There’s a lot of buzz about analytics translators these days. In general, I find the name to be a really poor choice for what effectively is a product management role applied to internal data science or analytics services. But, I think the skills are really important. The question is, do you really need another role/headcount, or can you stay lean and train your existing staff on some of the skills that may help them out, including some fundamentals of UX/service design, product management, and visual design? While it may mean there is a reduction in technical and “hands work,” might the increase in “head work” lead to more effective, valuable, and insightful data projects, products, and services?
Bill Franks is the CAO for the International Institute of Analytics and recently wrote an article on LinkedIn entitled “The Fastest Growing Analytics and Data Science Roles Today.” Bill knows his stuff, so take a read! I chimed in with my comments and there were several other good comments in the thread. (Feel free to connect with me on LinkedIn as well).
Incidentally, there is a skill set that I think leaders who feel they need analytics translators might want to look for. I also think many UX practitioners could also benefit from this as they are often focused on the “U” piece only (User Experience) without considering the business overall, or the additional stakeholders who may not directly interact with a particular application or service that is being built. This practice is called service design, and it can be modified for analytics and data science projects.
How does it differ from product or UX design?
Service design looks beyond the customer experience and adds/considers:
- Employee experience (employees may also be the customer or analytics service’s main user)
- Business processes (by uncovering dependencies, duplications, communication problems, cross-department issues, etc., most of which are - you guessed it - human considerations as well)
A full analytics service might have multiple applications/interfaces within it, and so it’s important to realize that a service design mindset is a “level up” from any one analytics application, data model, or decision support tool. Each application interface in an overall service may require application-level design attention as well (e.g. product/UX, UI, or data visualization). However, this entirely separate from how the overall service is designed.
Example: consider a bank: you might have customers applying for loans via multiple touchpoints (web, mobile, in-branch), data insights/decision support tools being used by employees to review risk/approve applications, and additional back-office business processes the customer doesn’t see, etc. Whether it’s an analytics translator, data scientist, analytics leader, or designer leading the charge, by having a service blueprint and service design approach, you might uncover where dependencies are, how to improve the customer experience (loan rejection vs. approval flows), how to ensure downstream data gets put back into a predictive model by the right people/processes, etc.
It’s also possible for tech companies creating cloud/SAAS analytics tools to benefit from service design thinking, and product design or product management leads should probably be tasked with this. While a SAAS business may not need a service blueprint for their own company, in a B2B context, your SAAS analytics tool/info product may be “just one box” within your customers’ service blueprint (a visualization or diagram with lots of boxes depicting how a service works now or could work in the future). Understanding how “your box” (your service!) fit into your B2B customer’s bigger picture (their service blueprint!) may help you design a more effective SAAS.
I don’t have any evidence to back it up at the moment, but I might hypothesize that a lack of service design consideration may be in part why some data science and advanced analytics initiatives at non-tech companies do not “make it into production” and do not end up providing value back to the business. Thinking “outside the box” literally applies here: while we often hear this as being a form of “creative thinking,” in this case, it means understanding that your analytics service, SAAS product, or decision support tool may be just “one box” that sits inside a larger service that exists, even if nobody has explicitly visualized it in a blueprint yet. If you’re having problems with low user engagement or the business value of your data science/analytics efforts don’t seem to be what you hoped for, putting together current-state and north-star service blueprints might help you get better results.