This post was originally written for Your Data Stories project blog. YourDataStories is a European research project which envisions to combine and fuse the two “suppliers” of open data (traditional producers and user-generated content) and to exploit the added value from this amalgamation in order to better satisfy the needs of the “demand side” – meaning citizens, journalists and others.
YourDataStories is all about enabling everyone – citizens, journalists, and other professionals – to create their own data stories. One might ask him or herself, why not setting up an own Data blog? Such can be for private amusement, serve as a portfolio for students and professionals, or simply function as a data blog directly associated with a company. This post sets up a guideline that leads you through the process of launching your own data blog.
Planning the launch
Start planning for the launch of the data blog around three months before the launch. Don’t forget that Data Journalism is nothing new. Journalists have been using numbers and data for a long while even before Data Journalism was trendy, and this would mean you will likely find some data stories in your archives, if you look carefully. So my suggestion would be to look into your archives, particularly in the past year or two to see if you can find stories with data elements or data visualisations in them. This could give you a head start. Whether you want to include these in your data blog or not, you can decide later.
You might have found a long list of data driven stories in your archives. You don’t want all of them. Go through all these stories to make a shortlist. This will help you to (1) familiarise yourself with all these data stories and (2) have a list of stories that you may want to include on your data blog, in addition to the number of stories that you decide to publish for your launch, and soon after that.
Ok, so we are at ‘stories for the launch and soon after that’. What you need at this stage is a pool of data story ideas. I suggest creating a Google doc for this and sharing it with the data team, and other people in the newsroom or in the organisation, who have interests in data journalism and data storytelling. Ask them to add their ideas and potential data source(s) in the document. I bet you will soon have a good pool of data story ideas, which you will be able to get back to for a while. You may decide to use this document as you continue for the data blog so that people can add their ideas and the data sources they are interested in for future data stories. This way you always have a back pool of ideas to go to if you need new story ideas.
Data Source Bank
Another useful thing to do is to compile a list of useful data sources. These could include a variety of public and private, national, European and international data sources. The most prominent and obvious examples are Government portals, Central Statistics offices or sources such as Eurostat, World DataBank, UN Data and Gapminder. Here you can find a list of some useful data sources on the YDS website.
While putting a list of data sources together, you might realise that many of these sources publish data on a regular basis, and some even have a predefined schedule for publication of data. These could include the rough release dates for data sources, as well as exact dates and links to the data when they were actually published. For example in the initial schedule it might say Housing prices index by the end of June. Later they might give a more accurate indication, for example 3rd week of June, and then you will have the release on an actual date. This leads us to the idea of “Data Diary”. You would want to keep the planned release date, as well as the recently released data sources in your data diary. This will give you a chance to plan for near future, as well as having a good idea of the datasets which have been published in the past few days or weeks. It would be a good practice to try to keep the data diary up to date on a daily or weekly basis depending on your resources.
Get more concrete
At this stage, we have a list of story ideas, a data sources, a data diary to be updated and maintained and a large pool of data story ideas. You are now ready to make plans for the actual launch. A first thing to do might be to pick a launch date, or week. Your newsroom and analytics team will have a good idea of the times when it is most suitable for launching a new section. For example certain times of the year, like christmas and new year’s time may or may not be a good time for launching a new section, depending on your audience.
So far we have been talking about data and story content and the launch date. What you need now is to start make other decisions such as the name, url, platform for publication (perhaps a blog or part of your daily publication/CMS system), logo, design, branding, social media, etc. More importantly it is now time to start working on a number of projects from your idea list. Put these all in a pipeline and assign appropriate people to the tasks, and start working away.In the meantime you may want to ask your video, audio or still team to start making promo pieces for you.
Discuss this with your production team. They likely have procedures in place for launching new sections. In any case think of test runs, and try to break the new section yourselves, or ask other journalists in the organisations, who have not been involved in the data blog or stories being published to have a look at it. Then go ahead for your actual launch as planned. Before the launch make sure you have the following in place:
- Have a launch/welcome story in addition to your data stories. Be open about your aims and objectives and introduce the team.
- Choose one of your stories as headline story. Consult your editor for this. A popular story, which everyone can relate to, may work better for your first headline than a sophisticated investigative work.
- Your actual, final url to be used
- Know where exactly the blog will sit in the web site
- Have your social media account ready (you may not want to put much there before the launch)
- Have your other sections prepared to tweet about you, for example the organisation handle or the news section.
- You may want to get an article about the data blog/section ready to be published in another section of the website/newspaper, for example news. You may want to publish this a while before your launch to create an anticipation in your audience.
- Have a video or a podcast ready about the new work and the team. You may want to include it in your launch story, as well as distribute it other handles of your organisation.
- After the launch and promoting the data blog, you may want to promote every single new story that you have published for your lunch separately, just in case some readers were not interested about the data, but an actual story.
It’s getting serious – but don’t forget to enjoy the excitement. Via GIPHY
One important thing to remember: Be ready for unexpected and be ready and flexible to respond to your audience. For example, imagine you have chosen a headline story which you are proud of and think will gain the maximum audience attention. After the launch you realize another story is trending from the data blog. Be prepared to change your headline and respond to audience reaction. You are likely to get more audience and follower that you would have expected. Celebrate it!
Some of you might think having a data blog/section you would need an army of journalists and programmers. Well it is good if you do, but in reality many of news organisations would not have enough resources to start a large data journalism team. Good news is there are loads of super cool, easy to use, and in a lot of cases free tools for you to use. It is great to have programmers handy, but in many cases the stories have such a short turn around time that you will simply not have time for programming. There are a bunch of great tools out there. Examples are: DataWrapper, Google Fusion Tables, CartoDB, ScraperWiki, import.io, Timeline.js, Tabula, Infogram, RAW, Tableau Public, Silk.co, Adobe Photoshop, Adobe Illustrator and Adobe Indesign, plotly, SPSS, OPenRefine, Gephi and of course Excel. Depending on your needs and interests pick a few of them and familiarise yourself with. You may want to team up with research labs and universities for specific projects instead of having expensive in house expertise. They are always happy to work with journalists, as it allows them to put their expertise and state of the art work in practice, and also will give them visibility. In addition to looking at the above tools, keep your eyes open for new tools. One good way to do it is to follow #ddj and#dataviz hashtags on Twitter.
Any potential issues to be aware of? Absolutely.
Mobile is a big concern when it comes to data visualisation. Depending on the type of your organisation, it is likely that you will have well over half of your audience come from mobile devices and naturally the production team may want to include any non-mobile friendly items on the website and app (if you have one). There are various issues with tools on mobile platforms, which you might come across as time goes by. From simple issues such as legends not showing properly on DataWrapper, to more complicated issues. Don’t be disheartened when facing such issues. Investigate the matters and try to find a solution, if you were successful then great, if not add the tool to the list of the tools that should not be used. You can contact the tools providers, they might have an answer to your problem, or may be able to put this in their production line to address for their future releases.
Another problem is reuse of charts for print if you are a newspaper. Print journalists and producers often complain that the Data blog visualistions and interactives have to be redone completely in the cases that the story also appears in paper. We surely cannot have interactives on paper, but it would be great if the tools could provide us with print friendly alternatives.
Posted in: Data Journalism