In March I travelled to Erbil, Iraq, a mere 30 miles from the frontlines of Iraq’s battle against the Islamic State. Unlike previous projects, I wasn’t there to bootstrap a news service or train local journalists. Instead I was taking advantage of the proximity to Iran and large Farsi-speaking community to localize the StoryMaker curriculum for Farsi speakers.

In 2013, the StoryMaker Coalition launched StoryMaker version 1, with a 62 lesson course in multimedia journalism. In December, ahead of the launch of version 2, I lead an effort to localize the content for Kirundi-speaking journalists in Burundi as part of the lead-up to the presidential election. This month, as part of a project to bring StoryMaker to Iranian users, we completed a Farsi localization. The process, which involved re-creating more than 150 multimedia assets, resulted in a number of conclusions valuable to anyone localizing their curriculum or using someone else’s.

Trainers with a good deal of experience working in cultures different from their own have all experienced a moment when the cultural content of an example diverts the focus of participants away from the lesson. What we’ve found is that content issue primarily arise when the goal is teaching elements of storytelling, rather than technical skills.

#1 Only localize assets where the content of the language is important.

Localizing assets for which the most important element is technical is a waste of time. For example, teaching someone the impact of machine noise on an audio recording doesn’t require the individual to understand the words being said. Nor is it necessary to provide a contextualized photo or video clip to teach someone how good or bad white balance affects the content.

If we spent less time planning how to recreate content where the lesson or training was purely technical in nature, more time could have been spent finding a broader variety of stories. A broader variety of stories and characters would ensure we had an even more effective localization. As it stands, our content is primarily focused on Farsi-speaking Kurdish Iranians, which may leave some users feeling their communities were intentionally left out.

#2 Remember that no “localization” will ever be local enough for everyone.

In order to create content for Farsi speakers, ideally we would travel to Iran to produce the content. Given the legal and political situation in Iran, this was not possible. However, even had we traveled to Iran, to make a truly “Iranian” localization, we’d need to produce a variety of stories across Iran, including Iranians who are Persian, Kurdish, Balochi, and Azeri, to name a few. We’d have to ensure we produced stories in a variety of urban and rural communities.

In the end, we chose to produce our content in Iraqi Kurdistan, which is relatively open, stable, and has seen a huge influx in Farsi speakers in recent years. This choice was also dependent on having a very good recommendation of a local, English and Farsi-speaking fixer. In all, this meant that we could focus on producing a variety of strong examples and interesting news stories.

#3 Focus on making the important material that doesn’t already exist.

Our greatest success in the Farsi localization revolves around the fact that we were able to create a wide variety of material, and in particular, more stories than required or expected in the initial curriculum. Attempting to create an exact replica, in its entirety, of an existing curriculum in a new local context is, as we’ve seen, unnecessary, and never going to please everyone. It’s also boring and reductionist. The stories and the approaches that are interesting to Egyptians or Libyans will not be the same for Iranians, or Burundians, or Cubans.

The important pieces, which should be the same across cultures, are the journalism and storytelling fundamentals. You need good characters. You need accurate, honest journalism. Trainees must understand they should focus on “do no harm” principles. There are some constants, most news stories are framed around specific events or people, otherwise they are more broad, and known as features or in StoryMaker’s lingo, issue stories. Video and photo stories should ideally contain a variety of images and shot types, audio stories should contain not only interviews and/or the reporter’s voice but also background audio. All stories should be built on a variety of source material and focus on informing the audience.

The results of the StoryMaker Farsi localization include more than a dozen video, photo, and audio stories not widely seen in the international media. In this way we succeeded not only to produce a great localization, but great journalism.