This post was updated on 14/10/2016 with links to latest findings. Give it a read for fresh insights.
Have you met Richard, the eCommerce entrepreneur? This is the story of how Richard found the perfect eCommerce metric to track and measure eCommerce performance.
Richard runs an online store selling two kinds of water bottles – a generic item worth $1 and a premium designer edition worth $100. As a data-driven marketer, Richard decided to look towards analytics.
Then it hit him.
The analytics universe is full of curiously named metrics. Which of these metrics should he track over time to measure his eCommerce performance?
That should be easy, Richard figured. He’d reach out to the experts.
Expert 1: Go For Conversion Rate
A conversion is any desirable activity performed by a visitor on your site. From a revenue perspective, conversion is checkout. Conversion rate(CR) is simply
Conversion Rate = Number of Checkouts/Number of Unique Visitors
If you have an average 1000 visitors to your site on any given day, and 50 of them become customers your conversion rate is 5%.
Optimizing for conversion rate will make more visitors into paying customers.
How does it help?
- Converting more of your current visitors is more cost effective than acquiring new customers
- It essentially gives more revenue at the same cost
Since you are already paying in some way to acquire traffic to your website — through PPC, SEO, Email — it would be a great idea to convert more of those visitors into customers. It brings you more revenue for each dollar spent on acquiring traffic.
It made sense.
But Richard wasn’t convinced. He had once conducted an A/B test on the product page and this is what resulted.
Overall conversions increased by 10% and his website that used to convert 1100 customers started converting 1210 visitors.
It was a moment of triumph.
And it lasted exactly a moment.
Later analysis showed that revenues had actually dropped because the conversions among high paying customers had declined.
Unsatisfied, Richard reached out to Expert #2
Expert 2: Without Doubt, Average Order Value is What You Should Be Tracking
Average Order Value(AOV) is just what it says. Total revenue/Number of Checkouts. It’s a direct indicator of what’s actually happening on the profits front.
Average Order Value (AOV) = Total Revenue/Number of Conversions
In the last A/B test he conducted, optimizing for conversion rate alone had left Richard susceptible to the blind spot — the average order value.
Despite the increase in conversion rate, Average Order Value had dropped by more than a dollar, resulting in an overall decrease in revenue.
How does it help?
- Comparing AOV against Cost Per Order gives a great idea of the profits you make on each order. Consider your Cost Per Order (shipping costs etc.) is $1 and your AOV is $10, giving you a profit of $9 per order. By increasing AOV by 10% to $11, you stand to gain an additional profit of $1 per order.
Here’s a statistic to remember: AOV in United States during 2014 Q3 was $72
- There are a lot of simple hacks to increase average order value, including upselling and cross-selling
- It’s just easy, see how Paperstone increased AOV by 18.94% using A/B testing
This was great news.
At this point I should tell you that Richard didn’t go alone to Expert 2. Tom, Richard’s best friend since that last A/B test hiccup, was there too.
Doubting Thomas asked,
“What if we successfully increase our AOV by bumping up the minimum order value for free shipping, but less people buy as a result? Our revenue could take a hit, harming Richard and his profits while still showing a higher AOV.”
Tom had a point, Richard thought. It was similar to what happened with his last test. There he had forgotten to take into account AOV and suffered. Tracking for AOV alone could make him blind towards conversion rate resulting in a revenue sheet like this:
There had to be something better. A metric that combined both Conversions and AOV to give the whole picture.
Hoping for better, Richard and Tom reached out to Expert 3.
Expert 3: Track Revenue Per Visitor, Dodge The Rest
Revenue Per Visitor(RPV) is deceptively simple. It tells you how much revenue each unique visitor is driving.
RPV = Total Revenue/Total Unique Visitors
Why is it so potent?
The trick is in understanding RPV from another perspective.
We already know that
Total Revenue = AOV x Number of Conversions (checkouts)
So we can rewrite the RPV equation this way:
RPV = (AOV x Conversions)/Total Unique Visitors
and since (Conversions/Total Unique Visitors) = Conversion Rate
RPV = AOV x Conversion Rate
The great thing about the RPV metric is that it combines both AOV and Conversion Rate.
What’s important for any eCommerce business?
For revenues, first you need traffic. Once you are able to attract traffic, increasing revenue is two-dimensional process:
- Convert more visitors into paying customers (Conversion Rate)
- Increase customer-spend per conversion (AOV)
RPV involves both these dimensions leaving no blind spots.
Avinash Kaushik recommends using an ‘actionability test’ before choosing any metric to track. The idea is that any metric you track should help you take definitive actions to correct/improve business.
Does RPV pass the actionability test?
With a crisp dollar certificate.
If there’s a drop in RPV, it could be due to
- A sudden increase in visitors without any buying intent (drop in conversion rate): Check if there has been any recent marketing activity that brought a lot of unqualified visitors with low buying intent. Use segmentation to understand what channels are bringing the right traffic.
- Customers are buying less of high-value goods and more of low-value goods (drop in AOV): Consider using a recommendation engine. Read the article I’ve linked to under the section above titled ‘AOV’ for 8 quick ways to improve AOV.
Touting RPV as a very useful metric to track does not take anything away from metrics like Conversion Rate or Average Order Value. It’s important to understand that metrics simply show symptoms, and different symptoms become visible through different metrics. RPV is simply one that helps you see the bigger picture.
Although it’s a lot of metric talk to take in, Richard feels he’s found what he was looking for – one metric that he could keep track of to measure his eCommerce success.
He thanked Expert 3 and got ready to leave.
“Wait!”, Tom had more doubts.
Why Use Unique Visitors and Not Total Visitors?
Expert 3 cleared his throat and explained.
Of all first time visitors to an eCommerce site, 99% won’t make a purchase. The typical buying cycle involves a visitor first visiting your site to check out the products, leaving to compare prices elsewhere, consulting a few friends, reading reviews and eventually a trip back to your site for the purchase (if at all a purchase decision is made). There could be even more steps involved here.
Using total visitors (unique and returning) bloats up your metric denominator considerably, resulting in small figures and giving you less credit than you otherwise deserve.
This is not to say it’s a bad practice, just sub-optimal. (In fact, if for some reason, you are getting many orders from repeat buyers it might even make sense to use total visitors instead of unique visitors.)
Using ‘unique visitors’, on the other hand, paints a real-world picture of what’s happening with your users, who are, of course, unique.
With this explanation, Doubting Thomas went poof, and Richard went back wiser.
What’s Your Doubting Thomas Wondering?
What metric have you found most useful to track? And why. Getting your perspective as a practitioner would be invaluable.
We’ll soon be coming out with a brilliant guide on understanding all the right metrics, including the bad-ass ‘Customer Lifetime Value”.