Understanding Metrics: Online Revenue
This is part two of our series on Understanding Metrics. For a high-level overview of the topic, check out the introduction blog.
Let’s start off with the simplest, most measurable goal there is online revenue, money received directly by your web ecosystem—either your own site or a third-party provider connected to it. This revenue can take several forms, but in all cases, the web interface is feeding money directly from the world to you.
- User payments: An online store accepting payment directly for products or services, or a donation system.
- Advertisements: Revenue paid by third parties based on how your users make their way through your site and what they see or click on.
“But Joe,” you say, “e-commerce sites don’t seem all that simple.” Certainly, that’s true, imaginary reader, but the same complexity that goes into their construction and management also lends very well to being measured and understood.
The effectiveness of online revenue as a goal comes from the direct, meaningful line from business goals to site goals to metrics. The goal of your business is to make money. The goal of your site is to make money for your business. You measure how much money your site made. Revenue is inherently quantitative. There are no proxy measures or intermediate assumptions that we’ll see in less concrete goals.
This doesn’t mean that there’s no room for variety in what, exactly, you measure. Online revenue has several variants, each interpretable in its own way:
- Revenue per customer. Has revenue per user dropped, indicating that it’s time for a little prompting? Has a particular user started buying much more, hinting at a change in circumstances? This value is the workhorse of customer measurement and understanding.
- Revenue per item. Understanding the characteristics of what products are money-makers is key to future product development and marketing strategy.
- Transactions per customer/item. Straight profits only tell part of the story. A user with high transaction count and low overall revenue may still have great value as a loyal user of the site. Often-purchased but low-revenue items may be a candidate for price adjustment, or a great flagship marketing item to get customers in the door.
It’s in the interaction of all these facets that the revenue metric also becomes predictable. Recommender systems and targeted advertising are grounded in the principle that users are alike if they buy similar products, and products are alike if they are bought by similar users.
All these handy metrics are great for the direct stream of online revenue, but what if all of your transactions happen offline, in a brick and mortar store? In the next installment, we’ll explore the pitfalls and struggles of linking online behavior directly to offline revenue.
Want to know more about understanding metrics or want to work with us to apply these ideas to your business? Visit GearLab’s website to view our offerings and send us a message.
This article was originally posted in 2017 as a series about Data Science. It has been updated for 2019.