Repeat customer rate in e-commerce?


November 27th, 2013

I am doing some due diligence on a business and am interested in what folks have seen as a good repeat customer rate. 

To provide some background, this business is in the food space, they only sell their one item online. They have been in business about 10 months so their customer list is very small at the moment (they have done no marketing). The rate they are seeing is about 22%. 

Does this sound good/bad/average to folks?

Thank you so much in advance for your help!

Daniel Ice

November 27th, 2013

The book Lean Analytics has a really good metric for this.  It looks at the annual repurchase rate: what percentage of people who bought something from you laster year will do so this year?  The book then break sites down into three modes, based on the percentage of repeat customers.
  • 0 - 40% - acquisition mode - 70% of e-com sites, think Blinds.com, you are only going to order blinds once every 5 years (at most)
  • 40 - 60% - hybrid mode - think Zappos
  • 60% - 100% - loyalty mode - 10% of e-com sites, think Amazon
Given this view, you have an acquisition based site.  So you need to really understand what it cost to acquire people are or think about how you can grow it into a loyalty based site.  One idea on an acquisition site are do really promo sharing your purchase with friends on social media.

Food products seem like a natural repeat business.  If you are interested in developing the reorder rate you could look at subscriptions or branding your product to relate to holiday dishes.

Also, the book says:

If 60% or more of last year's buyers will buy something this year, the company needs to focus on loyalty. Loyalty programs work well only if the retailer has this kind (over 60%) of engagement, and only 10% of e-commerce businesses end up in this mode when mature.  Amazon is a good example of a company in this mode.

Hope that helps.

Feel free to send me a message and we can do a quick phone call if you want to talk through this more.  I love e-commerce.

Anand Iyer CEO & Founder at Trusted

November 27th, 2013

Hi there,

So, 22% is the number of users who buy again within 30 days? If so, I don't think that's necessarily bad, but it depends on how your math over a course of time (3/6/9/12 months) pans out, IMHO. Ideally, you'd want to look at repeat purchase % over the course of 12 months, and the associated LTV over 12 months (this too depends on what they are selling and what is expected behavior for users who buy this product - can you expect a user to buy this product multiple times a month?). And you're correct in taking in to account the fact that the 22% repeat rate might not be based on statistically significant data, although it could be a good benchmark especially if you have done no marketing and the users have been acquired organically (typically the best performing users)

Granted, the business has only been around for 12 months, but you could extrapolate for recent cohorts based on trends from older cohorts. 

Hope that helps.

PS: RJ Metrics is your friend :)


November 27th, 2013

First of all, thank you so much for all your help!

Anand, great point, this number is based on the lifetime of the business so about 11 months, so no cohort analysis yet. 

Shobhit, you are way smarter than me! The 22% is over the course of about a year and we definitely don't have enough dat to know more at this point but your point is outstanding and made me think about things that are somewhat over my head so thank you! 

Daniel, this book sounds great, it's going on my read list! Yes, totally agree with you, currently there is no social media, no word of mouth etc. Also, most of the sales are occurring on Amazon so we would obviously want to move away from that or at least sell 1 item on Amazon, get folks hooked so that they come to the site to get a bigger variety of items (right now there is only one item so not really an issue). Would love to chat more, will send you a private note. 

Cynthia, they do not. No email or anything like that, obviously low hanging fruit :)

Shobhit Verma Ed Tech Test Prep

November 27th, 2013

Time information is missing.
Is it 22% every month ? [as in every month, 22% of the customers had ordered sometime before also]
Or is it 22% lifetime ? [ as in 22% of the customers have ordered more than once]
What about people who have ordered at least 3 times ? Are they some where in the range of .22*.22 =  4.8% ?
What about the sample size ? That can help figure out the confidence intervals around these numbers; This is important only if you are making a strategic decision which depends on a threshold like "I will invest only if the number is > 15%", in that case you want to know whether 22% is by chance or its significant. In that case some sort of poisson model for repeat probability should work. If there is too much data then you can even do fancy things like seasonality analysis and correlation with local weather etc.
Fun stuff! Good luck.


November 27th, 2013

I'm in a different space but a 22% return purchase rate sounds quite good to me. Do they actively market to past purchasers?