As we head into the week of the general election the race seems to be narrowing. Just like the 2015 election there are various estimates of who will win what but the polls appear to suggest a hung parliament becoming far more likely than we would have imagined six weeks ago.
Possibly. But if I was a betting man I’d have a few quid on the Tories to squeeze past the all-important majority of 326 to retain control of the country. The reason I am so bullish about the prospects of prime minister May is not down to her leadership, or her policies. I have no clue about how any of those stack up.
I think the Conservatives might win a majority on Thursday because they are better at marketing. Specifically, digital marketing. More specifically, Facebook advertising. And even more specifically, because Tom Edmunds and Craig Elder are smarter than me, you and probably most other marketers in the country.
I don’t know Edmonds or Elder but if they are as clever as I suspect they are, here’s how they will win the election for the Conservatives.
Step 1: Spend it all on Facebook
You will have the press and PR teams in place. You’ll also be spending the usual amount on outdoor media and a bit of print. But that’s really there to send very vanilla, general messages to the electorate and misdirect everyone that this is the main campaign. The real money will go on digital and specifically onto Facebook advertising.
In 2015 the Conservative Party outspent Labour seven-fold on social media advertising and 50 times more than the Lib Dems. The Vote Leave campaign that secured Brexit spent 98% of its £6.8m budget on digital media (and most of that on Facebook); the same proportional spend should win the election once again later this week.
In the era of dark Facebook political advertising you can block your rival completely.
Step 2: Geo-targeting
There might be 650 constituencies at stake on Thursday when Britain goes to the polls but the reality is that 85% of them are locked up as safe seats. This election will be won or lost on the 100 or so constituencies where two or more possible parties could win on Thursday. The first and most important strategic decision is to devote all of your digital budget to this small slice of Britain.
It’s here that two of Facebook’s big advantages come into play – its geo-location capabilities and social ubiquity make it the ultimate constituency winning tool. More than 60% of the UK population check their Facebook page, but it’s the small proportion of swing voters doing this in the key marginal seats that matter.
In 2015 the Conservative Party was able to serve ads to 80% of Facebook users in key marginal seats according to Facebook’s own data. If the party can again reach 80% of the users in the 20% of constituencies that matter they will win a majority.
Step 3: Micro-timing
Similar to geo-targeting, there is a wealth of data to support the idea that most voters have decided who they will support many months before the election was even called. But that’s not the point. By the very definition of their decision, these voters should be completely ignored.
Around 20% of the voting public will make up their mind in the three weeks prior to election day. Given the close nature of the swing seats we are targeting in step 2 this makes this small minority of voters incredibly important. This election campaign might have run for two months, but the digital campaign should last about 20 days.
Step 4: Segmentation
Now we get to the good part. If Facebook has one over-riding advantage over all others it’s the ability to segment at an incredible level of granularity. This enables a smart political campaign to estimate both the likely voting intention of the Facebook user and the main psychographic drivers that do and not motivate their thinking.
This is the part that many marketers find contentious. They question whether Facebook data can really predict who you will vote for and what issues you care about. But they miss the manner in which psychographics are typically used.
It’s crucial here to combine your own large, representative survey of the British electorate (in which you ask for voting intention and general attitudes) with all the demographic and psychographic data points that Facebook collects for its users. By segmenting the electorate using your own data on voting intention and drivers, you can then extrapolate this small sample to the voters in all the swing seats you intend to target using Facebook’s granular user data to identify them.
So, for example, I might identify a psychographic segment called ‘Worried Brexiteers’ (aged 50+, male, married, children, home owners, did not go to university, religious, small business owners, car lovers) who usually don’t vote but, because of economic concerns about the implications of Brexit, are very likely to be considering a vote for the Conservative Party.
Or perhaps I uncover another group I call the ‘Skeptical Socialists’ who are traditional Labour voters (aged 30 to 40, graduate, tablet owner, in a relationship, recent home buyer, football supporter, runner, engineer or healthcare, mid-manager) who would never vote Tory but has deep concerns about the current Labour front bench and its ability to manage the country. Imagine 20 or 30 of these psychographic segments and the ability to identify them with relative precision across all the 100 marginal seats.
It’s suddenly possible to promote military spending to John Smith at 25 Eden Drive while telling Jane Jones at 24 you will keep a keen eye on the budget.
Step 5: Targeting
Traditionally, in an election campaign a party inevitably had to explain itself to the whole nation. With Facebook, however, all kinds of new possibilities present themselves. First, we can ignore the vast majority of voters, even in a target constituency, because we already know we have either got them or lost them. Instead we will target voters that are still likely to be making up their mind and still possible to convert to our cause.
But let’s not stop there. While we try to get undecideds to vote for us, we will be spending at least as much effort and budget on those that will vote for our rivals and persuading them not to bother. Much has been written about how the Trump campaign managed to persuade many Hillary voters to stay home in the US election last year.
A vote for the opposition not placed is just as valuable as one placed in your favour. In a change from the norm, expect the Facebook political advertising this week to be as much negative as positive because a significant proportion is designed to demotivate voters.
Not that you will see much of the Facebook advertising in question. Unlike traditional political advertising, which everyone gets to see and critique, Facebook ads are ‘dark’. That means other voters, other parties, even other supporters have no clue exactly who you are targeting with your ads. Never mind under-the-radar, Facebook advertising has no radar.
Provided you don’t share the various videos you are showing on Facebook (and, ever the pros, the Conservative Party has steadfastly refused to reveal how many videos it has made or what they consist of) you can keep your rivals and everyone else completely in the dark. By the time the full scale and complexity of your Facebook advertising becomes apparent the election will have been decided.
That cloak of darkness has another big advantage for a party with big funds and most of it devoted to a digital spend. You can outbid your rivals for the same voter and all they will know about it is that the price has gone up. They will not know who is doing the targeting or what you are saying to the voters they are trying to target.
Forget the ancient, outdated idea of political equality in which everyone gets the same allocation of party broadcasts. In the era of dark Facebook political advertising you can block your rival completely. You get to shout, they stay silent.
Step 6: Messaging
Remember those detailed psychographic segments we built back in step 4? Well now we get to use them. Facebook doesn’t just offer a granular approach to segmentation, it also allows you to message them in the same micro-manner. Remember the ‘Worried Brexiteer’? How about we serve him a digital video showing Jeremy Corbyn being soft and all over the place on Brexit:
Remember the ‘Skeptical Socialists’ that were worried about whether the Labour front bench was up to it? Let’s show them Diane Abbott making an arse of herself again and encourage them to stay in bed on 8 June:
The advantage of dark advertising should not be underestimated here either. Because no-one else sees the ads you are running you finally have a politically expedient way to say one thing to one voter while promoting another to someone else down the same street. It’s suddenly possible to promote military spending to John Smith at 25 Eden Drive while telling Jane Jones at number 24 you will keep a keen eye on all spending in the budget next year.
Step 7: Misdirect in victory
The next bit is crucial. As the dust settles on Friday morning and the champagne runs dry you need to keep your digital heads down. No boasting about your role in the victory. No conference presentations later in the year showing how you did it. You pack up your stuff, delete everything and rely on those impregnable walled gardens at Facebook to keep your success a secret. Sure, the Electoral Commission will be able to work out how much money you spent on digital media but no-one, if you keep your traps shut, will believe that it made the difference between success and failure.
They didn’t join the dots in the 2015 election from the massive digital difference between the Conservative Party and everyone else to the the subsequent ‘surprising’ victory. They missed it when Vote Leave spent almost every penny on targeted Facebook advertising and managed to sneak past the post and trigger Brexit. They still don’t think it worked for Trump despite the mounting evidence that without the digital difference he would have fallen short. And people won’t see it on Friday morning as long as we all point to the powerful message of the prime minister and join the choir of voices questioning how the polls could have got it so wrong. Again.
Of course, the one big caveat in all of this is the assumption that Labour and the Liberal Democrats have not mastered the dark arts of political advertising on Facebook to the same degree. Apparently the Labour Party has confirmed they have more than 1,200 adverts in circulation which suggest they too may be playing the Facebook card this time around. But my money is still on Edmonds and Elder and the likelihood that the Conservative digital competence of 2015 has become even stronger in the subsequent two-year hiatus.
But in the end nobody really knows who is using Facebook or what they are doing with it. I’d like to tell you that we will find out on Friday but, while we certainly will discover the election result, in all that darkness it’s going to be hard to know who did and did not use Facebook to their advantage.