disparate social groups who seem to be disconnected, e.g. teenagers in different schools. They understand how to do movement-building. So they sent out messages to different communities to tweet about this as a same time- which turned it into a Twitter trending topic, and from there spiralled. (The algo loves trends that seem to be arising from multiple, apparently disconnected sources.)

No nuance got through that network - it became a very binary story, frustrating many (inc. Ugandans who were angry to be told how their country works.)

Also 2012 election in US - using fear and propagating through different social environments.

What’s disturbing - a lot of hatred of people who are different from us. Geek culture may be xenophilic (rly?), but mainstream culture isn’t. E.g. anti-Mexican and anti-Muslim sentiment.

To pretend this isn’t happening does a disservice - we aren’t building the societies we want. We want technologies to be used for our idealistic purposes - democracy, transparency etc - but this isn’t happening in practice. Not necessarily neutral technologies and not necessarily spreading information we like. What do we do about this? The tools are being repurposed

As geeks we have power in building the networks and shaping how they are used. We need to think as a community about how to shape them to combat fear. How do we see these power dynamics, recognise that the culture of fear is being propagated by social media, and solve this?

Social media can be a disrupter, but it is also being used as a force for political means.

A clear and convincing take-down of the notion that (a) influencers exist, and (b) marketing at them is worth doing.

To summarise the key points:

1. Six degrees of separation does in fact exist - but it doesn’t work the way most people think it does. When we imagine these chains we assume they must work through celebrities, leaders, & other social connectors - whereas in fact messages travel horizontally, not hierarchically.

The overall message here is that real social networks are connected in more complex and more egalitarian ways than Jacobs or even Milgram imagined— a result that has now been confirmed with many experiments, empirical studies, and theoretical models. In spite of all this evidence, however, when we think about how social networks work, we continue to be drawn to the idea that certain “special people,” whether famous wives of presidents or gregarious local businessmen, are disproportionately responsible for connecting the rest of us. Evidence, in fact, seems to have very little to do with why we think this way

2. In The Tipping Point, Malcolm Gladwell talked about “social epidemics” set off by a small number of “superspreaders” - what he calls “the law of the few”. This is obviously catnip for marketers - it makes manipulating a large audience seem straightforward - and is the model behind influencer marketing. Unfortunately it’s not true.

3. Influence may be subtle - e.g. from observing behaviour rather than explicit verbal recommendations. It may also happen in ways we don’t acknowledge because they don’t fit our model of how things “should” work - e.g. employees influencing their boss.
[NB This isn’t necessarily problematic for influencer marketing: it’s likely to be fine if it “works” by people copying the influencer’s actions rather than doing what they say/blog about.]

4. There are different types of influence - friends are influential in different ways to Oprah. People may also be influential only on their very specific topics of expertise.
[NB Again not necessarily a problem for influencer marketing.]

5. Measuring influence is very difficult, so usually we’re not measuring it directly at all:

For example, to demonstrate just one incident of infl uence between two friends, Anna and Bill, you need to demonstrate that whenever Anna adopts a certain idea or product, Bill is more likely to adopt the same idea or product as well. Even keeping track of just one such relationship would not be easy. And as researchers quickly discovered, doing it for many people simultaneously is prohibitively difficult. In place of observing infl uence directly, therefore, researchers have proposed numerous proxies for infl uence, such as how many friends an individual has, or how many opinions they voice, or how expert or passionate they are about a topic, or how highly they score on some personality test— things that are easier to measure than infl uence itself. Unfortunately, while all these measures are plausible substitutes for infl uence, they all derive from assumptions about how people are infl uenced, and no one has ever tested these assumptions. In practice, therefore, nobody really knows who is an influencer and who isn’t.

6. How much more influential is an influencer than a regular person? Let’s say fully 3x more. That doesn’t get you very far in reaching millions of people.

7. Doing so requires adding a second idea from network theory, that of social contagion. The hope is that this will multiply the “law of the few” so that the influencer’s choices reach millions of people. The problem is that networks don’t work like that.

8. Duncan Watts and Peter Dodds tested this using computer simulations, they found that the initial “influencer effect” didn’t hold across the whole network. An individual with 3x influence didn’t result in three times more people overall being influenced - in fact, often their impact was negligible.

The reason is simply that when influence is spread via some contagious process, the outcome depends far more on the overall structure of the network than on the properties of the individuals who trigger it. Just as forest fires require a conspiracy of wind, temperature, low humidity, and combustible fuel to rage out of control over large tracts of land, social epidemics require just the right conditions to be satisfied by the network of influence. And as it turned out, the most important condition had nothing to do with a few highly influential individuals at all. Rather, it depended on the existence of a critical mass of easily influenced people who influence other easy- to- influence people

9. Yeah yeah, but that’s a computer simulation. So Watts,Jake Hofman, Winter Mason and Eytan Baksh tested this on Twitter data, looking at link-sharing:

The nice thing about these shortened URLs is that they effectively assign a unique code to every piece of content broadcast on Twitter. Thus when a user wishes to “retweet” something, it’s possible to see whom it came from originally, and thereby trace chains of diffusion across the follower graph.
In total, we tracked more than 74 million of these diffusion chains initiated by more than 1.6 million users, over a two- month interval in late 2009. For each event, we counted how many times the URL in question was retweeted— fi rst by the original “seed” user’s immediate followers, then by their followers, and their followers’ followers, and so on— thereby tracing out the full “cascade” of retweets triggered by each original tweet.
As the figure on page 102 shows, some of these cascades were broad and shallow, while others were narrow and deep. Others still were very large, with complex structure, starting out small and trickling along before gaining momentum somewhere else in the network. Most of all, however, we found that the vast majority of attempted cascades— roughly 98 percent of the total— didn’t actually spread at all

10. Comparing half the dataset against the other - as a proxy for a past/future predictive split - they found that individual-level predictions (as a marketer would do, in identifying influencers to target) were very noisy. ” Even though it was the case that on average, individuals with many followers who had been successful at triggering cascades of retweets in the past were more likely to be successful in the future, individual cases fluctuated wildly at random”

Conclusion: The most effective way to influence a network is through a lot of ordinary-level people, not an elite of “special influencers”.

Marketing insight: You *might* do that with freebies and promotions, but it’s pretty expensive to give even something small to a mass group (e.g. 5% of customer base). What you actually have to do is:
(a) Not Be A Shit - because if you fall down on the hygiene factors they’re not going to recommend your brand; then
(b) Deliver gold-standard customer service. Making a problem into not-a-problem quickly and painlessly can delight people, and that’s how you create advocates.

Ah the challenges of having an original idea! In the process of transferring my longer-form posts here to my new domain JayOh.net, I find that the themes in my last post Influencer marketing: two speculative stories were prefigured by Tom Ewing over a year ago:

There are “influencers” who live in the world of “social media”, and if you influence these “influencers” they will say nice stuff about you and your brand will do better.

The reason they are “influencers” is that people who aren’t influencers - let’s call them, oh, “nobodies”, that’s catchy - anyway these nobodies listen to the influencers and spend money on stuff they mention.

So what you need to do is find a way of measuring who’s an influencer and who’s a nobody, and then be really nice to the influencers and give them free stuff - no strings attached of course, since otherwise you would go against BLOGGING ETHICS.

Of course in order to become influencers you need to have lots of nobodies who will listen to you. Some of them will be your friends, so what the brands are doing is giving you free stuff and specifically not giving it to your friends, in the hope that you’ll show it off to them.

This isn’t a brilliant concept of friendship TBH. It’s pretty much the way street teams operate but they’re aimed at 14 year old boys. The lack of reciprocity here also isn’t a problem if you’re a Social Media Rock Star and all your real life friends are Social Media Rock Stars too because they’ll also be getting the free stuff.

So, OK, what if you’re a nobody? (I am a nobody, by the way - I used to be an “influencer” but I guess I haven’t been tweeting enough lately.) DON’T WORRY though, you can easily become an influencer too. Then as your influence increases you will have nobodies of your own to influence and you will get free stuff too.

How do you get influence? Spend more time on social media building your “relationships”. A good start might be retweeting the influencers you know, so they notice you.

If you think this sounds like a massive Ponzi Scheme, you’d be COMPLETELY RIGHT.

Influencing the Influencer on tomewing.tumblr.com

01. The attraction of influencer marketing is in being able to leverage word-of-mouth and peer recommendations.

02. This is valuable because peer influence is the most effective form of influencing what we buy, or what we feel about a brand. [source]

03. Peer influence has this impact because it’s advice from people we trust. Key to almost every definition of an influencer is their credibility and realness:

“Broadly I define an ‘influencer’ as someone who follows their own path, is rooted in creativity, and is looking for new ways to change or redefine their world. Someone who is an ‘influencer’ not only has broad relationships but also has deep relationships. In short, they are building a community around shared beliefs, principles, and interest.”
Philip McKenzie, Managing Partner at FREE DMC and Founder of Influencer Conference: [source]

04. We trust our friends and family because we have known them a long time and feel emotionally close to them.

05. This means we believe that their recommendations will have our best interests at heart.

*

But what happens when brands come in and try to get a piece of the action?

1a. MediaCorp have identified Amelia as influential about Personal Personality Monitoring Devices (PPMDs). They send her their new product, SomaTech, in the hope that she will talk about it to her sphere of influence and generate increased purchases.

2a. As a personal branding expert, Amelia is savvy to influencer marketing. She knows she needs to deliver visibility if she’s to continue receiving shiny freebies, so she schedules a series of tweets about the product for the next week. Each is very proper, including the brandname, the hashtag the PR had sent, and exhorting retweets from her followerbase with one too many exclamation marks.

3a. Amelia has 4,000 followers on Twitter, but two-thirds are bots or other personal branding gurus (or both), and the rest she purchased at $2.50 - $4 per follower. [source]. Almost everything she tweets is automated from Oprah Winfrey’s Paper.li, but she does also auto-schedule interactions with her sockpuppet accounts to keep her Engagement score up. This helps keep the free shiny things flowing from the social media PRs

4a. Amelia’s tweets about the SomaTech are retweeted a respectable 30 times each, and along with Amelia’s 15 positive mentions of the product herself, MediaCorp are happy.

5a. However MediaCorp haven’t connected up their Twitter analytics with their store’s Google analytics or purchase data. If they did, they’d see Amelia only generated 20 clickthroughs and no purchases. This is because her highly automated copy-cat content is followed by almost no real, active human beings. (One of the bots did try to buy something but the transaction was declined as their card was registered in Zurich and their shipping address, Belarus.)

6a. Next week Amelia shoots a YouTube segment for MediaCorp’s competitor’s personality monitor, SimSoothe, who sent her not just a free device but $250 as well. The video gets trolled by 4-Chan and goes viral.

The kind of social media user who’s highly receptive to sharing brand promotions may not be generating content that real people value. Influence metrics are highly gameable and, if incentivised by freebies, attract game-players - not the “real people” marketers actually want.

*

1b. So MediaCorp improve their influencers algorithm and have another go:

2b. Bill’s a normal guy, albeit the linchpin of his group of friends. He’s a bit surprised to receive the free SomaTech device as he’d tumbld just the other day about PPMDs being a bit creepy. This had got a bunch of reblogs and sparked a bit of a debate.

3b. Bill’s not sure what to do with the SomaTech, but his mum raised him proper so he knows that if he receives a gift, he’s got to say thanks. So he writes a blog post as MediaCorp requested, which is auto-shared to his other social networks.

