2019 federal election: How Labor lost the unloseable poll

Labor's team is reviewing the loss to explain exactly how the party failed so abysmally.

The biggest disaster was inflated expectations. Wreathed in hubris, Shorten didn't even bother giving a National Press Club address in the last days of campaigning. The only other leader to do this in the past half-century was - you guessed it - Hewson. Both, also, stubbornly pushed new tax regimes and both were certain they'd be victorious. The electorate was as well. Feeling free to vote against change, people did. Voters punished an arrogant opposition.

Not caring very much for either Morrison or Shorten, people walked into the ballot booths, thought "yeah, nah", and voted (in record numbers) informal or for minor parties, before finally parking their vote, eventually, more or less, where they had last time.

Compared with the previous election Labor lost one seat, the government gained one, leaving independents controlling the Senate. Some "victory", so what's to see? How many reviews do you need to tell you a campaign based around an unpopular leader wasn't a good idea; that Treasury spokesman Chris Bowen shouldn't have tampered with Paul Keating's dividend imputation reforms; and that polling's more and more unreliable as people shift to mobiles.

Probe slightly further and one vital lesson does emerge. It should send shudders through both sides of politics.

What's changed, dramatically and irrevocably, is the way we communicate. This alters not just how we participate in society but something far more fundamental: its very nature.

Earlier, back in 2016, Labor thought it had a pretty sharp campaign machine. It did. The party had its demographic analysis down to a fine art, knew exactly which messages to send to particular electorates and how to target advertising to particular groups. Then, in the final moments just before the poll, it pressed "send" on the coup-de-grace: mediscare. These texts were just enough to shift wavering voters over the line and prompt votes against the Coalition.

It was a one-trick wonder. It provided tactical advantage but, not enough to push Labor over the line, set the stage for strategic defeat in 2019. Focus shifted to what Shorten would do if he became PM and the party's campaigning didn't quite hit the mark. But the key was how the Liberals managed to turn things around. This was quite extraordinary and new.

There were two vital contributions that ensured the coalition retained office.

First was removing Malcolm Turnbull. The polling was right: he couldn't win. Voter's had become utterly disillusioned with his government. But even though the Liberals weren't liked, Morrison was new. There offered space to develop a new campaign.

Morrison came from an electioneering background and possessed a mongrel desire to win. He was prepared to speak - at arms length - to Clive Palmer and harness the campaigning smarts of the polling and social research firm Crosby|Textor, created by former federal Liberal director Lynton Crosby and political pollster Mark Textor. We still don't have granular detail of how they managed their campaign. It seems, however, that they effectively inverted Labor's model.

Instead of chopping up electorates demographically, the internet now allows individual voters to be profiled. Thousands of data points (5000, to be precise) can be amassed, legally, on each voter. Detailed portraits are built up, beginning simply (sex; age; mortgage) but rapidly morph into massive databases.

These highlight potential pressure points which, when massaged appropriately, have the capacity to sway voters, shifting them one way or the other.

This can be particularly vital for the "undecideds", those voters who haven't really come to a decision about who they like least. As they stand with their pencils poised in the ballot booths, figuring out how to vote, they overwhelmingly listened to those carefully targeted messages identified through massive data-mining. Morrison's message hit home.

The Liberals' campaigning certainly became sharper once Crosby|Textor became involved. Interestingly, observers noted that Clive Palmer's spend also became a lot more focused as the campaign went on. Although he didn't manage to win a seat, his money resonated with a huge number of people and formed the basis for Morrison's victory. While Shorten and Bowen dismissed these voters' complaints, the Coalition was reaching out for preferences.

Despite the professionalism of the Liberal's campaign team, some individual Labor candidates still held on. In Eden Monaro, for example, Mike Kelly, managed to retain a naturally conservative seat through doggedly working the local communities. That wasn't even attempted up in Townsville, another bell-weather seat. Here the broader campaign dynamics swept out the hapless Cathy O'Toole replacing her with another ex-army officer, Phillip Thompson.

Electorates are now bigger than ever before with the last significant expansion of the Parliament in 1984. Back then there were 15 million Australians; today there are well over 25 million. Increasingly parties aren't bothering to reach them through the news media: they're reaching past the usual civic debate to profile individuals and message them through the so-called "social" networks.

The problem is these are focused on advertising. Should we really be surprised they offer up demographics in return for money.