In order to track the growth of your loyal customer base, the first thing you need to decide is how you define loyal.
If you don't currently have a definition, we would suggest you combine an element of frequency, and recency - so for example customers who have shopped at least three times, with at least one of these shops in the active period (as defined for your business).
So you would build a segment like this: (assuming an active period is 26 weeks)
Since this is set with a relative (rolling) last ordered date, this will dynamically update on a daily basis.
Ideally you will be tracking this on a weekly or monthly basis. However, if you want to go back in time to see how large your loyal customer base was at a specific date in the past, you would create a new segment, layering on order dates.
So your segment would look like this, to find out how many loyal customers you had on 31st December 2016:
(Logic is - has made at least 3 orders before 1st Jan 2017, AND has made at least one order in the six months prior to 1st Jan 2017 - i.e. 1st July-31st Dec 2016)
There are two ways to measure repeat rate.
1. Top level - what percentage of all your customers have ever repeat purchased? This is available to view in real time on the segment explorer screen as below:
2. Cohort level - what percentage of customers acquired in a specific month have repeated in their first 12 months?
For this you would create a cohort chart using the criteria below (with the 'cumulative' box ticked).
For tracking the movement of your repeat rate month to month, (this year vs last year), this is the most accurate way to compare.
Further reading on repeat rates: