Getting your Trinity Audio player ready...
|
In an increasingly tech-driven industry, the decision of whether to buy or build systems and platforms is complex. Christoph Schmitz, interim CTO at Norway’s Aller Media explains some of the complexities and challenges publishers face.
Where have we got to in the tech story and how well-placed are publishers to tackle the ‘to buy or build’ question?
So, as the media industry is working towards being more tech-oriented, we still have a background of not being a tech company, which means that we have a culture that has, in many cases and in many organisations, streamlined the operations. We have been cutting costs, we have made sure we are really efficient in what we are doing – and that’s in the DNA of many of the organisational cultures within the media industry. I see that very well in our organsiation. The commercial staff are really slick. The editorial staff are really efficient and using a lot of tools. And we have also seen this period of time where the systems have been switched out and changed.
A lot of people probably bought a print CMS or print solutions four or five years ago, for example, and that is probably an investment that they have finally stopped paying for now – and so they are looking to spend that money on something else. But all of these systems are holding them back. A lot of the new things we need in terms of data, data management, changes resulting from reregulation, payment methods, distribution models – which mean you have to drop something on Instagram, drop it into your CMS, add it to Facebook, drop it on Twitter etc – all of these things require a system that is flexible and streamlined. All of this means we need to become more tech-focused, because none of these solutions that we have been buying over the last four or five years are communicating together.
From an IP perspective, everyone would say we should have open interfaces, to make everything speak openly to one another. But the reality is quite different.
How are the suppliers responding and what are the challenges there?
The suppliers want to build these big systems but they can’t really build everything that we need. So how do we cope with that? I mean you either say ‘I am going to rely on one supplier and trust that they can handle all my needs for the next half a decade or so’, or you start building stuff yourself, which is pretty expensive because you need developers and there are cultural differences. We’ve all tried outsourcing, but outsourcing to a low-cost country is also difficult. We have some developers in the Ukraine and they are really, really good, and they have done some good stuff for us – but still there are some cultural differences and certain things they just don’t understand because they haven’t seen it. So covering that cost with local resources in Norway is also expensive.
So what’s your approach now and how are you tackling the issue?
So we say that we want to try to buy as much as possible and spend our money on integrations – because instead of saying ‘we are building a platform to do everything’, we say we are building an integration platform. We understand that we can’t build everything, and we understand that we need the different systems to speak together – so our strategy is to cherry pick the things that are best and have them run through our integration platform. So we’ve built a data platform ourselves, with the starting point being that the data platform is the core of the integration. We are now able to collect data from any service that’s available – and we have our recommendation engine, which is quite basic but performs really well, and we are able to let other services connect to our data platform and integrations service. Now, we have our core CMS for web publishing, made by a company that we own but which is loosely coupled, and we have other services connected to that. We have video, for example, and we are able to do recommendations through these integrations. So instead of saying we want to build everything ourselves, we are focusing on integrations.
How much is that influenced by not knowing what’s coming next?
Yes a lot. And I think that being a bit silly when you think about what you might beed in the future can get you a long way. I usually say. Ok, we need a CMS and a set-up that allows us to support telepathy. It’s about saying: ‘At some point, we may send and receive information through a channel that we haven’t had before.’ How do we organise our content and what sort of system can we set up that allows us to provide that? Two years ago, at the Facebook developer conference, they introduced a device that will let you receive information through wave scans. So it’s obviously going somewhere, and where exactly doesn’t matter, as long as you stack everything correctly. The same goes with payment – I was using bitcoin a few years ago and suddenly bitcoin is a valid currency. It’s not used by many, but it is still there as something we could never have predicted previously. Now, I’ve changed form using bitcoin. Will that be the currency I use later on? Possibly not – but there could be a system where different currencies work together. You have to be a bit silly to imagine the most crazy, out-there case that we might have to integrate with.
How difficult is the business-case side of build, given that it’s hard to go and ask for investment when you have just finished a project that might already be outdated?
Yes – it’s very hard, and if I ask for money to make something ready for integration, then someone is bound to come up with a supplier who has two or three things already in their box. And they will ask me: ‘Why don’t we just use this supplier?’ Then my challenge is again to go around the big conversation and talk about what we will not get and what kind of corner we will put ourselves in – what kind of risk we will be taking.
Over the last year to 18 months, those conversations have been fewer, and there are a lot of good suppliers coming in too, both in terms of recommendation engines, data platforms and all kinds of stuff that is high on the hype curve. We need to address all of those with an idea of how we can integrate them.