4b. Bill’s ex-boyfriend sees his post. He knows MediaCorp have been exploting child labour in the Philippines, and sees an opportunity to embarass Bill. “Since when were you such a slut that you did everything some big company told you to? I remember that time we were talking about….” An argument breaks out among their friends.

5b. “Mate, seriously? That time you suggested I should get those Blahphonic headphones? Was that you, or, you know, something you’d been told to say?” asks Cate. Bill feels really awkward - it’s the fifth time he’s been asked that question this week.

6b. Bill disables his Facebook account, Tumblr etc - making him an early adopter of the Going Analogue trend, and influencing three friends to follow. But as he no longer has a social media presence, his departure goes tragically un-curated.

If customers know that peer recommendations have been purchased by brands, many will stop trusting these friends’ recommendations. This is pretty corrosive to the friendship - and defeats the brand marketer’s purpose too..

So what if publicly disclosed “I did this because Brand X paid me” is individually toxic to personal credibility - but sneaky (non-disclosed) influencer marketing is socially toxic to friends’ trust?

Perhaps the only circumstances in which it works is celebrity influencer marketing - but only where there is no relationship of trust between celeb and fans. To participate in influencer marketing is to say, “I see my friendships as social capital, and I’ll exchange them for a fee”. Works fine for celebs who’re selling themselves (*kough, Kardashians*) but among almost every other social media user? Rather more double-edged.

Notes on the semantics of the internet

What I’ve been working on lately - @O2’s brand graph, i.e. the social and interest graphs of @O2’s followers.

Otherwise known as “Fran’s jellyfish”. Bit of a CPU-melter to put together this stuff - but pretty once it’s done!

Presented at WARC Datacentric conference yesterday.

Waste and rubbish are big topics in geography and urbanism at the moment - what happens to the stuff we throw out? Stuff we no longer use still exists and takes up space: urbanists are exploring where it goes and what this says about our society.

(cc. ICON magazine’s issue 101 out now on Waste)

But what happens to our digital waste? Facebook’s new Timeline feature has caused some consternation by its proposal to show key events from people’s whole time on the site - do young professionals really want drunken freshers-week photos (or worse!) still showcased as part of their digital identities? And what about all the old social sites still out there, but now hardly used: MySpace, Friends Reunited; LiveJournal; (the closure of Geocities etc)?

- Google etc as erasing the possibility of forgetting
- Managing memory as part of a process of online identity-making
- Social media’s tension between real time vs. history

(TBC)

“If urban history can be applied to virtual space and the evolution of the Web, the unruly and twisted message boards are Jane Jacobs. They were built for people, and without much regard to profit. How else do you get crowds of not especially lucrative demographics like flashlight buffs (candlepowerforums.com), feminists (bust.com) and jazz aficionados (forums.allaboutjazz.com)? By contrast, the Web 2.0 juggernauts like Facebook and YouTube are driven by metrics and supported by ads and data mining. They’re networks, and super-fast — but not communities, which are inefficient, emotive and comfortable. Facebook — with its clean lines and social expressways — is Robert Moses par excellence.”

Best internet metaphor ever

The Old Internet Neighborhoods by Virginia Heffernan, Sunday 10th July, NYTimes.

New fashion trends consultancy Editd is (finally) bringing big data, crowdsourcing and quantitative rigour to the hitherto very expert-driven world of fashion forecasting.

Pretty amused that this is in the same week I’m working on a pilot trend forecasting project at Face, seeing if social media buzz can be used to predict what’ll be the key looks at Glastonbury this weekend. Our monitoring tool Pulsar uses same range of social media datasources as Editd - Twitter, Facebook, blogs, forums, Tumblr and all the rest - bringing them all together on a bespoke dashboard designed for research rather than PR use.

