Can digital media be faster than Twitter? How to cut time to publish by 70%

Written by Joanna Kocik, ex-media professional and content strategist

When it comes to the issue of "speed vs accuracy" in the journalistic world, one can debate it for hours. But the honest answer to this dilemma has nothing to do with compromise. Your media should be as accurate and as fast as possible. And it just so happens that there’s a lot you can do with both matters, improving efficiency and cutting time to publication by up to 70%. Here's how.

I remember that day very well. It was early morning in Central Europe on 17 April 2018. I was in the newsroom and had the shift of the homepage editor. It was me who decided what to put on the front page and when.

There wasn't much going on. The usual stuff. And suddenly my colleague said, "Barbara Bush is dead".

It was already known that her health had deteriorated, and she had decided to discontinue further treatment. A day earlier, CNN had published that her husband, former President George H. W. Bush, was with her. So when I saw a tweet that said, "Barbara Bush dies peacefully in her Houston home," I didn't wait long. We immediately put the breaking news on the front page.

And then my colleague, who had broken the news to me, anxiously spoke up. "Wait," she said, "there's something strange about this tweet".

Yes, you guessed right. The tweet came from a fake account that resembled CNN's profile. And it wasn't true. We were posting fake news.

Barbara Bush died later that day. This time we waited for the official statement before publishing the information. But the milk was spilt. For two, maybe three minutes, we had a fake on our home page. And a lot of people saw it. Like, really, a lot.

Later, I wondered why we were in such a hurry. The reason was obvious: we simply didn't want to be late. We tried to avoid another question from our superiors: "Why did it take so long to publish this story?". We – humans – didn't recognise a fake Twitter account and fell into our trap: speed versus accuracy.

This may sound cruel, but VIP deaths are the easiest to handle in this context. Most newsrooms have a secure database of obituaries that is immediately available when a VIP passes away. Then a media worker simply copies the article, adds details and has an article ready within minutes.

But it doesn't work that way for other types of breaking news. A journalist has one minute to write an article, add a title, a picture, tags and linked topics. Then he or she clicks "Save" – and that starts the ride.

I probably can't even count how many times the system has been our biggest enemy.

"Your article cannot be saved; please add a publication date".

Are you serious? You want me to add a publish date manually? That's an extra 10 seconds.

"The image isn't framed properly".

Who cares? I need this article immediately, I will frame the picture later.

"The image isn't framed properly".

OMG, all right, let's frame it (that's another 30 seconds).

"You have exceeded the recommended title length".

Couldn't you just tell me that when I wrote the title? I've to do it now. 20 seconds.

While I’m struggling with the tool, the editor keeps asking. “Where’s the damn article?”. I want to let them know it’s ready. But to do this, I must paste a link to our communicator.

Ten more seconds.

You'll have to forgive me for this longish introduction. The reason is that I've been frustrated for a long time with the speed pressure in the newsroom, with faulty editorial software, and the lack of fact-checking tools. How was I supposed to do my job well if every time something was urgent, technical obstacles popped up? I know it sounds like Murphy's Law, but it's true in every newsroom: "Anything that can go wrong will go wrong.”

It's hard to justify publishing fake news – but hands up, who's never been baited. And hands up, anyone who has never said, "We were behind the competition because the system didn't work right."

As a high-level media professional, you don't want to choose between speed and accuracy. You want your media to publish verified, fact-checked stories ASAP. You don't want your staff to have to deal with annoying features in the CMS that slow down the publishing process. You don't want stories like the one with Barbara Bush. You want to minimise human error and system failure.

The good news is that this is possible – and now I will tell you how.

Read more: 13 numbers behind the improved publishing workflow

Journalists and media workers struggle with the fact that editorial tools are simply not comfortable to use. They are unintuitive - something is in the upper left corner instead of the usual bottom right. The text window is often too small, and most users prefer to write their content in Google Docs or Word. This is also because they fear the editor will not save the text and all the work will go to waste. Everyone has experienced this – believe me.

So, to write and publish a text, you often need several tools: one for writing, one for copyediting, and another for publishing the piece on the main page. And probably another one to monitor the results.

Moreover, content editor tools have too many features to fill in manually. Why not introduce an editor with predefined presets where everything which can be automated is automated? Less time devoted to deciding what to write and completing tabs means more focus and more time for actual writing.

Let’s go further down this road: you know that speed matters. Why, then you still maintain a system that doesn’t support that goal? In many newsrooms, excellent outcomes are delivered primarily because of the heroic engagement of people – not because of the tools they have. Automating repetitive tasks will give people more time to write and edit. They will produce more content, and you can catch up with the competition. Add innovative AI tools, and you can compete with all the big players on the market.

Imagine: you could introduce a “fast track” feature in the editorial software (like we did in our project for Wirtualna Polska), which would allow for a bare minimum to get breaking news published. You could improve the process of adding photos and other content elements – and avoid a work process where people spend more time filling metadata points than actually writing content.

Yes, we’ve seen many such cases. Yes, they probably exist in your newsroom, too.

Okay, we agree that better editorial software speeds up content creation. Another improvement it brings is speeding up communication. People in newsrooms typically use a communication tool – Slack, Microsoft Teams or others – and spend a lot of time sharing information, especially in a hybrid or remote work environment.

Journalists who have just written an article post and sent it for review have no idea what will happen to their content unless the editor messages them. They don't know when a piece will be accepted for publication or published. If there is a typo or a minor mistake, they usually get a written request to correct it, then do it and write back that they did it... Then, the editor asks them to create a different title version. Again, they write it; they send a message. They wait for the answer. The editor manually copies the new title of the article.

Precious seconds go by. The clock is ticking.

It doesn’t have to look this way. Your CMS could support additional features like publish status (visible to all stakeholders: journalists, desk editors, and proofreaders) to reduce the time to publish and minimise the time people spend exchanging messages.

Some newsrooms are already benefiting from such solutions. To give you some examples, let us show you the improvements we introduced in the editorial workflow of several media organisations.

We created systems that:

• eliminate the need to enter unnecessary information manually
• recommend article settings by analysing the author's history and statistics
• suggest options based on current trends in the newsroom
• add tools that automatically suggest metadata based on live analysis of the content
• allow photos to be added faster and more intuitively
• have a "fast track" feature that allows for quick publishing of breaking news
• insert additional content elements such as videos, quotes, and embeds faster include AI features that recommend images, videos, entire new paragraphs, links,

The results?

• Time to publish of urgent article reduced by over 70%.

• The overall efficiency in the entire organisation increased by 10% (journalists can write 10% more articles daily).

• Icing on the cake: over 80% of users are satisfied with the new editor features.

Do you want similar results in your newsroom? Let's talk --->

Okay, but what about fake news? Can a system function prevent situations like the one with Barbara Bush?

To answer this question, we need to move into the realm of AI, social media, and fact-checking. Let's break down this topic in a little more detail.

First, on AI. With the proliferation of language models, the race between fact-checkers and those who check them is unequal. We already know that artificial intelligence can produce natural-sounding text at the push of a button, essentially automating the production of misinformation. This asymmetry urges media organisations to develop their own AI-powered tools that help detect fake news.

Independent researchers and companies have been developing software tools to help journalists and fact-checkers identify and combat fake news for years. According to experts, now this help can come at three levels:

• detecting deep fakes and manipulated content
• checking whether the content itself makes false claims
• responding to fake news

In the past, the rule in most media was that a person's death must be confirmed by two independent sources. Today, in the race against time, that rule is often compromised or sometimes impossible to fulfil. I can recall the confusion we experienced at least a couple of times with the deaths of figures like Osama bin Laden or Kim Jong-il. You can't just call someone in the North Korean government and ask, "Is this true?" You either have to wait for official confirmation or take the risk and release the news after the first tweets have appeared.

But there may be another way. In the case of Barbara Bush, we'd need an algorithmic fact-checking tool integrated into the content management system that verifies the accuracy of the message using comparative data from posts and articles on social media. It could detect that the information hasn't yet been shared on other online platforms or by official channels and assess the source's credibility.

It's common for misinformation on different topics to come from the same root – we saw this with the large-scale misinformation on Covid-19, followed by the U.S. elections in 2020 and the war in Ukraine. In this case, back-office AI tools could identify which fake news stories have the same origin and allow journalists to watch out for them.

AI tools could also help assess the credibility of user-generated content (UGC). When something happens – a shooting, a fire, an accident – it's usually someone with a phone who is the first reporter on the scene. Judging whether the photos or videos someone sends are true is challenging. With the help of AI, it would be easier to check the metadata of the content and determine whether it hasn't been manipulated.

I can recall a situation where this feature would have been handy. It was the death of Osama bin Laden. Shortly after the news broke, an alleged picture of bin Laden began circulating on social media. It was even republished by some reputable photo agencies. It wasn't until about an hour later that it turned out to be a photo montage when someone who knew about photo editing discovered the image. But Osama died in 2011 – at that time, we couldn't even imagine what would be possible in the future.

So, having the context of fact-checking, let's return to the question of speed. Can a digital medium be faster than Twitter (or any other social medium) when witnesses can live stream current events? Can your medium be faster than Twitter?

The first answer is no, of course not. But as a reputable medium, you’re not competing with Twitter regarding speed. And – what’s more important – you have a huge advantage over it.

The first moments when a story leaps into Twitter are often very chaotic: it's usually a stream of shared consciousness with no contextualizing information, full of typos and misspellings. It's like a single firehose containing every tweet every user sends worldwide.

Journalistic news content provides much more context and explanation of actual events. That's because journalists in newsrooms have something Twitter users never will: years of experience linking distant facts and a database of contacts with people who can explain any topic in detail, making complex information digestive for an average user.

In this analysis for Forbes, Kalev Leetaru points out that Twitter's speed on breaking news comes mainly from retweeting, not from users sharing their new perspectives and thoughts on events. And when it comes to scripted events like political conventions, Twitter doesn't outperform the news. It acts more as an on/off switch that registers an event, but the media offer a rich, advanced, and fact-checked narrative about it.

It's hard to say if I could have avoided a Barbara Bush mishap. An AI fact-checking feature built into CMS probably would have helped, but here we were dealing with a human error.

I know that well-designed, intuitive CMS and editorial tools that form a system are irreplaceable regarding speed and accuracy. Journalists who don't have to spend time on redundant tasks and manual work can avoid mistakes caused by a lack of focus. In this case, a combined system where man and machine work together like a cyborg can bring the results you strive for. It makes your medium fast and accurate and increases the content you produce.

Let's talk – whether you're considering speeding up the publishing process, redesigning your back-office system or adding functional AI features to make journalists' jobs easier. We have plenty of other stories and numbers to prove that you can aim higher in your digital transformation.

Read also: Why media need to learn how to swim in the AI waters