The standard answer to the question of how you are measuring conversion rate, from someone versed in Amazon's 3P Seller Central might be Ordered Item sales divided by Sessions. This is a fine answer if you need to know what fraction of buying opportunities turn into a sale. Ask the same question of a first party (1P) vendor/rep, and you will not get the same, or any answer since they neither know the numerator -- the number of ordered (line) items, nor the denominator -- Glance Views (GVs), which Seller Central terms Sessions or their close surrogate, Page Views. So why is conversion rate important? Who should care & why? Two interested parties at least should care: advertisers and vendors. - Advertisers should use this insight to calculate their break even ad-spend per impression. For example, if you make $10 on a sale, and one in 10 who clicks buys, then in principle you should only spend up to $1 per impression to make a sale. The poorer the conversion rate, the less you should spend on ads. - Vendors (& their reps) should care because poor conversion rates may indicate opportunities for improvements in listings -- better keywords, better pictures, more compelling copy, etc.
This topic resurfaced when a client asked how he could better align his KPIs with his organizational structure. Like most vendor rep groups, he has an advertising group charged with getting the "right eyeballs to the right product", and another group charged with content/search engine optimization i.e. getting a product indexed for the right keywords to facilitate organic search discovery, and generating/maintaining compelling content so that the customer converts/buys. The content task is not trivial and several of the components of a 1P buying decision are beyond the control of the vendor. For example, Amazon sets the price of its 1P offerings, and reviewers chime in to rate the product. So what is controllable here, and what is the scare resource? Assuming the listings are already active, the scarce resource is people to improve the content. Recognizing what is under control -- what listings to turn their attention to, and how much effort to put into it, leads us to a good metric.
I propose that for performance measurement of the Content group, they use some variation of ShippedCOGS per Glance View.
This is because their inputs from organic search or ads are Glance/Page Views, and that they are likely trying to maximize Revenue, Commission or Margin as a result. The other part of the story is where to focus attention? I contend that benefits are capped by market size. If every Glance View converted at 100%, you've captured the Total Available Market (TAM). Knowing last week or recent sales for each ASIN and the conversion rate gets us to TAM (as it was last week). Knowing ShippedCOGS per GV (or some commission-weighted variant) gives us the expected value of a sale to us, especially if you put the cream on the cake by adding future forecast volume insight.
Here is a sample analysis:
This analysis identifies two classes of opportunity: - products or groups of products converting well that merit extra demand generation efforts - products getting discovered but converting poorly, perhaps because the offer is not compelling, or, because it shouldn't have been offered in the first place (i.e. a negative keyword opportunity)
Unfortunately there is a wrinkle -- Amazon doesn't tell you Glance Views, only the percentage breakdown of views across your products in each market last week or last month via the Traffic report. This is not a show-stopper as a little analysis over time can get you to a decent estimate.
Imagine being able to continually review such insights every week as new data becomes available. As the manager, you could look across your entire portfolio of clients, or within a client's portfolio to direct your team members on what to do next based on latest information. Such continuous focus on the smartest products to do next led to the near doubling of revenue per GV for one client
Such analytics are not trivial unless you have the right tools. Merchant AI routinizes such analytics and performance measures so that the latest understanding is automatically available within minutes or hours of the underlying data refreshing.
And the linkage of the Oliver Twist picture to the article? Nah ...that's the Content function asking for more work when properly incentivized!