Bringing data into Gainsight is a critical component of our CS product. Without data, you wouldn’t be able to inform the CS motions that Gainsight CS drives. Because data is such a good thing, you might think the mindset is the more the better! However there should be a qualifier to that statement. The more insightful and actionable data the better.
There are a few circumstances where it’s not uncommon for clients to grapple with very large data volumes.
- Usage Data (Discussed below)
- Object with too many columns, too many rows, or both (Coming soon in a future post!)
Usage Data
Usage data can refer to any data that tracks how your clients are using your product. It is a type of time series data, meaning that it is created steadily as time goes on. The exact format of usage data will vary based on the company’s products. It may be as simple as a count of licenses assigned per customer per week, or as complex as tracking every click each end user makes on a website.
In its raw form, usage data doesn’t tell us much. As a CSM, how would you act on the data points below?
- Customer ABC assigned 100 licenses the week of July 1
- End User Jane Smith clicked on the “My Account” tab 6 times on July 1
Without context, this data is meaningless. You need to take the raw data and determine what story this data tells.
For example:
- Has there been a change in the customer’s usage?
- Is the customer using more or less than they are contracted for?
- Is the customer exhibiting usage that indicates a healthy customer?
- Is the customer neglecting features that healthy customers use?
When bringing in time series data such as usage data, we recommend using Gainsight CS’s Adoption Explorer tool.
Why Adoption Explorer?
This tool was built to ingest large volumes of time series data.
- Aggregation (Daily and Weekly) is easy to configure
- Archival of old data is easy to configure
- The ingest is fast because it only performs inserts
Aggregation Prior to Ingest
In lieu of using Adoption Explorer, you could aggregate the time series data prior to bringing it into Gainsight via a Connector or Rule. If you are working with large enough volumes, aggregation would likely be necessary at some point in order to meaningfully make use of the data.
What is your strategy for utilizing usage data in Gainsight CS?