This post is by Helen Rose, head of the7stars' Lightbox.

The Sun's 'Well Hung' splash on May 7 left no doubt that Britain was heading for a hung parliament by morning. According to polls, the UK was gearing up for the tightest election in decades. By May 8, the Conservatives had won by majority.

Pollsters are now facing a "post-mortem", launched by the British Polling Council, to determine why their predictions, which vastly underestimated the Conservative's vote share while simultaneously overestimating Labour's, fell so short of the mark.

A pre-election study by the7stars' research and insight division, Lightbox, of over 1,000 18-24-year-olds revealed similar results to early polls, with Labour coming out the clear frontrunners taking 30% of the millennial vote. The study also found 80% said they planned to vote, far more than the 66.1% who actually turned up on Election Day. In short, the polls across the board didn't come close to reflecting the actual results. So what went wrong?

They say hindsight is 20/20 and on cross-referencing the results of the study with what came of polling day we can identify a number of factors that may have caused polls to misfire this May. In the first instance, the nature of traditional polls does not correlate with how people think or act in today's modern, digital age. The information available to us and the means to access it are propagating at lightning speed. Newspapers and radio have been usurped by Facebook and Twitter – and our decisions are no longer built brick-by-brick over extended periods of time but can change shape as quickly as a newsfeed is refreshed.

Passive polling and consumer mentality

We shouldn't let the immediate cries of media bias distract us from the fact that when it comes down to it, people change their minds. There is also a stigma around certain parties, meaning their supporters are likely to be less forth-coming in admitting their political persuasion. In April this year, YouGov proposed the 'electoral phenomenon' dubbed 'the shy Tory factor', whereby Conservative voters won't admit their political preference, would potentially skew results on polling day.

Rather than relying on one black and white polling system, we should use technology to better understand voters' motivations and behaviours. Passive data collected from mobile and desktop activity, just as brands and tech companies collect from their consumers, would give a more accurate picture of where people will cast their vote.

A quote from David Ogilvy, which outlines why the consumer research industry needs to move towards more passive data measurement, has never rung truer than in the context of the pollster debacle: "The problem with market research is that people don't think how they feel, they don't say what they think and they don't do what they say."

People are becoming savvy at projecting an image of themselves that doesn't always fit with their actual values or behaviour. The past 12 months have seen the media industry finally adopt neuroscience research methodologies to gain a deeper understanding of emotional behaviours, while passive tracking can provide insight into how and where people spend their time even on a minute-by-minute basis, which has always been so tricky to accurately record or articulate.

People prefer personalities over parties

While the7stars' Lightbox poll found Labour to be the most popular party with millennials, its leader came only fourth in the ranks behind David Cameron, Boris Johnson and Russell Brand. Similarly UKIP as a party was much more popular than Nigel Farage.

If the pre-election polls had focused on the person fronting the party rather than the party itself, they might have had more success in getting to know where voters' hearts truly lay. Parties would do well to learn from consumer brand marketing strategies. For any brand promoting itself, the core personality of the brand and the image it portrays is just as important as the brand values it imparts.

Make it easier to vote

If the respondents of online polls could have voted via social media, would election results have been closer to those of the pre-election polls?

The election turnout this year was the highest since 1997 but still considerably lower than the percentage who said they planned to vote. Given more than half the 18-24-year-olds in the7stars' Lightbox poll said they would like to vote online, with Facebook as the most popular platform, it may be that more modern voting methods are required as well as upgrading polling methods. Whether through the Conservative's investment in Facebook campaigning or Ed Miliband's tête-à-tête with Russell Brand, this year's election was allegedly centred on young voters. Perhaps it's time therefore for a rethink in terms of how young people are able to voice themselves.