Donald Trump's Presidential approval rating in the US this quarter is at an all-time low at 35%, well below the historical average of 64%. But who believes polls any more?
During the election itself, Trump was laughed off as a no-hoper, a crass star of Reality TV who lived in cloud cuckoo land and spouted words like somebody who needed the care of the community.
Even so, Trump was elected and opinion polls were derided. With the advent of fake news and the polarisation of views along party and political lines, it's difficult to believe anything, not least predictions.
One data company, however, has consistently called global elections right and even managed to predict the Labour surge in this year's General Election.
Qriously is a London-based startup that uses mobile phones to create real-time surveys based on its access to a network of more than 50,000 apps.
Just before Trump's visit to Asia, the company released data gleaned from research conducted in Japan, South Korea and China. The company interviewed over 2,200 people - around 750 in each country.
One of the astounding results was that in China, Trump had an approval rating of 24%. This compared with 13/14% in Japan and South Korea, and approaching the 34% approval rating in the US.
Does this mean Trump's recent comments about Chinese leader Xi will play well domestically? Perhaps Trump's recent statements imply that the Chinese man in the street likes Trump's style and rhetoric more than expected.
Unsurprisingly, the numbers are very different for Kim Jong-Un, the leader of North Korea.
In Japan, 66% see him 'very negatively', in South Korea this is down to 48%; but in China it's only 18%, with a plurality of people actually having a neutral opinion on Kim Jong-Un. How the other half think.
Qriously's data has also unearthed that Chinese people are most likely to say that the North Korean situation is impacting their political allegiance The data says 41% 'Yes' as against 28/31% in Japan and South Korea.
So why should Qriously's data be trusted more than any other pollster?
Its recent record is impressive. Aside from the Labour surge in the UK General Election, the company also predicted Brexit, Trump's rise in vital swing states and other elections in France, Turkey, The Netherlands, Germany and (naturally) South Korea.
It workings are simple and focused on the easiest channel possible; mobile phones. Consumers are served a banner question instead of an ad on their smartphones, and if they choose to engage are offered a survey of between 15 and 20 questions they can choose to answer or ignore.
Their surveys have all the typical features of traditional market research projects, including randomisation, piping, skip logic and more. Qriously's data and predictions are already trusted by the New York Police Department and many of the world's biggest hedge funds and consumer brands.
"We think polling needs to evolve. The current systems of panel-based polling is increasingly being gamed by professional poll-takers and results are suffering because of it."
"Complicated correction factors often have to be applied to compensate for this. We're seeing our programmatic sampling methodology on smartphones produce much more representative samples and we're getting great results with minimal data treatment," said Christopher Kahler, CEO of Qriously.
While Qriously says that its polling is not its primary focus, it's an important path to credibility and vindicating the technology.
"We want to be known for getting things right. It's much harder to argue with an accurate prediction," adds Kahler.
It will be interesting to see if the company conducts polls when Trump gets back home and how they differ from the current data.
While Trump seems to have the approval of only a third of Americans, the Chinese aren't as far behind as thought possible. Japan and South Korea, however, is a different story.