Amazon customer relationship management metrics are the numbers that show whether buyers come back to your brand after the first order: repeat purchase rate, Brand Follow growth, Tailored Promotions redemption, message response time, Order Defect Rate and the share of revenue from returning customers. Track them in Seller Central’s Brand Analytics and Customer Engagement tools, and review them on a fixed weekly and monthly rhythm.
Most sellers already watch sales, sessions and ACoS. Those tell you how well you acquire. CRM metrics tell you how well you keep, and on Amazon keeping is harder because the marketplace owns the customer record. You never get an email address, so every retention number has to come from tools Amazon gives you.
What counts as a CRM metric when Amazon owns the customer?
On your own website, CRM means a database of contacts. On Amazon, it means the signals you can see about repeat behavior and the levers you can pull inside the platform’s rules. The useful set is small, and that is a good thing: a short list gets reviewed, a long one gets ignored.
Think of the metrics in three layers. The first is loyalty outcome (do people buy again). The second is engagement reach (how many people can you contact through Amazon’s channels). The third is service health (are you giving people a reason to leave).
- Loyalty outcome: repeat purchase rate, repeat customer revenue share, Subscribe and Save active subscriptions.
- Engagement reach: Brand Follow count, Customer Engagement email audience size, Tailored Promotions eligible audience.
- Service health: Order Defect Rate, buyer-seller message response time under 24 hours, negative feedback rate, return reasons.
Where do you find each number in Seller Central?
Brand-registered sellers get the most data. Go to Brands > Brand Analytics and open the Repeat Purchase Behavior report. It shows orders, unique customers, repeat customers and repeat purchase sales by ASIN and by week, month or quarter. Export it monthly; the dashboard view does not keep long history in a way that is easy to compare.
Brands > Customer Engagement shows your follower count and the results of each email campaign you send to followers or recent customers: emails sent, open rate, clicks and attributed sales. Brands > Brand Tailored Promotions shows redemptions and sales for promotions aimed at audiences such as repeat customers, cart abandoners or high-spend customers.
Service health sits in Performance > Account Health and Performance > Customer Satisfaction. The commonly cited Order Defect Rate target is below 1%, and Amazon expects replies to buyer messages within 24 hours, including weekends. Both are published thresholds, not suggestions.
Which metrics should you review weekly, and which monthly?
Weekly review is for things you can fix inside a week. Monthly review is for trends that need a few hundred orders before they mean anything. Mixing the two leads to overreacting: a repeat rate that drops for one week usually reflects a promotion calendar, not a loyalty problem.
- Weekly: message response time, new negative feedback, A-to-z claims, return reasons on top ASINs, Tailored Promotion redemptions running that week.
- Monthly: repeat purchase rate by ASIN, repeat revenue share, follower growth, Customer Engagement campaign results, Subscribe and Save subscriber count.
- Quarterly: repeat rate by product line compared with the same quarter last year, and how the gap between first-order and repeat customers’ average order value is moving.
A worked example
Take, for example, a supplement brand with three ASINs. The Repeat Purchase Behavior report shows 2,000 unique customers last quarter and 340 repeat customers. That is a 17% repeat customer rate for the period. On its own the number says little. Compared by ASIN, it might show that the magnesium product drives most repeats while the sleep blend almost never gets reordered. That is a product question, not a marketing one, and it should go to whoever owns formulation and pricing.
How do you tie CRM activity to revenue you can defend?
The fair test is whether a CRM action produced orders that would not have happened anyway. Amazon reports attributed sales for Customer Engagement emails and Tailored Promotions, but attribution windows are generous and returning customers often buy regardless. Treat platform-attributed sales as an upper bound.
A simple way to tighten the number is a holdout. Run a Tailored Promotion to repeat customers on one ASIN and not on a comparable one. Compare the change in repeat purchase rate between the two over the next month. It is not a lab experiment, but it is far better than reading the attributed sales figure alone.
- Record the baseline repeat rate for both ASINs for the previous 60 days.
- Run the promotion on one ASIN only, with a discount you can afford across all redemptions, not just new ones.
- After 30 days, compare repeat rate change and net margin, not just revenue.
- Keep a short log of every test so the next person does not repeat it.
Which metrics look useful but mislead?
Follower count is the most common trap. It grows when you add the Follow button to your Brand Store and Posts, and it looks like an audience. But only followers who also receive and open emails count as reach, so watch campaign open and click rates alongside the total.
Star rating is another. It matters for conversion, but it is a lagging and blended measure; a product can hold 4.5 stars while a specific defect drives returns. Return reasons and review text tell you more about why people leave than the average does.
Average order value alone also misleads. A bundle promotion can raise it while lowering repeat purchase rate, because customers stocked up and have no reason to come back for months.
What does a minimal CRM dashboard look like?
You do not need a BI tool to start. A spreadsheet with one tab per month and one row per ASIN works for most brands under a few dozen products. Columns: unique customers, repeat customers, repeat rate, repeat sales, follower change, campaigns sent, attributed sales, ODR, median response time.
Add one notes column. It records what you changed that month: a price change, a new Tailored Promotion, an out-of-stock week. Without notes, three months later nobody remembers why repeat rate jumped. If you already use a tool such as Helium 10 or Data Dive, check that its repeat numbers match Brand Analytics before trusting them; third-party tools often estimate.
Where to start this week
Pull the Repeat Purchase Behavior report for the last two quarters, sorted by ASIN. Mark the three products with the highest and lowest repeat rate and write one sentence each on why you think the gap exists.
Then check Customer Engagement: if you have more than a few hundred followers and have not sent a campaign this quarter, schedule one for a new product or restock. Finally, look at your median message response time. If it is close to the 24-hour line, assign someone to clear the inbox daily before you spend money on any retention campaign.
Keep reading: Amazon Customer Relationship Management · Amazon