The Problem with Polling

24 Apr

This article was first published in the Association for Business Psychology newsletter on April 20th i.e. just after the upcoming UK general election was called, but written beforehand. I’ve consequently made a light edit to reflect last week’s events.


2016: The Year of the Underdog

By most objective standards, 2016 was a pretty remarkable year. In sport, we saw 2000-1 outsiders Leicester City winning the Premier League, and supposedly cursed perennial losers the Chicago Cubs winning baseball’s World Series. In culture, we unexpectedly lost a number of iconic figures, particularly in popular music. And in politics, we saw Britain vote to leave the EU and Donald Trump become President of the US – both of which were seemingly unthinkable last January.

Consequently, a £10 bet at the start of the year on Leicester winning the Premier League, Leave winning the EU Referendum and Trump winning the US Election would have returned a cool £45 million.

The unique property of political events is that our view of their likelihood (and that of the bookies) is largely determined by one thing: polling. Now UK Prime Minister Theresa May has called a snap election, media coverage will contain rolling polling updates predicting the outcome until the second results are confirmed. For those of us concerned with human behaviour, and how it often differs from our intentions as described to others, the consistent failures of pollsters to accurately predict recent elections is an interesting phenomenon. In addition to the Brexit poll and US Election, the 2015 UK and Israeli General Elections and this week’s Dutch election were also badly forecast by polling. By contrast, in sport events like Leicester’s win are the exception rather than the rule – the bookmakers always win in the long run (as I know to my cost).

But in polling, if anything, it seems like the problem is getting worse, not better. President Trump has caught on to this, and even used it to discredit his worst ever approval ratings.

Trump tweet

Understandably, the polling industry is concerned about dwindling credibility. As are those who have built careers on using this data.

At the Ideas42 Behavioural Summit in New York last October (just before the November presidential election), I saw prominent US polling expert Nate Silver – creator of the excellent FiveThirtyEight website – talk about the sophisticated predictive model he and his team built based on aggregated polling data. This correctly predicted the outcome in 49 of 50 states in the 2008 Presidential Election, and all 50 (plus the District of Columbia) in 2012.

Then, Silver pointed out that this model gave an increasing (but still slight) chance every day of Trump winning the electoral college, but losing the popular vote (it was about a 10-15% chance on that day, and up to 25% by election day).

That, of course, was exactly what happened.[1]

FiveThirtyEight eventually predicted a 71% chance of Clinton winning, although that was a significantly lower likelihood than other forecasters. In October they were forecasting the odds of Trump winning as almost exactly the same as the Chicago Cubs winning their two-horse race the week before the election.

I guess we should have taken that as a sign…

Bill Murray Cubs

The Chicago Cubs most famous fan, Bill Murray, celebrates their World Series win – after a short, 108-year wait



Poor Predictions

So why were these predictions wrong? At the conference (and more recently in posts such as this) Silver was at pains to point out at their “model” of behaviour is exactly that. It is designed as a forecast, and has a margin of error built in – just like their baseball model. It is not a perfect predictor of actual behaviour (if such a thing could ever exist), and he pointed out that (for example) the Brexit result was well within an acceptable level of polling error in most cases. The error was in perceiving that because all the polls predicted a narrow win for ‘Remain’ vote, that meant the likelihood was higher than 52%. He expressed disbelief at some of the odds ahead of the vote being offered by bookmakers for a ‘Leave’ outcome, especially given the UK’s track record of polling inaccuracies.

This applies to predicting sports events too, of course, but there you are predicting the behaviour of 18-30 people at a time. With an election, it’s several million, so that potential margin of error is several magnitudes greater.

But this doesn’t fully explain why the polls are consistently inaccurate, nor why this is especially true recently. Many theories have been put forward. Is it simply ‘shy Tories/Leavers/Trumpers’ (i.e. people embarrassed or uncomfortable declaring a preference for certain candidates or parties misleading the pollsters)? Bad methodologies? Or sampling errors?

Those with knowledge of behavioural economics may nod sagely at this point and say: “Well, if you ask people to account for their actions, of course it will be wrong.” If our actions are often unthinking and irrational, as behavioural economics has consistently proven, then how can people consciously account for them with any degree of accuracy?

This fundamental problem with polling (and traditional market research in general) was neatly articulated by Nobel Prize-winning behavioural economist Daniel Kahneman in conversation with Silver in New York. He pointed out that there is a key difference between building models of behaviour and how people behave in the real world: “The problem with algorithms is you feed the same data in twice and you get the same answer. The same is not true of people.”

But surely voting (one hopes) is a considered, rational, conscious (i.e. ‘system two’) act? Simple human irrationality can’t explain the error. What purpose would it serve people to lie to a census taker? Or is it being done subconsciously, without them realising?


A Voting Heuristic

My hypothesis is that there is a specific set of heuristics and biases that help to explain the incorrectly forecasted elections, which only kick in at the polling booth and which (I would argue) has been further strengthened by the unpredicted political events of 2015/6. Like most heuristics they are not open to introspection, so can’t be used to rationalise actions, and are ‘noise’ affecting decision-making (as Kahneman categorised heuristics in New York) that people may not be comfortable with in any case. And that’s why the polls don’t reflect them.

Let me explain.

Behavioural economics tells us that many behaviours are influenced by self-serving biases. We will often behave in ways that benefit ourselves and our ‘in-group’ (people like us), but our need to maintain self-esteem encourages us to represent our behaviour in ways that justify our actions favourably.

