Computers and artificial intelligence have come along at an exponential rate over the past few decades, from being regarded as oversized adding machines to the point where they have played integral roles in some legitimately creative endeavours.
Twitter has announced a new program to analyze some users' activity on the site to give marketers better information about how companies' tweets impact sales in physical stores.
The capability, which Twitter calls "offline sales impact," measures how certain users engage with promoted and organic tweets from advertisers, and then feeds that information back to marketers so they can see how their tweets may have affected in-store sales.
The program is based on a partnership with consumer data company Datalogix, Twitter announced Thursday in a blog post. Through the partnership, Twitter has already run initial tests for 35 brands spanning multiple categories of consumer packaged goods including beverages, food, wellness and household products.
Some of the first brands included Oreo, Wheat Thins and Trident gum. In those tests, followers who saw a company's promoted tweets purchased 29 percent more from that brand than followers reached by organic tweets alone, Twitter reported Thursday.
Starting out, the program will be open only to consumer packaged-goods businesses in the U.S.
Here's how it works: A measurement partner, or Datalogix, sends Twitter a list of scrambled email addresses and corresponding ID numbers that the social network then matches to its own scrambled, or hashed, email addresses that users have associated with their Twitter accounts. When there's a match, Twitter shares with its partner the matched ID number and information about activity on Twitter, like views of and clicks on the food brand's tweets, Twitter explains on its website.
The partner can then measure offline sales impact by comparing, say, the ice cream purchases of users who were shown the brand's content on Twitter with those who were not, Twitter said. The partner includes the aggregated, anonymous results of the analysis in a report for the brand.
Twitter faces ever-rising pressure to monetize its site. The site first launched promoted tweets in 2010 to place branded content from partner advertisers in users' streams. Still, more than 94 percent of retail activity still happens in the physical world, Twitter said Thursday, citing U.S. Census Bureau research.
"This presents an interesting challenge when trying to measure the impact of marketers' online advertising," said Ameet Ranadive, Twitter's product manager of revenue.
Twitter acknowledges that its offline sales analyses do raise some issues around users' privacy. The company has hired a consulting firm to perform an independent review of its practices, which found that Datalogix never sends Twitter information about individual users' purchases, Twitter said.
The reports that Datalogix creates for advertisers include only aggregated anonymous information, according to Twitter.
"We've taken extra steps to design our offline sales impact capability with privacy in mind," Twitter's Ranadive said.
Users can also opt out of Datalogix measurement studies at any time, Twitter said.
But in spite of the privacy considerations, the offline sales measurement will help to fill some informational holes for Twitter, said Brian Blau, an industry analyst with Gartner.
"Advertising platforms need to show how their services ultimately influence sales, and there haven't been many really effective services to do this type of offline tracking," he said.
"Having this data will certainly help advertisers better measure how effective their ad campaigns are on the sales of those products," he added.
Meanwhile, Twitter is continually trying to give marketers more effective ways to advertise on its site. In April the social network rolled out a feature to let advertisers place promoted tweets in users' timelines by targeting certain keywords or phrases.
The social network also has a television ad targeting program to let TV marketers push out branded content to people on Twitter if those users are tweeting about certain shows.
A recent Nielsen study suggests that Twitter's TV marketing program may indeed provide a revenue boost for the social network. The study's findings, announced Tuesday, showed that broadcast TV tune-in rates led to increased Twitter activity, and vice versa.
The television study may not be connected to Twitter's offline sales analysis program, "but I can say that Twitter is pushing the message that their service is a very effective platform for a diverse array of brands," said Gartner's Blau.
Twitter said it hopes to offer its offline sales measurement capability to more verticals and more geographies in the future.