[ad_1]
The publisher Buzzfeed has rebuilt its badysis processes around the BigQuery data warehouse and Looker dashboards for consumption by reporters and editors.
Buzzfeed has been a big challenge to unify his content badytics. com on Facebook, Snapchat and Twitter.
Buzzfeed has nine billion content views & # 39; monthly on its platforms, including social networks, so a lot of data to aggregate.
Talking During Google Cloud Next '18 [19459010NickHardybadysted'BusinessReviewedBuzzfeeddeclared:"IncludingthatcompetenceoftheContentoftheContentisoftheChallengeofWewouldwanttoecrosstheappropriateAllInCompanyandOnly"
With Regard to This Growing Variety Of Plates -forms, Hardy admits that some play better than others when it comes to open data via APIs
.a data varies, we like to use APIs when we can but some platforms, without names of names, are not so beautiful, so we did some creative things to get these data, "he said.
Buzzfeed's challenge was to consolidate these data and to make them consumable for a wide range of users with different data skills.
The first step was to build a database data warehouse in the cloud, and then turn to Google's BigQuery cloud. Buzzfeed previously used Amazon Redshift and wanted to switch to BigQuery because this system "required a lot of maintenance and development resources," according to Hardy.
He also notes that the team had already chosen Looker as a downstream business intelligence tool. , mainly "because it gives us a lot of features and is well integrated with BigQuery".
Before Looker, Buzzfeed relied on home-made tools for downstream badysis. Hardy says they have struggled to evolve as content has become more distributed across channels. "
" It was becoming difficult to do dashboards using engineering capacity and we needed a better system to iterate faster. he wanted out of the box, with minimal development costs, including central access controls to give publishers access to the information they need.
Buzzfeed also liked that Looker came in two flavors: Explorer for hardcore users and Dashboard for simple high-level badysis
] Hardy explained, "Let's take questions about Disney as an example, if that's all right, how can we run that along the way?" Social media publishers were heavy users of these data, so we built dashboards to see what works well in real time. "
These content creators are particularly interested in" quiz complements ".
According to Hardy, the main lessons learned by the construction data team for non-technical users were twofold: "First, people want answers quickly.
"Looker allows us to open all the warehouse to people, but they did not want it, we have so simplified and deleted what users wanted with the option of scaling, rather than opening the valves. "
Buzzfeed now has 100 terabytes of data on 300 tables in BigQuery, Supports 100,000 Queries a Day Finally, Hardy said that Buzzfeed is an early access partner for BigQuery ML, testing the new machine learning tool to improve the performance of its recommendation systems. Buzzfeed sites and Facebook pages, as well as give an overview to social media publishers to help them optimize content distribution.
(Scott Carey Report, Computerworld UK)
Sign up for the newsletter!
<! – Leading provider of daily news on the computer channel, covering business, technology, products and services. ->
Error: Check your email address.
CloudGooglebadyticsBuzzFeed Tags
[ad_2]
Source link
