In all honesty I must log onto the BBC website and my local news website at least 20 times a day mostly through my smartphone or work computer, and I can honestly admit that I learnt more about today's events by logging onto this 3 times today
This quote from an early user of the ITV News prototype illustrates what's exciting about the news stream; this is the kind of behaviour we wanted to tap into, but it's ill-served by most news sites because they are tied to the idea of the article - something that's published once and is rarely updated (like a newspaper or news broadcast). When we started working with ITV, we knew we had to do something different. We wanted to explore realtime news and use it to show how news stories develop over time, i.e. that news is an ongoing process not a finished product
An early diagram of the possible components of a filtered stream, and throughput across the day
Testing the concept of a news stream and what that might entail with users was central to the creative process. It was through understanding use cases and user needs that we were able to develop the key design principles that underpin the stream: 'tell me what the world's talking about today' and 'whenever I come back show me something different'. We needed to make change highly visible.
We kicked off the project in May last year. After an initial warm up period getting to know stakeholders, digesting research and shadowing the news teams, we moved quickly to sketching ideas and simple keynote concepts. Within 5 weeks of starting the project we built a prototype in Node.js.
Building a working prototype we could let people play with
At Made by Many, we believe in taking a lean approach to product development. This means testing ideas quickly with real customers by making things -- not asking people to tell us what they think.
Over the next 5 weeks we iterated on our prototype to test out new features, test design ideas and mock-up an admin tool for the journalists - because we were testing their responses, not just the audience's.
Prototype front and back end v. 1
We had some great feedback from people who were testing the site on a weekly basis. We worked with a group of users drawn from across the country and representative of the five audience segments ITV were most interested in reaching. Over the course of the prototyping we gradually grew the number of users who were testing the site - our aim was to get a mix of feedback from people looking at with fresh eyes and those who had built up some familiarity with the format.
Prototype v. 4, the second variant of the filtered stream, with story filter on the right
Of course not everyone liked it, we were trying something quite different and we didn't expect it to be a universal favourite.
The homepage is much too long..... I have to navigate what seems like miles to view all the headlines. There is a lot of homepage space wasted with Tweets and videos.
But there were some strong signals, particularly from repeat users, that it was proving an addictive and exciting way to follow the days news.
I think its a really good idea. I like that everytime I viewed the site, the news was totally current by minutes of my log in.
Some people said they had spent a couple of hours checking the site, even though we hadn't asked them to spend a particular amount of time on there.
I spent a couple of hours using it and checking on it every 15 mins or so whilst at work
Here are five things we learnt from the testing:
Combining all today's news stories together in one stream was the simplest, most compelling part of the proposition
Being able to filter this stream by the main stories was critical to the user experience - we tried this out for the first time in the third prototype
Video worked incredibly well within a stream - users were more likely to watch it in the context of the full story and it just felt natural to click to play as it popped into the stream
The stream works for local news too - it was the thing most users asked for early on and we even tried mixing local news into the national stream (an idea we want to develop further in future)
Having a mobile optimised version of the site was a must - it tested particularly well with those on the go and a number of people wanted an app
Working directly with users over a period of time helped us to develop our understanding of what was innovative and valuable about this news proposition and their feedback informed our design thinking.
Growing our understanding of the audience
Part of our brief was to develop a proposition which would draw in audiences outside of the ITV heartland. One early assumption was that the concept of the stream would appeal most to people who are very digitally switched on, but that it might alienate an ITV TV audience.
Interestingly, it proved particularly popular with two quite different audience types - those who were pro ITV but heavier TV than web consumers, and the news junkies who always want to find the latest, breaking news but don't necessarily check ITV.
In the case of the former, they liked the snacky feel of the format and the way that you can quickly grasp an overview of the main stories of the day.
Very likely. It would be the sort of thing I would look at even just to pass the time, but if it was there I would use it.
With the latter, it was the speed that new updates appeared on the site and the ability to easily jump into more depth or see how a story had unfolded.
Like it a lot, its better than waiting for the whole story to be published, I like being able to get as it happens information
Although we found that digitally savvy people did get the stream, we also found that they were less likely to become loyal consumers for ITV - they are quite flighty by nature, going wherever the story is on the web.
Transforming a newsroom from broadcast to web first
ITV has 300 journalists around the country many of whom were not used to producing content for the web. One of the challenges of the project was helping them to make the transition from a broadcast organisation, geared towards peaking at certain points in the day, to putting the web first.
How the news cycle has changed
Part of the prototyping was geared towards helping us (Made by Many and the editorial team) to understand what it takes to produce news in this format. Live blogging all the day's news stories is a big job and we had to work out the right balance of content.
We gave journalists feedback direct from users and this played a big part in helping them shape the content proposition.
Over the past 6 months, the newsrooms have been using the prototype to learn how to produce and curate the best kind of content for the web. We even took the unusual step of updating and maintaining the prototype during development of the production site so that journalists could keep playing with it until an Alpha of the production version was ready to go.
Make tools that work for journalists not against them
We were also working closely with the newsrooms so that we could better understand their workflow and how to make tools that fit easily into their existing habits.
As part of the prototyping process, we created an email-in tool so journalists could send in pictures and text to the team in the newsroom. These would appear with a notification in the admin section of the site. It was a very simple prototype of what an app might do in future to get journalists accustomed to sending in content and make it very easy to use it immediately on the web.
Email-in tool notifications in the admin
Email-in tool expanded
An early learning was that the web teams need to be small and agile, acting as curators of the content produced by the wider group of ITV journalists and rest of the web. This frees up the correspondents and editors who are out following the story - they just need to remember to tweet
and email in content as part of their news-gathering process.
We've released a minimum viable product so that we can learn from real use in the wild
Testing and prototyping behind the scenes can only take you so far. We used it help us understand the core appeal of the proposition and what features we needed to develop first to deliver an innovative, coherent product.
In that process, there were lots of things we tested which we discarded or put into the icebox for future releases. For example, we were very excited about how we could help users follow a big long running story which develops over months, like the Libyan revolution. We realised we could test out the beginnings of this idea through tags and that to do anything more would be making too many assumptions for a first release of the product. We need to find out how well the main proposition works first.
Another early idea was to use a sliding bar timeline to navigate the stream, which tested very well with users but it, too, ended up being set aside as it didn't provide enough value on top of the stream to merit making it into the first release; it would have got in the way of building more important features.
Likewise, when it came to mobile we had to deprioritise building a mobile app for reporters in favour of building a mobile optimised version of the site. We felt it was more important to deliver additional value to users than editors in the first release of the product.
But of course the real test will be in seeing how people use the site now that we're live to the public.
Already, there have been a number of opinions voiced on Twitter - as you'd expect, we've got our fair share of fanboys and haters.
What's going to be more interesting in the weeks to come will be monitoring actual use. Will the findings we found in testing be backed up in the metrics?
Our testing told us that people found it addictive, that they would come back repeatedly during the day and that it made them more likely to watch the ITV news bulletins on TV.
We'll be watching closely to see how long people spend on the site, how much content is shared and whether audiences grow, both online, on mobile and on TV.