Accepted Solutions
Indeed. Here’s how I approach this challenge.
- Charting is pretty basic in Airtable but if you have good data, you can make it look pretty useful to analysts and managers with very little effort and fully integrated.
- Mapping is also pretty basic and despite good data, it won’t put a dent in any location science usability.
- In mapping and charting, analytics (i.e., crunched data) is key - I call this viz-ready or report-ready data. It is typically summations, aggregations, etc where some degree of data science has been applied to elevate the data to tell deeper stories or explain causal outcomes.
- I use a number of platforms to blend data science into Airtable and there’s no shortage of posts from me about this topic. Some of them are actually driven by Script Blocks like this one which allows data science solutions and machine learning to exist wholly within Airtable.
A key requirement in data science is the ability to crunch data locally (i.e., in Script Blocks, roll-ups, etc) and then push those results into other systems. Airtable has an ideal pathway in javascript for building processes to make local aggregations and computations inherently prepared for higher-level analytical conclusions.
I recommend you try looking inward by finding someone on your team that can code in javascript; use the Script Block environment a little and see if this isn’t a good approach to building a unified solution at least as it pertains to the data crunching.
The map shown below is 100% computationally dependent on a script block, but rendered as a custom Mapbox solution. It performs extensive geometries concerning distance and nearness of B2B and B2C cohorts.
Almost every one of these apps are driven by Airtable data.
I tend to steer clear of Tableau for a variety of reasons too many to touch on here - the biggest is how tightly it is tied into Microsoft and its ability to embrace web services and modern web architectures represent a retrofit into a long and storied past of desktop analytics.
