Last checked at 
By David Ingram and Jeffrey Dastin
NEW YORK, Nov 10 (Reuters) - A handful of small publicopinion polling companies that bucked consensus and accuratelycalled the U.S. presidential election for Republican DonaldTrump are reporting being flooded with calls from investors andclients seeking their services.
Most pollsters wrongly forecast Democrat Hillary Clinton asleading Trump ahead of Tuesday's election in the latest fiascoto hit the $20 billion public opinion research industry, onlymonths after it failed to predict the British vote to leave theEuropean Union in a June referendum.
Among the few that got it right was a new industry playerusing a different method, South African firm Brandseye, whichanalyzes social media posts.
With offices in Capetown and Johannesburg, Brandseye took anentirely different approach from traditional polling.
The data-miner pays people around the world to sift throughsocial media for relevant posts, a process known ascrowd-sourcing, and then uses a computer algorithm to rateconsumer sentiment about products or politicians.
Its method pointed to a Trump victory. It also correctlycalled Britain's Brexit vote.
"My phone has not stopped," CEO and part owner JP Klopperssaid in a telephone interview on Thursday.
Kloppers said venture capitalists have been calling aboutinvesting in the firm, and polling companies have inquired aboutits methods. Kloppers said the company was open to licensing itstechnology.
Existing clients include ride-sharing firm Uber, banks,telecommunications firms and some government agencies, he said.He declined to identify the governments.
"Elections come every couple of years, but there arecompanies and governments that need to understand every day andevery week what's driving customer satisfaction anddissatisfaction," Kloppers said.
New methods, if successful, are of high interest to pollingcompanies that are having difficulty reaching cellphone users orsurvey-weary Americans.
"The answer is ongoing research," said Krista Jenkins,director of the PublicMind survey research center at FairleighDickinson University in New Jersey. "It's basically digging inas an industry and comparing the findings from your traditionaltelephone polls to data that is collected online."
POLLSTER ERRORS
Traditional pollsters base their results on questions posedto randomly selected people, often in interviews conducted liveover the telephone. Among other errors that pollsters made aheadof the U.S election, they almost universally miscalculated howturnout would be distributed among demographic groups.
High-profile London-based polling agency YouGov PLC's stock dropped 4 percent on Wednesday after its newmethod of surveying people online failed to accurately predictthe U.S. presidential election's outcome. Analysts were split onthe reason behind the drop, and YouGov stock partly recovered onThursday, closing up 2 percent in London trading.
No one from YouGov, which primarily does research forcorporations, was immediately available to comment on thebusiness implications of the polling miss, a spokesman said.
The Reuters/Ipsos States of the Nation project gave Clintona 90 percent probability of winning the 270 electoral votesneeded to secure the election. On its website, one of thepossible scenarios was how Trump could win the election byadjusting turnout levels. Ipsos SA will assess what itgot wrong, said Clifford Young, president of the French researchcompany's U.S. public affairs unit.
Election polling is generally not a big money-maker. Manypolls are done by universities or the media, while privateresearch companies use it to market themselves to corporateclients that provide more profitable work.
Pioneering U.S. pollster Gallup last year abandonedhead-to-head polls for the presidential race, saying it wantedto focus its resources on polls about issues.
The outfits that wrongly foresaw a Clinton victory willsuffer damage to their reputation, said Alex DeGroote, aLondon-based analyst at Peel Hunt LLP who follows the industry.
"They attach their brand to these polls. If these pollscontinue to flop, it's not going to help their brandcredibility," DeGroote said.
Among this year's winners may be Remington Research Group.The Kansas City, Missouri-based pollster correctly predictedthat Trump would win the election battleground states ofFlorida, North Carolina and Ohio, although it missed Trump'ssurprising Wisconsin victory.
The company, which frequently works with Republicanpoliticians, conducts a mix of automated phone interviews andlive phone interviews. It may have done relatively well becauseother firms were afraid of standing out from the crowd, saidTitus Bond, the group's director.
"A lot of pollsters, they want to be at the same number aseveryone else," Bond said.
Bond said he received three calls on the morning after theelection from corporate clients in the pharmaceutical andfertilizer industries and three others from advocacy groups, allseeking to do business.
(Reporting by David Ingram and Jeffrey Dastin; Editing by WillDunham)
