When I started my ecommerce company, I was a regular reader of top analytics blogs.
I wanted to know every trick in the book, anything that would give me an edge.
I’d spend hours every week running reports, emailing them around, and setting up custom reporting for various ecommerce metrics.
Analysis paralysis. Google Analytics is just packed with data — unless you know what you want you can waste a lot of time in there.
Eventually I figured out that the key to increasing my productivity and getting the most out of Google Analytics was to be consistent, not comprehensive. It mattered more to review a small set of key ecommerce metrics consistently than it did to review all metrics every so often.
So I developed a short list of key performance indicators (KPIs) for my business. And I reviewed those KPIs according to a regular schedule.
In this article I’m going to review the actual KPIs I used in our first year to drive the business to one million in revenue. They represent the essential metrics for a new online retailer. Use them as a basis for your Weekly Metrics Review — all you need to add is coffee.
Word of warning! The clip art options for “ecommerce metrics” are limited to say the least — you try finding something off of Flickr for ‘abandoned shopping cart’. Accordingly I have done the best I could with what the Internet gave me.
Most ecommerce startups begin with a “throw everything against the wall to see what sticks” approach to online marketing. There is a strong argument for this — you usually don’t know exactly what will work in your industry, so you need to do a fair amount of testing early on.
The problem is that resource-constrained — manpower-constrained — startups rarely go back to clean up the mess. Some campaigns stick to the wall and will drive your business; others fall into a big costly pile on the floor.
Cost Per Acquisition (CPA) will help you clean up that pile:
CPA = ($ Campaign Cost) / (# Acquisitions)
To calculate CPA you need, obviously, cost information on each of your marketing campaigns. As a Google product, Google Analytics plays nice with AdWords and is able to import your AdWords cost data automatically (Traffic Sources > Advertising > AdWords). For all other campaigns, you must either upload your cost data (hard) or run a quick calculation in a spreadsheet (easy).
Here’s an example of the spreadsheet method:
First get revenue data from Analytics. Navigate to Traffic Sources > Sources > Campaigns. Click on “Ecommerce”. Copy your transactions / visits / revenue into this spreadsheet template as conversions / clicks / sales for each campaign you want to track.
Now you need cost data for these campaigns. Again, outside of AdWords, costs are not present in Google Analytics — you’ll have to go into the reporting for your various marketing campaigns to find them. Be sure to use the same date range as you did for your revenue data. Pro tip: the longer the period you examine, the less lumpy your data will be: I recommend a trailing 30 days.
Calculate the CPA per campaign (that’s the blue line in the spreadsheet).
Log this CPA data over time (monthly is good) so that you can monitor improvements.
What do you do with this information? You kill your underperformers, grow the others. Look at my template: Facebook Ads are costing us $5 per customer, Amazon Ads cost $25 per. Work to improve the CPA of Amazon, maybe by tweaking your ad copy. Then increase your Facebook budget. Voila!
CPA is good, but it’s only part of the glorious marketing story. You want to consider return on your marketing dollars to determine which campaigns are really working for you.
I look to two ratios for this: Return on Ad Spend (ROAS) and Return on Marketing Investment (ROMI).
ROAS is a simple ratio of dollars revenue to dollars spent on a campaign:
ROAS = ($ Revenue) / ($ Campaign Cost)
ROMI is a return calculation that takes your gross margin into account. For ROMI, you can use the exact costs on the product sold for that campaign (hard) or your catalog-wide gross margin (easy):
ROMI = ($ Net Profit) / ($ Marketing Expense)
Look to the Facebook Ads column in our spreadsheet. Let’s assume the average gross margin across our catalog is 50%, which tells us the cost of whatever-we-sold was $500 and gross profit was $500. To get to a net profit, however, we need to subtract our costs for all items sold. These costs include the campaign costs (in this case $50), plus overhead, rent, personal, service time, returns, etc.
We will make things easy here and just assume the $50 campaign cost is the only additional cost. This leaves us with $450 net profit. All generated from $50 of ads. Our ROAS is 20 and our ROMI is 900%.
When you spend $50 to pocket $450, that is a good thing, no? Like an investment that makes you 900%? Sign me up!
Now compare that 900% to Amazon’s 300%. Both good returns, true, but if you had an extra $100 to spend on marketing, you’d double-down Facebook to net the better return. Also remember always double-down on “11”.
ROAS and ROMI show you where to double-down, spending any extra dollars in your marketing budget. Make sure you are tracking them.
Remember in high school when you’d plan a big party when your parents were out-of-town? You and your buddies would clear away the D&D twenty-sided die and the Nintendo, bring in the alcohol and enticing smoke machines, then invite all the hot chicks from school?
Remember how those hot chicks would walk into the party, take one look at you air-guitaring Back In Black, then walk out never to be seen again?
You didn’t know it at the time, but this is a great analogy for a website’s Bounce Rate.
Bounce Rate = (# of visitors who leave immediately) / (# of visitors)
Bounce Rates (Traffic Sources > All Traffic and Content > Landing Pages) are the percentage of visitors to a page who leave immediately — probably because they don’t see what they are looking for or the air guitar guy scares the hell out of them. It allows you to determine the relevance of your marketing campaigns: the more relevant, the higher the engagement, the lower the bounce rate.
