In the aftermath of Twitter’s big data division GNIP announcing that it would no longer make firehose reseller agreements with third parties, Twitter has provided an update on how one of its firehose licensees is being affected. In a blog post in Japanese, Twitter confirmed that NTT Data, which resold Tweets and their metadata in Japan, will indeed no longer have firehose reselling rights come August (as we reported first over the weekend). But it also pointed out that the relationship is not ending altogether: the two are now working on enterprise services.
“NTT Data has been and will continue to be a very close and strategic partner in Japan,” the company wrote in the post. “Our two companies are working together now on transitioning NTT Data from being simply a data reseller to a far more strategic solutions partner based on the strength of NTT Data’s breadth and depth of expertise in social analysis, systems integration and business insights.” A full translation is below.
A Twitter spokesperson confirmed to us while NTT will no longer have the firehose to resell, it will transition into a partnership with Twitter along the lines of what the social platform has with IBM. That deal, originally announced in October 2014, was made to “let enterprises incorporate Twitter data into their decision-making through an established set of IBM tools, solutions and consulting services.”
The two finally turned on the first of those services in March 2015: a set of data analytics services and developer tools that run on IBM’s Watson artificial intelligence platform and its Hadoop platform, which will let developers write applications that utilise Twitter data.
The fact that NTT and Twitter are not completely ending their direct relationship in light of the changes to firehose reselling policies could have been one reason why Twitter continues to list NTT Data as a key partner on its site.
The reason that Twitter has provided an update today could have been to reassure NTT Data’s customers for Twitter data — which include NHK, RECRUIT Technologies KK, JR Souken Information Systems KK and the National Institute of Informatics Information Research Center — that they will not be left high and dry, and to make it clear both to Twitter’s and NTT Data’s investors that there will continue to be a connection between the two, along with all the value that this would impart on both companies.
It looks like DataSift does not have a similar kind of relationship to announce at this point. It has published a fairly blunt blog post outlining the differences between what it offers to customers versus what they will be getting as a result of DataSift losing its firehose deal. It also noted that the reseller agreement that Twitter had in place with DataSift charged $.10 for every 1,000 Tweets received through the firehose — business paid for what they used. On average Twitter’s users produce 500 million Tweets per day.
Twitter Japan’s blog post translated below:
In order to better serve our customers and partners after acquiring Gnip in May last year, we worked to bring all data licensing activity in-house. We began this process nearly a year ago and will conclude it in August.
NTT Data has been and will continue to be a very close and strategic partner in Japan. Our two companies are working together now on transitioning NTT Data from being simply a data reseller to a far more strategic solutions partner based on the strength of NTT Data’s breadth and depth of expertise in social analysis, systems integration and business insights.
We are very excited about the initiatives we’re working on together and will share those with you soon. In the meantime, our arrangement will not change the way NTT Data accesses Twitter data. We will work together to best support each and every NTT Data customer that licenses Twitter data, and our teams will reach out to these customers together to help them make best use of our new and growing solutions relationship.
The account transition from NTT Data to Gnip will be reviewed individually and each account is evaluated based on the use case of Twitter data.