But Editd are using a few extra datasources too - included the Top 20 most Liked garments on ASOS (and elsewhere?) to produce a really strong indicator of current trend buzz (this week: cut-outs). I wasn’t aware that ASOS had an API to pull this data from - and I’d love to know where they’re getting retailers’ sales figures for individual items too - but this mash-up of datasources is the future of business intelligence and it would be the future of market research, except it’s pretty clear this won’t be happening in market research companies per se.

Having these different stages of purchase intent - from general social media buzz, to what people have been browsing and Liking, to actual sales figures - should allow it to be a very powerful prediction tool, capturing different stages of the decision-making process to probably do some pretty prescient stuff.

With a bit of expert curation on top, Editd looks really exciting - follow at @EDITDtweets to see what’s next.

That said, ASOS have half this stuff in-house already: at least all the Likes and all the purchase data, and I’d be very surprised if they’re behind on the social media monitoring… Wonder what we’ll see develop there too?

Over at FACE we’re exploring social media monitoring as a prediction tool - does what people say they’re going to do indicate future behaviour?

The test arena: Glastonbury 2011, and what people are going to wear. Are the style blogger “influencers” really influential, or do mass mentions have greater predictive power?

ASOS may be talking about sheer maxi skirts and £240 Prada sunnies, but we’re staking our money on gloss black Hunter wellies and short shorts - or the attack of the furries with Kigu-style full-body animal suits looking strong.

Though it’s now looking like we may be on for a re-run of Glasto 2007, in which case it’s mud head-to-toe for everyone. Let’s see what happens!

Image by @marionren

Comment by thrawnpop on Guardian article, Is this the start of the second dotcom bubble? by Dominic Rushe, 20 Feb 2011.

This.

Everyone’s like, OMG, Facebook means we know so much more about our consumers - that’s got to be worth sofuckingmuch.

They seem to be forgetting that “consumers” - aka people - don’t have limitless money to suddenly start buying loads more stuff. Better targeting may make more people buy the socially-marketed product, but only at the expense of not buying stuff they currently purchase.

It’s essentially zero sum.

In the UK and America, salaries are flat and inflation is over 5%. In real terms people are getting a bit poorer every year. What’s different is that this time, we know it - after a decade’s economic growth funded by consumer borrowing (funded by debt, not wealth), individuals - and lenders - have cottoned on that that gravy train has halted. With little in the way of savings and nervous in mood, now we spend what we earn. Which is decreasing.

Facebook marketing could be magic mind control juice, but all consumers could do would be to want more stuff, not buy it.

The only way Facebook can justify these valuations is if it manages to take essentially all advertising budgets for everything everywhere.

Size of the global advertising market: estimated to reach $500bn in 2012
Google: valued at $189bn
Facebook: $50bn

1. There is a New New Thing that trancends the Old Economics, and you cannot value It the Old Way. This Time It will Be Different

That’s from the now-famous article, 10 steps to see if you’re in a bubble.

A comment I left on a Research magazine article about the uses of social media in market research - which incidentally I also contributed to writing along with fellow FACErs @abc3d and @sharmilasub ;)

Definitely pretentious to cite Hegel, but perhaps we need to call in the big intellectual guns if we’re going to break down this persistant obsolete notion research has that binary distinctions are meaningful/useful/valid!

(Next up, poststructuralism 101?)

The Atlantic | Delusions Aside, the Net’s Potential Is Real

Brilliant response from Zeynep Tufekci to @evgenymorozov’s The Net Delusion. It’s particularly relevant in the UK given the recent student and #UKuncut protests. In a situation where the coalition lacks a coherent mandate and writing to your MP has never knowingly changed shit - “a fundamental collapse of citizen leverage on institutions of power” is exactly what we’ve got. And I think or hope it’s going to be online organising that gets us out of this mess if anything can, because it ain’t going to be the “Big Society” meeting in draughty church halls…

Hear Morozov’s argument in person - he’s speaking at the London School of Economics on 19th January.