I just got this question from a colleague who I consulted with to help get them set up on Amazon Seller Central a couple of years back. Their goal was to manage the platform in-house and we were able to help them get to that point, but they still like to ask a few questions and get some input here and there – which I’m happy to provide.
The question they asked today was about their top product’s “Best Seller Rank” (BSR) on Amazon, and I figured since I’m already typing out an answer that I might as well share this with others too.
I’ll summarize and anonymize the client’s question to protect their privacy. But in general, it went something like this:
A few weeks ago our main product’s BSR was anywhere from 5,000 to 10,000. Then our listing got suppressed for a couple of days, and since coming back live the BSR was closer to 20,000. It has now come down to 17,000. That said, our sales before and after the suppression are virtually unchanged. So why did the BSR change so much?
My response:
Before I answer this directly, I want to advise that you don’t give a ton of weight to BSR as a key performance metric. BSR is relative within a category. So hypothetically you could sell exactly the same number of units this month as you did last month, but the BSR could move up (worse) if other products within the same category sold more. Likewise, the BSR could move down (better) even though you sold the same volume, if other products sold less this month.
Now this is assuming we’re talking about a product where the BSR is not in single or double digits, like the case above. If you are in a category where your product’s BSR is anywhere close to reaching #1, then it is worth obsessively tracking this and trying to get the #1 BSR in that category – as that will give your product the coveted “Best Seller” badge, which can have a significant impact on clickthrough rates and conversions. Pro Tip: It can be worthwhile looking for smaller niche categories where your product can be listed and potentially get the #1 BSR value, instead of keeping it in a very broad category where the BSR is much higher (worse).
With that disclaimer aside, in evaluating the performance of a product over time these are the things you should mainly be focused on:
- Impressions (i.e. detail page views)
- Sales
- Conversion Rate (sales / impressions, or use “item session percentage” as a proxy for this)
- ACOS (for paid ads)
- Organic keyword rankings for top keywords (via a keyword tracking tool)
Plot those over time and then you should be able to see the impact or net effect of changes you make to the listing – like changing prices, photography, titles, changes in reviews, etc. Or just your overall performance trend in light of competition.
Now I’m sure you are thinking “well that’s nice Jon, but I still want to know why my BSR is so much higher?”. Okay fine, I’ll answer the question already. There are 4 possible explanations that immediately come to mind for why the BSR would get worse even though sales stayed the same. They are:
1. BSR is Category Specific
A first thing to check is the category. Some products are listed on Amazon in multiple categories (“browse nodes”) and the BSR is specific to each category. So if you see wild swings in BSR, first make sure you are comparing the BSR specific to the same category.
2. BSR can be Child-Variation-Specific
This one is frustrating, but Amazon is inconsistent with how they calculate BSR in the case of parent-child relationships. Let’s say you are selling widgets and have variations for three different colors. Depending on the category/store, Amazon may combine the sales for all three variations to compute the BSR at the parent level, OR they may calculate it only at the child level. In the latter case, you might notice big BSR shifts among whichever is the top-ranking child product when your sales shift from one child to another over time – even if the total sales at the parent level is the same.
3. Seasonality Impacts BSR.
Once again, BSR is category-specific. But if you are in a category where sales increase in October (i.e. related to cooler weather) then this would make a lot of sense. Also, BSR is funny when it comes to variations and usually just pertains to the top variation so a shift in sales between the different child variations could conceivably reduce (well increase the number, meaning worse) the BSR for the top variation whereas your total sales could stay the same. Put another way, your product might not be seasonal and might sell consistently which is just fine. But many other products in the category could be seasonal and thus move up or down over time and impact your BSR as a result.
4. Sample Size – BSR is Computed Hourly!
One more thing… I’m assuming here that the 17,000 BSR figure you referenced is averaged out from multiple readings. BSR is re-calculated by Amazon hourly, so it is constantly changing. Certain products tend to sell more at certain times of day (or certain days) just depending on the market. So if you have a product that tends to sell more on weekdays and you check the BSR on the weekend, it will look much lower than it actually is if you average out the data points over all 7 days. Likewise if its a product that sells more in the evening and you check in the early morning. So you need to make sure you are taking that into account.
So in summary, I do think BSR is useful and worth keeping an eye on (especially for key products)… but its also important note to pay too much attention to it or stress about it. Sharp changes can help us identify potential issues, but really if you are monitoring impressions, sales and conversion rates then we should probably catch those issues anyway. At the end of the day its sales revenue and margins that we really want to focus on, whereas BSR is a very indirect figure and has its limitations as discussed above.
Any other thoughts I missed? If so, let me know!