We would rather not confront the fact that we often behave selfishly, because it harms our ego and makes us feel bad (“I won’t give my spare change to that beggar because how do I know if he’s genuine? Besides, I give plenty to charity anyway”). Everyone likes to think they are behaving (largely) for the good of others, and we seek to rationalise our actions in ways that are consistent with that self-image. Because people who don’t think that lack empathy – which could be sociopathic.

Professor Dan Ariely at Duke University in the US (and colleagues) has conducted a huge amount of research in this area, especially on dishonesty and ‘self-signalling’ (effectively priming our own behaviour through projected self-image). As an example, committing one act of mild dishonesty (wearing counterfeit goods, for example) can then lead us to behave dishonestly in more overt ways (like cheating on a test). But we won’t acknowledge this when confronted with it, because we want to maintain a positive self-image, and will post-rationalise our behaviour accordingly.

In short, we all regularly cheat ourselves about our levels of selfishness, because it is critical to maintaining our self-esteem. We have all experienced occasions where we have been surprised by the levels of self-interest displayed by others (“who keeps stealing from the office fridge?”).

However, the national scale of elections only increases this effect, and as a civic duty (and a privilege reserved for democracies) the cognitive dissonance of behaving in a self-interested way is much more salient. And, as in so many areas of life, when that dissonance kicks in people ignore evidence contradicting that behaviour and misrepresent their intentions.


The (Self-Serving) Ayes Have It

I think it is possible to view many votes cast in the UK and US General Elections, and the EU Referendum, in terms of this kind of self-serving bias. Perhaps more so than any other elections in recent memory. Consequently that need to avoid cognitive dissonance based on preservation of our own self-image led many to (subconsciously) deceive the pollsters.[2]

At the EU Referendum, a vote for Leave was ‘Putting Britain First’ as the campaign slogan had it. The entire premise was that a vote for your ‘in-group’ i.e. the British, would put your interests ahead of others.

And in the US… Trump couldn’t have been more clear about his protectionist, pro-US stance. In fact it was the only clear policy intention. ‘Make America Great Again’ – if you are American, there couldn’t be a clearer appeal to the self-serving bias of putting America ahead of the needs of the rest of the world. We’re now seeing that manifested in his defiantly protectionist trade and immigration policies.

UK GE pic

The 2015 General Election had a bigger blue/yellow divide than ever before – and than predicted


The 2015 UK Election is more complex, but also explicable in the same way. The polling errors then were thus: under-estimated votes for the Conservatives in England (a recurring error across elections) and for the Scottish National Party in Scotland. The Conservative manifesto was largely based around cuts to public services that were going to disadvantage certain groups, and benefit others.

Post the independence referendum in 2015, the SNP manifesto commitments largely focussed on how they would seek to increase public services and defy these cuts. So Scots could confidently vote SNP safe in the knowledge that (a) it would not lead to independence in the short term, and (b) it wouldn’t mean the same cuts to public services that the rest of the UK would endure. A win-win if you were Scottish, but inherently self-serving for Scotland versus the rest of the UK.[3]

Consequently votes for both were self-interested to varying degrees, depending on the perceived ‘in-group’.

These contexts were more stark than they had been for some time, and consequently the potential for polling error was higher than normal (in my view). But the dissonance for a large number of those voters of voting to benefit their ‘in-group’ to the detriment of others may have been too much psychologically for many to bear, and they either misdirected the pollsters or genuinely couldn’t attribute their subconscious intentions.


But Why Isn’t Everyone Doing It?

But if we’re all subject to this kind of bias, why isn’t it manifest in everyone’s (stated) voting preferences? Wouldn’t it be as true of Remain/Clinton/Labour voters too? Well, maybe it is – it’s just that if the self-serving bias can be more easily post-rationalised as not being self-serving, it won’t create a dissonance (and therefore a predictive error). Some of the data on voting trends supports this.

The commonality in Leave/Trump votes is that they are strongly correlated to level of education. In short, if you had higher education of some form you were much less likely to vote for either (hence all the discussion of ‘intellectual bubbles’ and ‘echo chambers’).

Therefore a difference in ‘in-group’ and ‘out-group’ distinction derived from education level may determine the nature of the bias. If education and the pursuit of knowledge is about exploring and understanding things that sit outside your immediate environment, for the university educated, your ‘in-group’ is wider. In a literal sense, going to university often means leaving your immediate vicinity, friends, family, perhaps even going abroad. Those whose interests you hold dear are more likely to be from farther afield, and your concept of ‘self-interest’ is likely to include people from across Europe, or outside the US. It is well known that the majority of Americans don’t hold a passport – I would love to know the stats on how that differs amongst Trump vs. Clinton voters.[4]


We know how you voted…


Thus a higher educated person, voting based on the basis of a self-serving bias, would be more likely to confidently state a preference for Clinton or Remain without dissonance – as it would still benefit their ‘in-group’ but not be overtly detrimental to ‘out-groups’. Because their ‘in-group’ group is much more likely to include people from outside their immediate vicinity.


So what do we do about it?

In summary, I think it is as simple as this. Everyone largely votes in their self-interest. But we maintain an illusory visage that we vote for the benefit of others, to fit in with our own self-image as good citizens. But for some, the cognitive dissonance of voting in a way which is patently self-interested leads us to misattribute our actions, be it through post-rationalising, or (in a minority of cases) misleading others on how we will vote. Hence the polling errors – and they have been magnified by the specific contexts of the recent elections.

Will this problem persist, or is it a 2016 phenomenon? Whatever the merits of increased globalisation, the world is getting smaller and consequently our exposure and association with other cultures one would assume should (gradually) increase. As that perceived in-group increases in size, one could assume the scope of the self-serving bias becomes wider and more inclusive. But if politicians continue to pursue protectionist policies that dissonance may only get greater.

By contrast, recent events in the Netherlands perhaps also indicate that when that dissonance is made plain, then behaviour may change accordingly. For the upcoming UK general election, the key policy ‘battlegrounds’ will determine whether this bias comes into play. If it does become a second Brexit vote as some are predicting, we may see some interesting effects – under-representation of votes for the Liberal Democrats for example, as they are (currently) the only major overtly pro-Remain party.

One thing is for certain: recent events reinforces the importance of recognising, and accounting for, the subconscious drivers of human behaviour – and how we survey them. Simply asking people how they intend to behave is not going to be effective, and we need to use other (more implicit) techniques to actively identify how they will behave on election day. At CSG, we use these techniques to more accurately gauge how people will behave, and (where possible) assess and test actual, rather than claimed, behaviour.

Importantly, it also emphasises that the only way to avoid the ‘echo chambers’ that lead to the perception of common consensus (when none exists), we need to respect that others’ behaviour may be influenced by a completely different context.

As this guy knows well.



[1] An additional, sobering, fact that their model threw up: if the vote were only conducted by white men (as it was before 1920, with black voters only being enfranchised in 1965), then Trump would win all but 10-12 states.

[2] There is a (valid) counter-argument that the very act of voting itself is a denial of self-interest – in that people vote for who they think is best at running the country, rather than who is going to best serve their personal interests. Whilst some may vote on this basis I would suggest this is naïvely optimistic at best, and is a more accurate reflection of post-decision rationalisation than most people’s real motivation for voting.

[3] Though of course this has now changed post-Brexit, as the SNP are now actively pursuing a second referendum on the basis of Scottish opposition to leaving the EU (Scotland voted 62-38 in favour of Remain).

[4] There’s also potentially a ‘mere exposure’ effect going on here i.e. simply meeting people from outside the US or from Europe may influence your behaviour accordingly, by generating more empathy with those groups.



Thoughts from New York(2)…The Trump Bias

20 Jan

Today marks the inauguration of Donald Trump as the 45th President of the United States. It’s also just over three months since I attended the Ideas42 Behavioural Summit in New York (this being my somewhat delayed second post related to that event).

It’s taken me a while to write this post, because there have been more (literally) profitable things keeping me busy (more on that to follow). It’s also gone through a LOT of edits. I guess it’s also taken me a while to process the election outcome, given how unlikely it seemed in October.

And I don’t think I’m the only one. At the conference, there was inevitably a lot of talk about the election – on stage and off. But by far the dominant tone of that discourse was: “Trump can’t possibly win…can he?”

Leaving aside all questions of ‘echo chambers’ and ‘intellectual bubbles’ – from a behavioural viewpoint, what made Trump electable, especially considering a majority considered him “unlikeable”? And was he consciously using psychological or behavioural techniques to persuade voters? Daniel Kahneman (father of Behavioural Economics, Nobel Prize winner and featured prominently on this blog before) gave a slightly cagey, but still revealing interview to Nate Silver at the conference where he suggested “Trump has a very permissive system 2” – which is a great academic ‘dis’ if ever there was one. In effect, he was saying he has no filter between his brain and his mouth. That certainly seems borne out by a lot of the evidence (particularly his Twitter outbursts).*


Trump attacks Streep on Twitter. Because he’s got nothing better to do.

But whether it was conscious or not, he’d clearly tapped into a compelling vein of motivational rhetoric. Even if a lot of it didn’t make rational sense. Kahneman talked about an issue like climate change being one where “evidence alone is not persuasive”. This is because “our brains are not geared for an event like climate change”, where the issue is not perceived as urgent because of our biases towards short term gain. One could easily see the ‘hot button’ topic of immigration in a similar way – most evidence suggests increased migration has a net positive effect long-term (this Freakonomics podcast covers the issues well). However, the short term impact on shared resources could be the bigger driver of voter behaviour (it brings to mind a lot of the shameful racism directed towards Britain’s ‘Windrush’ immigrants in the 1950s, despite active government encouragement of the capable workforce coming to help rebuild a country ravaged by WWII).

Irrespective of intention, what else made him vote-worthy, in a psychological sense? In New York Kahneman categorised (some might say controversially) biases as “a form of noise, in that they are things that affect your decision-making that you would not want to be affected by.” That partially explains the appeal, given some of Trump’s policies seemed at face value to be patently racist or xenophobic (e.g. registration of all Muslims, building the Mexican wall). It would undoubtedly be wrong to assume that all 63 million Trump voters are simply racist. But if racial prejudice were an unconscious bias affecting voters it would explain why so many felt comfortable post-rationalising those policies as potential outcomes of a Trump presidency.


