Chances are most of you didn’t attend WooConf this year, where I presented on Value Metrics for New Stores 🙂 I’m going to give you a recap of that talk today, and if you want to check out my slides as you read, you can view them here (UPDATE: or view the full presentation here).
I come from a baseball family — it was always a part of our lives at family gatherings, and I played softball from kindergarten all the way through to the division I level at college. As a result, you learn stats, and you memorize famous ones to face off in trivia against your cousins. Leading MLB batting average of all time? Ty Cobb, at .367 (don’t give me .366 Wikipedia-ers, check the official MLB stats). Second? Rogers Hornsby, .358.
I learned some interesting stats about college baseball as I grew up, too. One that always stuck out to me: Larry Bird (yep, the basketball player), had a college career batting average of .500. However, he doesn’t have the career batting average record. Rickie Weeks gets it at .465 — why? At a full 35 points lower, why does he get the title and Larry Bird doesn’t?
The answer lies in statistical significance. The NCAA won’t count a batting record unless you have at least 200 at bats (the MLB requires 3000). The reason is that they know you need enough data points for an average to be significant. In Larry Bird’s case, he was 1-for-2 in college, and 2 at bats doesn’t cut it for a significant metric 🙂 His college career was a publicity stunt to encourage attendance at baseball games. (He did have an RBI single though!)
Without enough data points, we don’t know if a metric is representative or not — Larry Bird may not have ever gotten another hit again after that second at bat, and his “true” or representative batting average may have been a lot lower. It seems natural we shouldn’t base records on data samples this small, right? Otherwise a lot of players might quit after getting a hit on their first at bat.
However, we count small or insignificant metrics like this all of the time on eCommerce sites. We talk about optimizing conversion rate before we even have enough orders to know if this conversion rate is a representative, significant metric. One quote on significance I love is from Peep Laja in the eCommerce Fuel podcast:
If you have less than 1,000 transactions per month — the transaction could be either purchases or whatever you want to test, like email sign ups…it’s probably too early for you to [split] test.
Tons of shops I’ve worked with don’t yet hit this 1,000 orders or so per month benchmark, so testing can take an extremely long time to reach significance. Optimizing your conversion rate before this point won’t necessarily give you conclusive, significant results.
Instead, there are some key metrics your store can focus on instead as it grows until you get to a larger number of orders, allowing you to test and optimize other metrics in a significant way. We’ll look at 3 metrics that are helpful, regardless of the number of orders you receive.
We’ve written a fair bit on average order value already — it’s the amount of money you generate from every order, and you can calculate it by doing:
AOV = Total Revenue in a time period ÷ Total order count in that time period
Since we’ve already established we don’t want to optimize metrics that rely on our order count, as we can’t measure or optimize them in a statistically significant way, optimizing average order value is a good way to generate more revenue as you grow. Instead of focusing on generating more orders, you can focus on getting more money on each order you already get.
So how can you get this metric? My talk focused on getting average order value if you use WooCommerce, but we’ve written about getting this with other eCommerce plugins before.
For example, a lot of eCommerce platforms will give you total revenue and order count by date, which you can use to calculate your own AOV.
If you can’t use built-in reporting, you can have analytics programs like Google Analytics track this for you instead. You’ll need a Google Analytics plugin that adds eCommerce events to your site tracking, such as:
- WooCommerce Google Analytics (free)
- WooCommerce Google Analytics Pro ($29)
- Easy Digital Downloads Enhanced eCommerce Tracking ($59)
Once you know your AOV (which you could get within minutes of reading this article!), you can use it to drive marketing decisions and begin to optimize it. We’ve got 8 strategies you can use to optimize AOV, along with a series of articles on helpful tools.
While knowing your AOV can help you drive more revenue, you also need to know how much of your revenue is going into your bank account. This is where Average Order Profit comes in. You can calculate it by doing:
AOP = (Total revenue in time period − cost of goods for that time ) ÷ total order count
This will help you determine what money is available and how you can allocate that to marketing, customer acquisition, and other operational costs. You’ll know how much you can afford to spend to get each order, especially since you may not yet know your customer lifetime value.
The problem, however, is that most eCommerce plugins have no way of knowing what your set costs are for each order. As there are tons of different kinds of shops, such as digital-only shops, that may not have a cost per good, this usually isn’t included in a core platform.
You can track costs externally in your accounting software or spreadsheets, or if you use WooCommerce, you could look into using the Cost of Goods plugin ($79). This lets you set a cost for every item you sell, so that profit for a date range and average profit per order are automatically calculated.
Once you know how much you profit on each order (based only on set costs like item costs, shipping, taxes, etc, not your “operating costs” like salary), you can more intelligently allocate costs, and try to optimize it as you grow.
For example, you can negotiate rates with your shipping provider, or optimize your supply chain by buying in bulk at lower prices.
Now that you know how much revenue you bring in, along with how much profit you have overall and per order, you can take a look at other costs, such as customer acquisition cost. It’s important to know how much your growth and acquiring new customers will cost. You can calculate CAC pretty easily:
CAC = Total marketing spend in a time period ÷ total new customers in that time
Most businesses track marketing spend already in their own P&L, but you could track this as simply as just adding every advertising or marketing purchase you make in a spreadsheet. Just have this number somewhere.
Now what about new customers? Here’s where things get dicey. Most plugins will give you some basic customer reporting. However, they tend to only give you customer registrations, not how many “new” customers (first-time purchasers) you have in a time period.
Unfortunately I wasn’t able to find a great way to do this with most platforms or analytics tools like Kissmetrics or Mixpanel. If you’re a WooCommerce store owner, you’re in luck at least since I decided to solve this problem for my presentation 😉
You can download WooCommerce New Customer Report to get a count of first-time purchasers in a given time period, whether they’re logged in or guest purchasers (based on email address).
Once you know how much you’re spending per new customer, you can make some decisions about how you’ll grow — do you have investment money available for growth? Are you trying to stay within AOP to grow until you know how many repeat purchases you’ll get? What channels can you afford to use to acquire customers?
You often hear “measure everything!” and “what gets measured gets managed” in reference to eCommerce sites. I don’t necessarily disagree with this advice; I just disagree with the idea that you should try to do something about each metric you measure, or that each metric should hold the same weight in your mind while growing your store.
It can be crippling to divide your focus among tons of different metrics and trying optimize each when your primary focus should be on growing in a sustainable, long-term way — especially if you can’t even measure certain metrics reliably yet.
Tony Perez from Sucuri wrote a great post on this recently, and this quote sums his ideas up for me:
The harsh reality is that while it’s fun to spend time reading case studies on the wacky testing schemes employed by organizations we all aspire to be, they are rarely practical insights for you. It’s not because testing is impractical, but because most of us don’t have enough data to actually test against in a timeframe that makes sense to make any meaningful decisions from.
Until you have enough data to run meaningful tests and optimize metrics that require a significant number of data points, there are other metrics you can be focusing on to ensure the viability of your eCommerce business.