Days after Scott Morrison became Prime Minister, Liberal MP Tim Wilson called up with an idea. He wanted to be put in charge of the lower house economics committee - but on one condition: he would run a no-holds-barred inquiry into Labor's franking credits policy.
The recently dispatched Malcolm Turnbull had already branded it a "retiree tax" that would rob people of their savings after a lifetime of work. But the rhetoric had not cut through.
Along with colleagues Jason Falinski and Trevor Evans, Wilson wrote a proposal to smash Labor on the policy: to highlight human stories and link it to a wider narrative about Labor's fiscal management. It eventually became part of the Coalition's core message: "Labor can't manage money so they'll come after yours."
Of the myriad factors that probably contributed to Scott Morrison's stunning surprise victory on Saturday, franking credits might be the archetype. It underlined how Labor was exposed by a big, ambitious policy agenda on which the government could capitalise. It played into fears about change. It gave voters a selfish reason to stick with the devil they knew.
"They did have a very broad vision," says the Liberal Party's federal president Nick Greiner of his Labor opponents. "We believed our best chance would be to say to people: you don't want more taxes and more government, you want less taxes and less government. I think you can see from the result ... that's pretty much what happened."
Early in Saturday's count, as blood drained from the faces of the Labor faithful and their sure win became a horror show, Labor operatives started briefing journalists that they had been crushed by the Coalition's campaign on dividend imputation.
The powerful NSW Right faction - never happy about the show being run from Victoria, Bill Shorten's home state - was livid. They had long been worried about how the campaign was tracking. In seats like Macquarie, Reid and Dobell, the results seemed to prove them right.
"They ran a Victorian campaign nationally," says one NSW Right source. "But Victoria isn't north Queensland or Brisbane or western Sydney or the Northern Territory."
NSW conducted its own research that suggested topics such as electric cars were not hitting the mark. Labor's primary was also soft, a finding reflected in some of the published national polls, including Ipsos surveys in May which put Labor's primary accurately at 33-34 per cent.
There were big disagreements about strategy and seats. An "intransigent" national secretariat in Melbourne wanted resources thrown into seats such as Banks in western Sydney and Page on the north coast. NSW thought they didn't stand a chance. In the end, both swung to the sitting government member by more than 5 per cent.
Shorten's camp sees things very differently. They argue NSW was late to get into gear due to the March state election. That was a particular problem because Labor's path to victory was ultimately quite narrow: picking up five seats in Victoria, a couple in Western Australia and some in NSW: Reid, Gilmore and Robertson.
In the end, Gilmore was the only seat Labor won off the government that wasn't already a notional Labor seat. It was a crushing result worsened by the loss of possibly two seats in Tasmania, two in Queensland and a failure to realise anticipated gains in Victoria.
On Sunday, very senior Labor hard heads conceded their problems went well beyond the Sunshine State. They pointed to several big-picture issues with the campaign: they lost the advertising game, with the finger pointed squarely at party ad man Darren Moss. They had over-complicated their messaging; especially problematic given their complex platform.
[By contrast, the Liberal Party's relatively new go-to advertising gurus - Adelaide's KWP! agency - has now run ads in the winning South Australian, NSW and federal campaigns.]
There were also reservations about the performance of first-time Labor campaign director Noah Carroll, particularly over a confused communications strategy which did not always match paid advertising with what the party was talking about in the media.
But all these explanations are retro-fitted. The bottom line is that Labor did not expect the routing it received - its internal polling is said to have broadly matched the public ones - and the party is struggling to determine what went wrong and how to rectify it next time.
Meanwhile, the Liberals are surprised - but not entirely stunned - at their surprise victory. For months they worked on the basis that there is "no such thing as a uniform national swing", and focused on ways to get over the line in key seats: "guerilla warfare", as moderate powerbroker Michael Photios called it on Saturday night.
"You could argue that was the only way we could win," he said. "The government was not going to be swept out of office and neither was Shorten going to be swept in. In that context you move to a more sophisticated matrix which is seat by seat."
Photios said Labor, thinking it was destined to win, adopted a high-risk class envy platform when in fact "there was no one out there waiting with baseball bats" for the government.
The Coalition was tracking 20 marginal seats almost daily and oriented its campaign towards them with slavish devotion. Strong numbers in the key western Sydney seat of Lindsay put similar seats like Longman, Dobell, Bass and Braddon into play. Corangamite and La Trobe looked good too: the government ultimately lost the former but held strongly in the latter.
"On the marginals we were always ahead and the momentum was going our way," a senior Liberal source said on Sunday. The final week of the campaign proved especially positive.
Why was this not picked up anywhere else? That conundrum will be interrogated by party operatives, pollsters and commentators ad nauseam. But Liberals point to the fact their tracking polls are aided - meaning voters are asked specifically which candidate they will vote for, rather than just which party. It is seat-specific and more accurate.
Armed with this spreadsheet, the Liberals picked their battles and hoped for the best. By the end of the first week of the campaign, Morrison had clocked up his seventh visit to Tasmania since the start of 2019 - a good investment, it transpired.
Two weeks from election day, the Liberals polled the Brisbane seat of Longman and found it might be up for grabs. The retiree stronghold of Bribie Island was "on fire" over franking credits. The Coalition threw resources into area and ultimately won it off Labor with a 4 per cent swing.
The franking credits issue - along with negative gearing - had helped the Coalition craft what it called the "cumulative tax argument" against Labor: it was addicted to taxes and anything could be next. Labor-aligned economist Stephen Koukoulas called the government's victory a "huge result for Tim Wilson - the hero of the election for the Coalition".
On Sunday, NSW Labor types were closely examining the swings against them among over-60s. In dozens of booths across Banks, Robertson, McMahon and others, swings in that age group exceeded 7 per cent.
This helped the Coalition win the national two-party contest with 51 per cent of the vote - something neither side expected. Morrison is credited with relentlessly driving home the message, Shorten condemned for never connecting with voters in the way he needed to.
The results have Labor insiders scratching their heads about where to go from here. The huge swing in Queensland, always a volatile state, puts seats there further out of reach. And the failure to convert anti-Liberal sentiment in Victoria into significant gains means things will only get harder for Labor once things start to rebalance.
Labor figures reluctantly agree that no one will take another ambitious policy agenda to an election in a hurry; certainly not one that attempts to claw back entitlements or tax breaks.
The government, meanwhile, has carte blanche for the next term. As one jubilant Liberal MP observed at Saturday night's euphoric function: "We just ran on a strong economy - we've got a mandate to do anything."
- SMH/The Age