It’ll probably need to be bigger than this, fellas…

I’ve spoken here before about the challenges of encouraging voter turnout, and in particular the phenomenon of what I call the ‘meh’ segment (people who are so disengaged with politics they couldn’t be bothered to turn out to vote, even with significant incentives such as the threat of a fine). I discussed this with Christopher Mann (Assistant Professor of Political Science at Skidmore College) in New York, who presented compelling research on the most effective framing of incentives to encourage people to vote. He agreed that there is a sizeable proportion of people for whom politics will never enter their list of priorities. Maybe Trump’s unusual approach successfully engaged with these people, though I’ve not seen compelling evidence yet as to how many Trump voters were infrequent or first-time voters (though turnout was generally higher in states where he won). But other research demonstrates that for some people there is more ‘noise’ than others, that may lead to more biases coming into play.

One circumstance that generates particularly loud ‘noise’ is poverty. Professor Eldar Shafir from Princeton presented compelling research demonstrating that for those in poverty divided-attention tasks are much more challenging. Professor Anuj Shah from Chicago Booth then supported this with his own research which showed that people subject to scarcity of resources (i.e. the poor) are more likely to miss healthcare appointments, which then leads to worse health outcomes. When given a decision to make in a health based scenario, for the poor money is much more top-of-mind. This may seem obvious, but in this context it means that simple lack of money leads to far worse decision-making, and a vicious cycle of worsening health outcomes. A treatment should of course primarily be evaluated on its ability to help the patient, not it’s cost, but this additional ‘noise’ created by poverty makes that simple decision much harder. The natural consequence is a reinforcement of pre-existing social status (i.e. the poor get poorer, in both health and wealth). Interestingly they found that when running this experiment in the UK (where universal healthcare is available – for now – via the NHS) these results did not hold.

Many have questioned the logic of the disadvantaged in America voting for someone like Trump who (on the face of it) would seem likely to make the lot of the poor worse, not better. This may partially give an answer, if we can hypothesise that poverty leads to generally worse decision-making on areas relating to personal welfare. I think that’s a fair assumption given that a key Trump policy was the reversal of the Obamacare program which ensured millions of poor Americans now had access to affordable healthcare (as was eloquently demonstrated to Republican Paul Ryan recently). **

However, the evidence is not that Trump won the election because he was voted for by the poor (at least, not exclusively) – the biggest indicator seems to be level of education (specifically college/university qualifications or above). And he won votes in areas that were most at risk of economic downturn, even many areas that had previously voted Democrat. Education level also was a predictor of ‘Leave’ voters in the EU Referendum, and in that sense at least Trump was right to call the US Election: “Our Brexit… plus, plus, plus.”


Because nothing says ‘populist heroes’ like a solid gold elevator…

Whilst correlation does not equal causation, this may all help us with an answer as to what was driving Trump’s appeal, on a psychological level. The level of ‘noise’ around this election was perhaps greater than we have ever seen before (as Nate Silver put it in New York: “the Overton Window has been widened”) and as such the susceptibility to biases and heuristics (such as racial stereotyping) would have been greater. Consciously or not, Trump and his ‘post-truth’ assertions were the perfect vehicle for that environment.

In short, if you want to ‘win like Trump’:

  • Identify the places and people where the level of psychological “noise” is greatest – those under the greatest stress financially or academically, or areas which have seen significant recent changes in circumstances;
  • Focus on those voters’ short, not long, term fears (in academic parlance, exploiting our biases towards loss aversion and hyperbolic discounting).

And, of course, it also helps if you can win without getting a majority of the votes



What all this doesn’t fully explain is why the opinion polls didn’t take this into account and were so incorrect – not only about the US Election, but the EU Referendum, 2015 UK General Election and numerous votes in recent memory (e.g. the Israeli election). Whilst Behavioural economics tells us our behaviour is not always open to introspection, which is the flaw in traditional polling, why has this inaccuracy seemingly become so much worse lately?

I have a related theory on this, also clarified by some of the discussion at the conference, which I’ll talk about in my next post. In the meantime, I’d love to hear your views on ‘the Trump bias’.



*Coincidentally on the day of Trump’s victory I was at a conference with a BBC journalist who had interviewed him in the nineties. “He was very charming,” she recalled. And less surprisingly: “Smarter than you might think.”

**Interestingly, research from Paul Niehaus (co-founder of GiveDirectly and associate professor of economics at UC San Diego) found that poverty itself was more successfully addressed through their experiments in Africa by (a) transferring money directly to individuals using mobile technology (rather than via an intermediary, or NGO), and (b) doing so in a lump sum annually or biannually (not monthly). Contrary to popular belief, this money was not spent frivolously or instantaneously on non-essential items but the majority invested wisely, for the long term. It will be interesting to see if a Trump administration would attempt a policy initiative along these lines. But I won’t hold my breath.

Thoughts from New York (1)… How to motivate people

8 Nov

This is the first of three posts recapping some of my key takeaways from the Ideas42 Behavioral Summit I attended last month. A warning: this is quite a long post, possibly best read in sections and divided accordingly. As I said previously – there was a lot of good stuff at this conference.



Very broadly speaking, there are two ways to change behaviour. You can motivate people to change, or make something easier (or harder) to do.

A number of behavioural models (such as that of BJ Fogg at Stanford University) are based on this premise. In essence, once motivation and ease reach a certain critical mass, all you need is the correct ‘trigger’ to generate a change in behaviour.

