A tech stack can become complicated in any organisation, yet is further complicated in multi-title organisations as different brands and teams want different things. A centralised team will have to make trade-offs on what gets developed when. I am seeing more and more companies move to a modular system that allows a lot more flexibility. 

This approach isn’t just for large companies with centralisation challenges; it can work for any organisation that uses a monolithic platform.

While the problem is not solely limited to multi-title organisations, it is certainly more acute in such organisations. If different brands or divisions are doing the same thing, it makes no sense for numerous people to tackle the same problem independently. The important point here is that pulling this into a central resource promotes efficiency and, when done well, allows local brands and titles to be a lot more nimble and should limit tech debt. 

One solution is to break down the tech stack into several main layers, using multiple APIs as interfaces. For instance, content storage, editorial workflow, and front end may be decoupled into separate systems. A couple of leaders have found that bringing some of the “boring” tasks in-house to a smaller group of people allows local product and engineering teams to be more creative.

NewsCorp Australia has adopted this strategy and shows the difference in approaches in the chart below:

This approach involves building a suite of tools and processes that can be mixed and matched, and possibly switched in and out with other solutions, including third parties. It’s much easier to A/B test solutions without committing to a whole new suite of tools, which is a huge upside for organisations as they can be more flexible with their integrations.

This modular system allows individual brands to have more control over the design, UX, features, and functions of their products. However, it has to operate within a framework to avoid becoming a free-for-all. This only really makes sense when different parts of the organisation are using the same tool. For example, editorial workflows would remain on one platform, not have every brand working with different tools and technologies. 

It’s essential to give teams autonomy to focus on the biggest customer problems and let them own these areas. Often this will relate to OKRs but must allow for some creativity, not just hard data. Culture is also vital in this system, which works on a sharing and listening basis.

And once the system is set up, it allows for new products to be spun up much more quickly, as laid out below in the slide from NewsCorp Australia:

A modular or atomised system makes a huge amount of sense and can transform the way a news organisation works, allowing more nimbleness.

But we need to be clear that this is no easy feat. It takes years, not months, to get this right. It can only be transformed when there is a full understanding of workflows around the business, selecting the right vendors and systems, and building pretty much from the group up. I have spoken to one leader who has seen a team broken from this process because they tried to move too fast. 

I am convinced this is the way forward. I don’t envy those undertaking the project — just those benefitting from the fruit of the labour!

If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .