What to Promote? An e-commerce dilemma!

Understanding your fuel and getting the maximum mileage

Introduction:

E-commerce marketing usually involves 3 verticals: User Acquisition, Content Promotion, and Retention & Reactivation. This article mainly deals with content promotion bit.

If you are also an e-commerce marketer, you would often get into discussions like which brand to promote or which categories to promote more, whether you should focus on completing sales targets for one brand or focus on ad revenues from others, etc., etc. And then you get confused about which metric to give more importance to and which holds more financial relevance. Is it the Conversion Rate (CR), commissions, RTOs, Returns, product prices, or anything else? Which metric among these would give me my answer to…

What should I promote?

Sell everything they said!

Understanding what is my Fuel & what defines my mileage?

The purpose of any on-site content promotion is to get maximum purchases and eventually reach the maximum revenue for the platform. But to reach this level the most critical thing you have for your customers is Product Views. What customers see is what they buy. Therefore,

product views are essentially the fuel

for an on-site promoter and his/her goal should be to get maximum revenue from the limited fuel (Product Views) s/he has in his hand.

Now before we move forward, revenue also needs a clear definition here.

Revenue simply means the earnings per product view for the ecommerce marketplace.

Metrics that play a crucial role in defining this revenue are:

• Conversion Rates (Product View to Purchase)
• Delivery Rates
• Commission per product
• RTO/Returns
• Is a product on promotion or not?
  etc.

But I need only one number to look at!

A marketer always has an argument that what sells is in his hands but what happens with the order after s/he sells it, or how much margin they make on that product is not something that is in his hand and hence s/he should not be looking at those numbers.

This argument, though for a specific domain (marketing) stands true, but at an organizational level it proves to be untrue. There might be products that sell very easily but might have high returns or there might be products that sell in large numbers but have very low commissions (or margins) behind selling them. The purpose therefore of the entire organization is to make money more efficiently and hence the goals and performances of every department of the company need to be aligned.

In an attempt to incorporate all underlying metrics which affect the business under one common KPI, we came up with the concept of RPMV i.e. Revenue Per Mille Views (Or Revenue Per Thousand Views).

Revenue Per Mille — Only Metric to Look At

The attempt here was to simplify the business aspects for a marketer and establish only one metric s/he needs to look at and optimize for.

RPMV in simple words means:

Net money you make per thousand product views

To calculate this metric we simply took a sum of the total product views vs. the total revenue (earnings) our portal was making out of the same.

Where Total Revenue can be defined as:

Definitions:
Commission     - Net absolute commission on the product (basis ESP & commission margin)
COD Fee        - COD charges (if applicable)
Delivery Fee   - Delivery Charges (if applicable)
Marketing Fee  - Any promotion charges levied for the brand (if applicable)

Advantages of RPMV

As mentioned earlier, RPMV helps a marketer look at only one key metric s/he has to look at and hence simplifies a lot of things in his/her life. They therefore just need to look at which brand/category generates the maximum RPMV is the one that needs to be promoted more.

This also removes the price sensitivity among categories and brands. For example, in the shoe category, by nature, a premium brand like Steve Madden would have a low conversion rate as compared to a brand like Lancer. For that matter, it may even have a lower commission percentage. In such a case, if we just look at conversion rate numbers, we would probably never promote Steve Madden. But if we dig a level deeper and see the total commission made, there might be a different picture. See the below example (with sample values):

Brand 1: Steve Madden (Sample values)
ESP = ₹ 5,000
CR = 0.5%
Margin = 10%

Brand 2: Lancer (Sample Values)
ESP = ₹ 700
CR = 0.8%
Margin = 20%

For the same 1,000 product views, RPMV for the respective brands will be:

RPMV (Steve Madden) = (5000 * 1000 * 0.5% * 10% ) / 1000   = ₹ 2.5

RPMV (Lancer)       = (700 * 1000 * 0.8% * 20%)/1000       = ₹ 1.12

Here, we can see that though the Conversion Rate of Lancer is more than that of Steve Madden and in fact, even the commission percentage is also more, the fact that Steve Madden has a high ticket size, it eventually gets us twice the money as Lancer with same 1000 product views. And this is where the thought process should be aligned towards — Getting maximum revenue per 1000 product views.

Best Way to Study RPMV

The question then comes is what is the best way to slice and dice the data according to RPMV. The best approach so far seems to be by studying RPMV at 3 levels:

• RPMV by brand
• RPMV by category
• Total RPMV

This would give us the best-performing brands on the portal (considering all underlying key metrics), best-performing categories, and overall improvement of the number month on month.

The on-site content team will now also have a localized target attached to them which will be much more focused and purposed around making the highest RPMV on an overall basis.

The more noise you cut out for your team, the more concentrated can they work on their respective targets.

Further Detailing:

Though RPMV can not be the end of the optimization, it certainly is a start. A couple of things that a person can work on next are:

1. Improving the top line funnel itself i.e. increasing your fuel itself per user. Example: Do certain changes in your home page, category pages, notifications, etc. that same count of people now start looking at more products (basically discovery getting easier for them)
2. Incorporating post-order metrics in RPMV. These can be by not only adding RTO and Returns number to the RPMV formula, but also the cost implication they had.

Now, we did mention both of them in a line or two, but in actuality, these have complexities far behind what I intend to cover in this article.

So, for now, if you are reading this and you haven’t even reached the level of optimizing on RPMV, then get started with that. And in case you want me to go into further details or the next level of optimizations which one can focus on, then do let me know in the comments.

This article is an initiative to share the minimal learnings I had during my career journey. Idea is to share my knowledge with people who are in the early phase of their careers and help them understand concepts as easily as possible.

In case you liked the article or would want me to cover any specific topic, do write to me at puneet14@micamail.in