The trouble is, behavioural science tells us that motivations are often irrational or counter-intuitive (for example, it’s generally more effective to pay people based on the inputs, not outputs, of their work), and that ‘ease’ is not only physical (the gym is too far away, so I won’t go) but psychological (I don’t understand how the gym equipment works, so I won’t go).

For example, how would you classify the famous fly in the urinal from ‘Nudge’? Does this make people more motivated to pee accurately (no), or make it easier to pee accurately (not literally).


The answer lies in the concept of ‘cognitive ease’ i.e. things that require less mental effort are easier for us to do, and are what we are naturally predisposed to do (appealing to system 1). It’s often referred to as ‘going with the grain’ of our behaviour. In this case, men are pre-disposed to ‘aim’ whilst peeing, and so having the fly there makes it psychologically easier for men to pee accurately.

A lot of behavioural science research and literature has focussed on ‘cognitive ease’ as a source of nudges, but relatively little on the motivation side. That work that has (such as the work of Dan Ariely) has tended to focus on the negative consequences, such as disincentives and dishonesty. Having worked on encouraging positive behaviours for a number of years (particularly quitting smoking) it was refreshing to see a number of practitioners talking about motivation in terms of encouraging positive behaviours at the  conference.



So how does one take account of our irrational drivers to encourage positive behaviour? One interesting new field of research is ‘temptation bundling’. This is something that most of us practice instinctively (I think): incentivising people to complete a necessary, but not necessarily enjoyable behaviour (such as chores) by coupling it with a pleasurable reward. Certainly it seems to be an important weapon in the arsenal of parents trying to get kids to tidy up, complete homework etc. I tend to bundle my chores (putting the bins out for example) with the tempting reward of quality time with my Playstation.

Assistant Professor Katherine Milkman from the University of Pennsylvania talked about her fascinating experiments in this field (you can hear more about it in this Freakonomics podcast). In her case, participants were rewarded for attending the gym by getting chapters of an audio book of popular novels such as The Hunger Games. She concludes that three things are particularly effective at boosting motivation: temptation bundling, prompting planning (i.e. getting people to plan out a behaviour in advance, as was very effectively used by President Obama’s campaign team to get people out to vote), and making use of ‘fresh starts’ (such as major lifestyle changes). This is because it is easier to break multiple (bad) habits at once.



Recent advances in digital technology have provided new tools for boosting motivation, as several speakers demonstrated, although they have to be carefully employed to be effective. Jordan Goldberg is the CEO of StickK, an online motivational tool that allows you to make monetary commitments to reward achieving goals. He talked about the important difference between extrinsic and intrinsic rewards – the evidence being that intrinsic rewards drives greater sustained behaviour change. The site has hard evidence of the effectiveness of disincentives in particular (loss aversion means that imposing financial penalties for not achieving goals is more popular than rewards), and Goldberg revealed the most popular list of hated charities that people can choose to donate to as punishments (see the drop down menu below). He didn’t reveal the ranking, but mentioned that a certain Presidential library in Texas has apparently done extremely well out of their users.


On the same panel Kelvin Kwong from fitness tech company Jawbone (an amateur magician mentioned in my previous post) talked about the importance of trigger points. Jawbone found that timely reminders (through push notifications on mobile) increased rates of going to bed on time by 28%. This is certainly something borne out in my work too – timely reminders were a key functionality we built into My Quitbuddy, the world’s leading quit smoking app that my team and I built for the Australian Federal Government (see this post for more). Another key motivational tool we employed, also used by Jawbone, is providing salient information on progress via the app (e.g. the person icon on the homepage changed colour the longer people are smokefree). Kwong confirmed that their research showed that it was when progress shown and felt by users that motivation spiked.

An interesting (but unanswered) question put to Kwong was whether actually mobile interventions are less effective amongst heavy mobile users/those with many apps installed (because of increased “noise”). This data exists (it is possible to target mobile ads based on the number of apps installed) so I’d love to know if anyone has done this analysis. My suspicion is it doesn’t make a difference – certainly our heaviest users of My Quitbuddy were younger audiences who tend to have more apps installed.



Following this discussion was a brilliant keynote by Adam Grant, Wharton Professor of Management at Princeton. You can see a very similar talk by him here, about the key themes in his book ‘Give and Take’. His work is hugely informative on whether people are motivated differently in the workplace.

Yes, he argues, and you identify people as three broad types: givers, matchers and takers (categorised based on the time and resources they demand from others).  His research shows givers as the worst performers in terms of productivity, but that they add the most value overall (they tend to occupy the top 25% and bottom 25%). However, matchers tend to have the greatest long term value, because givers eventually get tired of being exploited (understandably). Takers are generally bad news – intrinsic takers are sociopaths, and they rationalise their behaviour by assuming the same (bad) standards of everyone else.

So how to use this information? Grant advises three things:

  1. Selection decisions matter. You can distinguish between an ‘agreeable’ taker and a ‘disagreeable’ taker in an interview, for example, by asking them to name people whose careers they have fundamentally improved.
  2. Changing your reward system. Specifically providing collective, not individual incentives, as the evidence is these are much more effective (and they discourage ‘taking’ behaviour).
  3. Encourage help-seeking behaviours, through allocating time for helping others and allowing time for individual projects (as Google do with their ’20 percent time’ employee program).



The last speaker over the two days was Professor Angela Duckworth from the University of Pennsylvania, whose work looks at the importance of grit and self-control. She referenced research which has identified self-control as being critical to academic achievement and better life outcomes more generally. The famous Stanford marshmallow experiment is one example – children who were able to resist eating one marshmallow on a table in front of them for the promise of two marshmallows later generally out-performed those who caved in to temptation.

Duckworth explained this in terms of strategies that individuals employ to boost self-control, some of which I’ve discussed already (e.g. planning prompts and temptation bundling). Others include ‘situation modification’ (e.g. having a dedicated space or ‘zone’ to do work in, or avoiding people or places that encourage bad behaviour), ‘selective attention’ (e.g. such as using internet restriction apps to prevent distraction), and ‘cognitive re-appraisal’ (effectively, using a negative event to spur one on to do better next time). These can be self-taught, and consequently become habitual (i.e. instinctive, system 1 behaviours), which explains why those with greater self-control do better academically.



So what can we conclude?


Overall, motivation is a fickle beast. The best analogy I can think of is to the wind – some days it’s stronger than others, and some places (like individuals) get blown about more than others. But if we can identify those times and places where it blows strongest, we can make best use of it – and even give it a bit of a boost when required. Such as:

  • You can motivate people to do a non-pleasurable task by bundling it with a pleasurable one.
  • People are more motivated to change behaviour if they are changing their routines more generally following major events in their life (e.g. moving house), and if you get them to plan out the behaviour in advance.
  • Use loss aversion to frame intrinsic rewards i.e. demonstrate what people will personally lose by not completing this task.
  • Time your task reminders to when they are most salient, and show progress towards the goal.
  • In the workplace, build teams of ‘givers’, encourage giving behaviour and reward them accordingly.
  • Use whatever works to boost self-control, and keep using it until it becomes a habit.



One important (and somewhat depressing) final point re: motivation. There is one group who face fundamental challenges to motivation: the poor. Professor Eldar Shafir from Princeton University and Professor Anij Shah from Chicago Booth, whom I will talk about more in my next post, talked about how those in poverty are subject to greater ‘noise’ in their decision-making. In effect, scarcity of time and money occupies the mind making many tasks more challenging, and lead to more erroneous decisions, creating a vicious cycle when it comes to healthcare, education and other critical decisions on wellbeing. Motivation for the poor for (perceived) non-critical behaviour change is also less. Certainly this is borne out by my own work in stopping smoking, where the ‘busy poor’ (typically those in low-skilled, low paid jobs with multiple dependents) tend to find it the hardest to quit.


Behavioural Science…or Magic?

24 Oct

I was recently fortunate enough to attend the first Ideas42 Behavioral Summit in New York. It was an extremely well-organised and inspiring event, and gave a great opportunity to meet like-minded practitioners from a wide-ranging field. You can read about it here –  it was an interesting counterpoint of US best practice in applied behavioural science to my own experience in the UK and Australia, in both public and private sectors.

Interesting to me too was comparing the event itself to the UK’s largest behavioural science “festival”: Nudgestock. I guess I’m uniquely positioned to do this – one of my responsibilities in my previous role as Head of Ogilvy Change was for me and my team to deliver that event in June this year. I was the host for the day, Ogilvy UK Vice-Chairman Rory Sutherland being chair. The videos are all up on YouTube, including my brief pitch of Ogilvy Change. It serves as a pretty good summary of what I did in my short time there, including establishing the core products, developing the new branding, as well as generating over 20 new business leads. Oh, and managing the team’s delivery of Nudgestock. It was a busy (and productive) time.

Overall, the feedback was generally that this year’s Nudgestock (the fourth) was the best yet. Certainly I was proud of what we achieved. The tone was fun (think of a more irreverent version of TED), the speakers were informative and entertaining, and we even managed to run to time despite Rory’s reputation for “flexible” time-keeping. Also we were able to accommodate more people than ever before, including Ogilvy staff, charities and students at subsidised rates thanks to the sponsors I got on board (the first time we’d done so).

That talk I gave was one of my last acts in the role, which in itself tells a story. It was a little more…functional…than the conference speeches I normally give, and some feedback I received afterwards (both positive and negative) was that it was a bit of a contrast to the other speakers. Being either academics or writers/practitioners, the other speakers tended to focus on the theoretical rather than the practical (and had leeway to be a bit more fun), whereas my role was purely to articulate “how to apply this behavioural science stuff most effectively”. It was for the benefit of clients (or potential clients) by and large.

To return to my point, aside from my talk what we didn’t do at Nudgestock (that the Ideas42 team did brilliantly) was have a wide selection of real world case studies from people ‘at the front lines’ of applied behavioural economics/science. This was deliberate – one of Rory’s aims for Nudgestock is to expand knowledge and awareness of the field rather than any specific organisations, so the focus is necessarily broad – but driven by the fact that application in the UK is still more niche than in the US. We showcased some great work of practitioners via the ‘Nudge Awards’, but most of these were relatively small-scale interventions.

Given that the USA can perhaps be considered the birthplace of behavioural economics (although Israel has a strong claim based on the heritage of Daniel Kahneman and Dan Ariely, but both have lived and worked in the US for most of their careers), this perhaps seems inevitable. Speaking at the Ideas42 conference were representatives from companies such as AirBnB, Microsoft, Domino’s and the Boston Red Sox, which I think it is indicative of the difference in profile of applied behavioural science in the two countries.

