Whether this version of NewsLynx was best constructed inside or outside the CMS was never a real question. In order to work with a variety of news organizations that used different CMS technologies, we always knew it would have to be its own platform. That leaves open the question of whether an organization should build its own internal measurement tool within its CMS—if indeed it has the engineering capacity.
This question is different for every organization, but we can discuss the pros and cons. The biggest advantage is that staff members do not need to sign-in to another system other than the one they are used to. This consolidation reduces friction and raises a user’s comfort level when adopting a new technology, which should be a high priority for an organization looking to start seriously measuring impact.
Another large advantage to living in the CMS is that you save having to follow one or many RSS feeds to ingest every article. You also get information like authors, publish date, and tags for free. Chalkbeat’s MORI tool also imposes the requirement that every article must have an impact goal assigned to it as a requirement for publication. That design decision might not work for every newsroom or every type of article—breaking news needs to go out as quickly as possible—but it does force staff to be more conscious of what to expect from each article and keeps monitoring and evaluation fresh in everyone’s minds.
Arguments against integration are also persuasive. Most of the metadata information can be acquired if the previously mentioned JSONLD standard is adopted and some effort is put into creating a standard RSS feed. Moreover, customizing a CMS is a task whose monetary cost, time cost, and required level of expertise cannot be overestimated. As a result, it’s no small undertaking to design and implement any change to the system. This reason alone, rightly or wrongly and also depending on the organization, is most likely enough to outweigh any efficiencies gained by integrating with the CMS and tip the scale toward building something outside of it.
As a compromise, we recommend designing a modular system, similar to NewsLynx, as the best balance. We built an API-driven platform that handles all data collection and standardization completely separate from the interface and user-facing code base. In this way, a newsroom could integrate a portion of a NewsLynx-like system into its CMS if it so wished while still keeping the impact tracking infrastructure in a separate code base. If a newsroom decides to upgrade or change the CMS, it would have to rebuild the visualization and interface—also a task not to be underestimated—but it wouldn’t lose the core impact tracking mechanism.