Why do we need this when we already have CPA and ROMI? Because early in the life of some campaigns — or early in the life of your company — you won’t have enough sales data. As a result, it is often hard to conduct tests or garner insights by looking at conversion-based metrics like CPA and ROMI.
In these cases, bounce rates are a decent proxy for conversions.
If the bounce rates for your marketing campaign landing pages are abnormally high (say, 80%+), you have one of these problems:
When you notice such pages, fix them! Think about why visitors might be leaving — what can’t they find? Is there a disconnect in the marketing messaging?
Conversely when you find a page with a super-low bounce rate, increase marketing spend on campaigns targeting that page — it’s a keeper.
If I had just one metric to focus on in my consulting practice, this would be it (this sounds good as a sentence, but think about it — this would be one incredibly boring life. . . There goes Drew, he’s the “checkout abandonment rate guy”). Why? It’s a crowd-pleaser: CoAR represents the lowest-hanging fruit for most internet retailers.
Visitors to your site already have decided to give you their money, and for some reason they’ve stopped along the way. Maybe the dog needed to go out, maybe they found a form on your site confusing. The first issue you can’t control; the latter you can.
So find which it is and fix it. Here’s how . . .
In Google Analytics, navigate to Goals > Funnel Visualization to get a picture of where visitors are defecting from your checkout process. Look for highest abandonment percentage steps and develop hypotheses as to why people are leaving — is it a long registration form? Because you don’t take AMEX?
Ideally you’d want to AB test various solutions to these hypotheses, but let me save you some time . . . these four items have worked in my experience and for client after client of mine:
Remove fields from forms,
display trust and security badges,
show shipping costs up front (on the product page, in the cart, etc.)
reduce checkout steps to one page,
automatically create an account (don’t ask users to create one), and
start an abandoned checkout email campaign.
If you take on one thing from this post, this is it: focus on the CoAR.
Conversion rate is the driver of your business — it’s the percentage of visitors who buy from you:
CR = (# transactions) / (# visits)
What most people don’t realize is that each marketing campaign has its own conversion rate. In Google Analytics, navigate to All Traffic Sources (Ecommerce). Sort by visits and look to the CR column. See? There is a real range of values there.
Check this report regularly; improve your poorly converting campaigns.
Pro tip: I like to create an Advanced Segment here to see if I can garner more insights. Create one for visitors with AOVs over 2x normal. These are your “Whales” (Being in internet retail allows you to refer to customers as “whales”, BTW — don’t try this in a brick-and-mortar store). See if they are converting higher than average from certain campaigns or keywords or sources. Focusing on these “Whale”-generating campaigns will really move your top line.
Most ecommerce sites are so focused on CR that they neglect 99% of their visitors who don’t convert.
This is a massive, massive mistake.
Why? Because most people won’t buy on their first visit to your site — they’ll want to check out your offering, think about it, talk to the husband, etc. A good email marketing program is a way you can still capture these visitors, market to them, build up trust, and eventually ink a sale.
Treat your email opt-ins as a secondary goal of your site. Know your Opt-in Conversion Rate (oCR):
oCR = (# opt-ins) / (# visits)
Add an email sign-up as a goal in Google Analytics — then whenever you view goal reporting you’ll be able to track your oCR. Eventually, after three or so months of email campaigns, you should be able to put a monetary value on each email subscriber. This is easy — just look back a few months and divide the total revenue by the average number of list subscribers. Calculate it, then add it to the relevant goal in Analytics.
Watch the oCR funnel as you would your CR funnel, and optimize it in parallel. Thank me when Google rejiggers its search algorithm killing your SEO and leaving your business with one viable channel you control: email marketing.
This final metric is for all the merchandisers out there.
Product Revenue is the amount of revenue driven by each product in the catalog. You can see it by looking at Conversions > Ecommerce > Product Performance and sorting by Product Revenue.
This report tells you your product workhorses — the items that really drive your top line. Flag them in your catalog. Use that flag to feature them in blog posts, newsletters, and on the homepage. Add them as related items to garner up-sells.
I should note that when I view this report, I like to set an advanced filter restricting Unique Purchases over a certain number to ensure that I do not optimize my site around a random big order.
By paying careful attention to what is selling, you’ll be better able to merchandise and market your offering going forward.
That’s it — my seven go-to metrics for new ecommerce retailers. I’d be remiss if I didn’t mention a few things that aren’t on this list, things like Lifetime Value, repeat order rates, cohorts, inventory turns, and multichannel attribution. All these are important metrics and reports, but for most online retailers they aren’t essential in the early years of running a company, especially on the path from zero to one million in sales.
When bootstrapping my company, I had tons to do in limited time, so I had to adopt a Pareto approach and focus on the metrics that mattered. If you are an early stage bootstrapper, I suggest you do the same. Pick from this list and set a schedule you will stick to.
Remember that “pressure over time” wins in ecommerce — I hope this list will get you started.