4 ms·

"It's not "get data on it" that's being flagged, I suspect, it's "who cares what editorial thinks?". Because as a recipe for success, riding roughshod over the

"It's not "get data on it" that's being flagged, I suspect, it's "who cares what editorial thinks?". Because as a recipe for success, riding roughshod over the opinions of 2000+ people who make all the content in your product is not a great one. What's more, BBC News editorial has reasons to be wary of pure data-driven decision making. It's not a pure Buzzfeed competitor -- it exists to serve the public purposes, not to drive eyeballs to ads. So arguments like "the data says nobody is reading our Welsh-language output, why don't we stop doing it?" or "The expensive pages covering Cornwall are getting terrible numbers, who care what editorial thinks, CLOSE 'EM DOWN" are not the slam-dunk you might be imagining. That's not to say there isn't useful data that can be put at the heart of the newsroom to try and capture things that matter to the BBC, as BBC News is well aware. They're building tools right now to support that [1]. But I think any editorial team will always want a way to say to its audience "we think this is important right now and you should know about it". So a constructive approach might be something more like "here is the most effective way you can highlight news to audiences" not "we don't care what you think, people don't like the breaking news ticker so it's gone, welcome to cultural change granddad". [1] https://www.newsrewired.com/2016/07/20/how-editorial-analytics-can-be-used-effectively-in-the-newsroom/ https://www.newsrewired.com/2016/07/20/how-editorial-analyti...

nailer 9y ago
> Because as a recipe for success, riding roughshod over the opinions of 2000+ people who make all the content in your product is not a great one. That's a gross mischaracterisation of data-driven decision making. The point is that non-data backed opinions belonging to anyone don't matter so much as the results of what you actually measure. > It's not a pure Buzzfeed competitor -- it exists to serve the public purposes, not to drive eyeballs to ads. Acknowledged, but effectiveness at that purpose is still measurable, and certainly should be measured as the BBC is publicly funded. > not "we don't care what you think, people don't like the breaking news ticker so it's gone, welcome to cultural change granddad" I'm unsure whether you're unfamiliar with media engineering, or simply creating a bizarre straw man of what was being suggested. Assuming good intention, and therefore the former, here's a more likely version: "We've measured the breaking news ticker and find out it that X% of users find it useful, Y% read it, Z% engage with it, engagement with other stories happens A% more/less when the breaking news ticker is displayed over the top of it. Given that our goals are B, this indicates we should.
- bonaldi 9y ago
  > That's a gross mischaracterisation of data-driven decision making. It is! And I'm v. pro D-DDM. But as I read your post -- given it has been flagged to death I don't think I'm alone -- you didn't appear to be calling for that. As I recall, you appeared to be saying "who cares what editorial think, work out what you want and get some damn data" (which is where the straw man comes from, fwiw).
  - nailer 9y ago
    > which is where the straw man comes from, fwiw No. I wrote: >> Act like engineers, work out what you want, test it and get some damn data on it. Which is quite different from: > riding roughshod over the opinions of 2000+ people who make all the content in your product I hope you don't need it explained, but just in case: we're ignoring all things not backed by data, which very different from ignoring a specific subset of people. The downvotes are for my saying 'damn' and writing angrily (ie, being pissed off with the poor engineering discipline of the public service). I'm intending to leave this thread, however, and enjoy the downvotes as "editorial doesn't like it" is frankly pathetic.

nailer 9y ago

> Because as a recipe for success, riding roughshod over the opinions of 2000+ people who make all the content in your product is not a great one. That's a gross mischaracterisation of data-driven decision making. The point is that non-data backed opinions belonging to anyone don't matter so much as the results of what you actually measure. > It's not a pure Buzzfeed competitor -- it exists to serve the public purposes, not to drive eyeballs to ads. Acknowledged, but effectiveness at that purpose is still measurable, and certainly should be measured as the BBC is publicly funded. > not "we don't care what you think, people don't like the breaking news ticker so it's gone, welcome to cultural change granddad" I'm unsure whether you're unfamiliar with media engineering, or simply creating a bizarre straw man of what was being suggested. Assuming good intention, and therefore the former, here's a more likely version: "We've measured the breaking news ticker and find out it that X% of users find it useful, Y% read it, Z% engage with it, engagement with other stories happens A% more/less when the breaking news ticker is displayed over the top of it. Given that our goals are B, this indicates we should.

bonaldi 9y ago
> That's a gross mischaracterisation of data-driven decision making. It is! And I'm v. pro D-DDM. But as I read your post -- given it has been flagged to death I don't think I'm alone -- you didn't appear to be calling for that. As I recall, you appeared to be saying "who cares what editorial think, work out what you want and get some damn data" (which is where the straw man comes from, fwiw).
- nailer 9y ago
  > which is where the straw man comes from, fwiw No. I wrote: >> Act like engineers, work out what you want, test it and get some damn data on it. Which is quite different from: > riding roughshod over the opinions of 2000+ people who make all the content in your product I hope you don't need it explained, but just in case: we're ignoring all things not backed by data, which very different from ignoring a specific subset of people. The downvotes are for my saying 'damn' and writing angrily (ie, being pissed off with the poor engineering discipline of the public service). I'm intending to leave this thread, however, and enjoy the downvotes as "editorial doesn't like it" is frankly pathetic.