Which metrics can reveal how well your online shop is performing? Here are the essential KPIs you should watch in order to monitor and optimize your shop’s performance.
A primer on KPIs
A performance indicator is a quantitative metric that is associated with the accomplishment of a goal, and hence can show how well you are scoring against it.
For example, if your goal is to increase customer loyalty (good one!), your Repeat purchase rate would be a great performance indicator. An increase in the repeat purchase rate would mean you are accomplishing your goal. If your goal is to increase your basket worth, a nice performance indicator could be Average order value (AOV).
As you can see, virtually any measurable goal could have its performance indicator. If we limit our choice to a few metrics that monitor only key activities and goals for shop, those will be our Key performance indicators.
Choosing the right KPIs
While every business is intrinsically different, every online shop will share some common business processes: attracting potential customers, engaging with them and converting them, determining the business impacts of conversions, and following up on customers.
For each process we will identify some key goals which will point us to which KPIs to track.
Segment your KPIs
KPIs fully express their informative potential when you start segmenting them. If you register a peak or anomaly in one of your KPIs, segmenting will let you explore the causes and find opportunities for improvement.
It is particularly valuable to segment by: acquisition sources, single campaigns, landing pages, product categories, customer cohorts, and various time spans. A business intelligence tool will allow you to easily segment and filter your dataset by these variables.
1. Acquisition
The first key objective for an ecommerce website is to attract potential customers. There are several ways to attract traffic and you should track how well each of your acquisition sources is performing in order to optimize your marketing spend.
Key goals: Attract traffic. Monitor how well your acquisition channels and single campaigns are performing, in order to optimize marketing spend.
Cost per acquisition (CPA)
Segment by: Acquisition sources.
Why it matters: CPA compares your investment on different channels with how many customers they delivered. It tells the costs of attracting a single customer, and which channels are more cost effective.
Data sources: You can find conversion data either in your shop’s database or web tracking tool, while marketing spend data will be in your different ad platforms. You can collect and merge your data in Excel or use a BI tool which will automatically join them for you, and calculate CPA.
Context: When segmenting for acquisition sources remember to look at the absolute number of customers for context.
Benchmark: Highly depends on your margins and customer lifetime value. Your cost per acquisition should never be greater than your customer lifetime value (CLV).
Cost per order (CPO)
Segment by: Acquisition sources.
Why it matters: CPO is analogue to CPA, and shows you the cost of getting a single order, and which sources are more cost effective.
Data sources: Conversion data lies in your shop’s database or web tracking tool, while marketing costs lie in your different ad platforms. You can collect and merge your data in Excel or use a BI tool, which will join sources and calculate CPO for you.
Context: When segmenting for acquisition sources remember to look at the absolute number of orders for context.
Benchmark: Highly depends on your margins. Your cost per order should never be greater than your average order value (AOV).
Cost income ratio (CIR)
Segment by: Acquisition sources.
Why it matters: Cost income ratio (CIR) compares your spend on each acquisition source with the actual revenues each one generates, to show which has been more effective.
Data sources: Revenues are in your shop database, while marketing costs are in your different ad platforms. You can merge your data in Excel or use a BI tool.
Benchmark: The lower the better, CIR should always be smaller than 1.
Marketing ROI with Customer lifetime value (CLV)
Segment by: Acquisition sources, single campaigns.
Why it matters: Marketing ROI shows the return on your marketing investment for each channel and campaign. There are multiple ways in which returns can be calculated. The most accurate and sophisticated, is to use customer lifetime value, or the maximum profit a single customer will generate for your company throughout his/her life. This calculation encourages you to shift your focus from immediate revenues to healthy long term customer relationships, and lets you maximize your marketing efficiency, spending the right amount on acquisition, considering the entire lifetime of a customer.
Data sources: Conversion data, revenues and COGS are in your shop database, while marketing spends in your different ad platforms. Simple calculation of Marketing ROI, with gross profits as returns, or with a basic CLV formula, assuming margins and retention rate are constant over time and using an infinite time horizon
![CLV=margin\times[\frac{\text{Retention rate}}{(1+\text{Interest rate}-\text{Retention Rate})}]](https://s0.wp.com/latex.php?latex=+CLV%3Dmargin%5Ctimes%5B%5Cfrac%7B%5Ctext%7BRetention+rate%7D%7D%7B%281%2B%5Ctext%7BInterest+rate%7D-%5Ctext%7BRetention+Rate%7D%29%7D%5D&bg=ffffff&fg=000000&s=2 "CLV=margin\times[\frac{\text{Retention rate}}{(1+\text{Interest rate}-\text{Retention Rate})}]")
A more accurate and sophisticated ROI calculation with Customer Lifetime Value can be obtained with BI tools, as it involves statistical modeling.
