Setting up Dynamic Benchmarks
We are a price comparison website and our building a cockpit where our partners/providers will be able to monitor their own KPIs in real time (e.g. new contracts won) as well as market trends (e.g. prices, churn). This information will give them more transparency about their performance, the market and help them control this sales channel.
Providers should only be able to see their own data as well as a Benchmark. The Benchmark should be dynamic e.g. if company applies product or geographical filters the benchmark should also adapt. The challenge is that if we restrict data to one provider (Row-level security), we automatically remove all other providers that make up the benchmark aggregate. Does anybody have any experience with this? Or any ideas how to solve this?
Thanks in advance for your help,
I have not attempted this in ThoughtSpot, however I have solved this issue with other tools. The approach I took was to load the data twice.
The first dataset you load to calculate the actuals. This dataset has the RLS applied to it.
The second dataset is a copy of the first however you remove identifiable attributes. I.e. specific company. This data is used for benchmarking measures. No RLS is applied.
This may also require you load a second set of dimensional rows for the de-identified attributes. I.e. Assume product dimension has 10 columns, 7 of which you want to 'hide'. The hidden columns could have a value of hidden. These dimensional rows are joined to the benchmarking dataset, rather than the 'real' dimensional values.
You could model this is a chasm trap. Again not something I have done in ThoughtSpot, however I believe this should work.