In many ways that’s a positive thing, I think. Certainly for me in my new venture in applying behavioural and marketing science for organisations being able to point to the fact that leading companies and governments around the world are using these insights to change behaviour successfully is a great advantage. There’s a significant social norming effect in being able to point to a Microsoft or AirBnB and say: “They’re doing it. Why aren’t you?” And if no-one else in your market is doing it, that’s a competitive advantage.

I’ll go through some of these examples (and some theory also) in the three posts that follow, which summarise some of my thoughts from the conference. I hope you’ll find it interesting (and useful). I’ve arranged them into three broad themes:

  1. Motivating people
  2. Behavioural science in the real world (especially the upcoming US Election)
  3. Nudging through technology

One further thought in conclusion. One commonality between the Ideas42 event and Nudgestock were speakers who were also professional magicians (Kelvin Kwong from Jawbone and Adam Grant from Wharton in NYC, and Paul Craven at Nudgestock). There’s a natural synergy between magic and behavioural economics I guess – both rely on misdirection and understanding how our subconscious influences our decisions. And to an outside observer, they can sometimes seem impenetrable and baffling.

Taking the complex and theoretical and making it practical and useful has always been important in my career, and I believe is what those of us who work in roles with ‘strategy’ in the title should strive to achieve. It’s certainly what my new business is all about.

Sadly, there are too many in the communications industry who don’t seem to appreciate this. They either view behavioural science as too theoretical to have value, akin to magic tricks, despite being wholly grounded in how people actually behave rather than how they claim to. Or they simply don’t believe that it can generate the significant effects that have been repeatedly generated with seemingly such little effort/budget (but a bit of knowledge). Like people who dismiss magicians, mind-readers, and mentalists – such as Derren Brown – as simply being in league with the devil.


As an example, just prior to Nudgestock I was discussing how to use behavioural science to benefit clients at a senior level meeting. During it, someone said to me:

“This stuff is very clever. But as my grandmother used to say: ‘You don’t get anywhere by being too clever.’”

I was reminded of this whilst at the conference. As I sat listening to thoughts from Nobel Prize winner Daniel Kahneman, multiple New York Times bestselling author Daniel Pink, and the respected US election forecaster Nate Silver (seen here talking to Stephen Colbert about the latest US Presidential debate), it kept popping into my head.

“They seem to have done all right,” I thought.


A quick update

23 Oct

So…It’s been over 18 months since my last post, and for me quite a lot has happened. In short:

  1. I left my job at Match Media in Sydney in July 2015 as my wife and I decided to move back to the UK from Australia;
  2. I got a job almost immediately to be Head of Ogilvy Change, Ogilvy and Mather’s behavioural practice co-founded by Rory Sutherland (mentioned in my last post);
  3. After a successful few months there, I left that job shortly after we hosted the 2016 Nudgestock Festival (and immediately after the Brexit vote, go figure), and was on gardening leave until September;
  4. I’m now working with a number of different partners and agencies, as well as starting my own consultancy business.

As the legal cliché goes: those are the facts. I give them briefly here to bring you up to speed, and give some context for the posts to follow. I’ll be fleshing out a few details and returning to my theme of the lack of a scientific approach in the communications industry in general (as described here and here) as it help explains much of these events, as well as talking about some of the interesting stuff I’ve worked on, seen at events, talked about with smart colleagues and friends, and stolen from the internet.

All in all, it’s an exciting time for me and I’m very optimistic about the future. Not least being free to do great work with great people, without the structural constraints of the communications/advertising industry (see posts passim).

That oil tanker is VERY slow to turn…

The only certainty is uncertainty

22 Feb

At the MSiX conference back in September I had the pleasure of seeing Rory Sutherland speak. I’ve mentioned Rory (and the conference) here before – he is probably the marketing industry’s foremost (and most eloquent) champion of behavioural economics, and of how a more insightful understanding of consumer behaviour can transform the industry. By virtue of his immense knowledge and wit he’s one of the most popular TED speakers of all time (check out his greatest hits here).

One of his most commonly cited examples of communications combating an irrational human behavioural trait is the dot matrix boards on the London Underground, which show how long it is until the next train arrives. This short animation illustrates his description of why this has been so effective in combating consumer unease and uncertainty, with amusing contrasts of how this has been applied in South Korea and China.

The ultimate calming device?

The ultimate calming device?

The conclusion is we hate uncertainty about the duration of a wait more than the act of waiting per se, to an irrational degree. Hence why customer service helplines now tell you how many people are ahead of you in the queue, or give you an estimated wait time, rather than leave you in purgatorial limbo.

I was reminded of this in a recent client meeting (the principle, not purgatorial limbo – though I’ve had meetings like that too). We’ve been working with a tech start-up company who have developed a great taxi booking and payment app (not one of the dodgy, semi-legal ones you may have heard about). I’ve had the pleasure of working with a few tech and app companies now, as well as spending the day at Fishburners (a tech start-up incubation project), and working with such dynamic innovators is really inspiring. Coming from a tech or development background means they have a great appreciation of the value of data and testing in understanding consumer behaviour – regular readers will know this is a passion of mine.

Anyway, we were being shown a new product in development – a dashboard that enables businesses (such as restaurants) who make multiple taxi bookings simultaneously to track the progress and location of each booking. The functionality looked great, and testing has been going well.

However – not for the reasons you might expect.

The developer told us that he’d just been to see one of the restaurants testing the app. He asked the head waiter for feedback on what had happened the previous Saturday evening, peak time for both restaurants and taxis.

“We had four tables ask for taxis at the same time, so we booked four. None of them turned up.”

The developer then asked what they thought of the app – understandably with some trepidation.

“It’s great, we love it,” he replied.

“But none of them turned up?” asked the developer, confused.

“Oh, taxis never turn up,” he said. “They see someone on the street on the way and pick them up instead. But now… we can see when they’re not coming!”

Mythical Viruses and Evil Queens – The Content Conundrum

8 Dec

“Let make this go viral!”

“This is highly shareable!”

Common cries heard in boardrooms up and down the land. I’ve certainly heard it said (I may even have said it myself, once upon a time). Ever since some babies roller-skated their way around the world, marketers have been chasing the dream of creating a piece of content that can be seen by millions without a dollar being spent on paid media.

And to the lay person, they seem to be succeeding. Every day we hear about a new content piece generating millions of views, and we see a new TV countdown show of ‘Top 20 YouTube Clips of Cats Riding Household Appliances’ or some such.

But here’s the thing. Viral videos are a myth.

Or at least the effects are wildly over-stated. Simply creating a piece of great content, shoving it on YouTube/Facebook and sitting back and waiting for the masses to share is likely to result in disappointment. Even if it’s a super-cute video of a hamster eating a tiny burrito.

And we know this because of science. Specifically, a rigorous study undertaken by Dr Karen Nelson Field and colleagues from the Ehrenberg-Bass Institute for Marketing Science, at the University of South Australia. The study is detailed in her book Viral Marketing: The Science of Sharing (you can read a good summary here).

I first became aware of the study when I saw Dr Nelson Field speak at the MSiX conference here in Sydney in September, and pleasingly have subsequently start to see the findings being quoted more widely, including by content creation specialists (most recently by Unruly, programmatic video distributors).

To summarise, the research found distribution is the most critical factor driving sharing of online video content. The quality of content itself (in terms of engagement) only gives an incremental benefit beyond the distribution and seeding of that content (i.e. how much money and effort you spend on generating views, as opposed to ‘organic reach’ or sharing).

After two years of work, five different data sets, more than 1000 videos, nine individual studies and with a large team of researchers Dr Nelson Field found that on average social videos were shared by around 1 in 24 viewers. And there is not much deviation from this norm, as you can see from the chart below (apologies for poor quality photo from conference, annoyingly the slides from MSiX still haven’t been shared…)

Viral Video Sharing Rates - Surprisingly Consistent

Viral Video Sharing Rates – Surprisingly Consistent

Surprisingly, even the multiple award-winning, and seemingly widely-shared, Dove Real Beauty Sketches had a ratio of 30:1 of views to shares – slightly worse than the average. The sixty million plus views were a result of extensive seeding and distribution – in much the same way as buying a load of TV spots would have done (though admittedly probably for a much lower cost).

But, you may ask, what of those content pieces that do perform better than average in driving shares? What determines this?

The key drivers seem to be twofold: emotional impact and social motivations (i.e. what impact does sharing the content have on the person’s desired self-image). In effect, does it move me and will it say something positive about me to my friends and followers?

And so we come back to our old friends system 1 and system 2 (see posts passim). System 1 being our instinctive emotional response, and system 2 more rational and considered – when we get a reaction from both we are more likely to hit the share button.

The research found that the content that most consistently drove greater sharing behaviour were stories of personal triumph (other creative ‘devices’ varied in effectiveness). These positive emotional responses led to greater sharing (although still only incremental) – a recent example being this video from Animals Australia (whilst the content was clearly created for a very low budget, I’d love to know how much they spent on distribution). I’ve talked here before about the ‘power of cute’, and this research seems to corroborate it’s power in generating organic reach.

Interestingly, there were slight differences in the emotional stimuli that impacted sharing behaviour in different countries. Unruly say that France was the only country where they have found negative emotional reactions to content prompted sharing more than positive emotions. Coincidentally, the very next day I met with Buzzfeed (the social content aggregator site) who told me they also found the most popular articles in France differed hugely from their sites elsewhere. Cuteness doesn’t play with the French clearly – they prefer to get their social kicks from gritty brutal reality.

(France appears to be psychologically unique in many ways – following on from my previous post about choice, I found this great TED talk from Sheena Iyengar where amongst other fascinating findings her research found the French are much more willing to delegate hard ethical choices).

Another critical finding was that the level of emotional arousal, and therefore ‘virality’, was entirely independent of the level of branding in the content i.e. how obviously ‘commercial’ it was, how many times the product is shown etc. Again, traditional wisdom is that commercially created video content needs to be more subtly branded (or completely devoid of it) in order to be effective.

So, despite all the industry brouhaha around “content” – are we to conclude that branded content works in exactly the same way as “traditional” advertising?

Most important is getting people to actually see the content. Tick.

Emotional engagement is required to get people to hit the share button – surely in much the same way as a great ad generates ‘water cooler’ conversation (i.e. traditional word of mouth). Tick.

And people don’t care if the content is heavily branded or not. Well, you very rarely see an unbranded ad, so…Tick. Full house.

Unruly summarise it as: “Distribution is king. And content is a highly emotional queen.”

In most parts of the world, that queen is inspirational – like our dear old Queen Elizabeth II.

In France, she’s much darker and manipulative. More like Cersei Lannister.

France's true queen?

France’s true queen?