Context: When segmenting for acquisition sources remember to look at the absolute number of orders for context.
Benchmark: Varies greatly for each initiative, should always be positive at channel and campaign level, optimize or kill anything that is not.
2. Engagement & Conversion
How does the traffic we just acquired behave once it lands on site? The second key objective of an eCommerce website is to provide potential customers with an optimal experience that will lead them into converting.
Key goals: Monitor how well the website is able to engage users and lead them to conversion in order to optimize website layout and contain funnel leaks.
Bounce rate:
Segment by: Landing pages.
Why it matters: Measuring how each landing page lives up to the promise it made in organic or paid search will show the site’s lost potential and where you can improve.
Data sources: Your web analytics tools will provide you with bounce rates for each page. (E.g. Google analytics > acquisition > SEO > landing pages)
Benchmark: Bounce rates can vary from as low as 10% for established sites to as high as 35% for small ones. Source: Clicktale 2013 Web Analytics Benchmarks Report.
Cart abandonment rate & Micro conversions
Segment by: Acquisition sources.
Why it matters: Cart abandonment rates will show you the lost potential coming from leaks in your shopping checkout process. This can help you optimize your site layout in order to fix the leaks.
Data sources: Your web tracking tool will show with this metric provided you set funnel steps for each check-out step.
Benchmark: Abandonment rates are around 60%, add to cart rates can vary between 5%-12% of visits. Sources: Baymard research, Monetate ecommerce quarterly
Conversion rate (macro)
Segment by: Landing pages, acquisition sources, devices.
Why it matters: Conversions mark the precise moment of revenue generation and you should monitor where traffic that converts comes from, which landing pages and which devices convert better.
Data sources: Your web tracking tool will show your conversion rates, although to assign an economic value to the conversion some tweaking is involved.
Context: When segmenting, remember to look at the absolute number of conversions for context.
Benchmark: Conversion rates can vary between industries and countries, from 1.5% to 5%. Source: Monetate eCommerce quarterly .
Honorable mentions
Assisted Conversions
New and Returning visitors Conversion rates
3. Outcomes
Congratulations, your visitors are converting into sales! It is now time to measure the business impacts of these conversions. This phase holds the largest share of KPIs, which monitor operations, and will inform you on how well your core business activities perform.
Key goals: Firstly, it is important to always have the pulse of the business under control. Then, to know how different products and brands contribute to the bottom line in order to fine tune merchandising and inventory, as well as spotting up-selling opportunities.
Net sales & Gross profits
Segment by: Product categories, acquisition sources, customer cohorts.
Why it matters: Revenues are the first outcome of your conversions, and the lifeblood of your business. You should track daily, weekly, monthly and annual development.
Data sources: In your shop’s database. Use Excel or a BI tool.
Orders & Items per order (Average order size)
Segment by: Product categories, acquisition sources, customer cohorts.
Why it matters: Just like your revenues, you should always keep track of daily, weekly, monthly and yearly orders. Tracking average order size will let you know if and where you could improve it in order to increase revenue.
Data sources: In your shop database. Use Excel or a BI tool.
Return rate
Segment by: Product categories, acquisition sources, customer cohorts.
Why it matters: Return rates are important to watch because they are a pitfall for your profitability and could also point out where to improve suboptimal merchandising, product descriptions or pictures.
Data sources: In your shop’s database. Use Excel or a BI tool.
Benchmarks: Return rates vary greatly from industry to industry. The fashion industry has return rates as high as 50%. More than 33% of fashion online shops have a return rate higher than 20%. In Germany, the average return rate is 13%. Sources: Ibi Research institute, German trade and invest eCommerce report.
Average order value (AOV)
Segment by: Product categories, acquisition sources, customer cohorts, devices.
Why it matters: Average order value is a key metric because it shows the value generated by every transaction. Increasing it by up-selling or increasing basket size, is a way to increase revenues with the same purchase volume. Average order value is also useful to segment customers into different revenue groups.
Data sources: Data lies in your shop’s database. Use Excel for for calculations, or a BI tool.
Benchmark: AOV varies between countries, devices and industries, though it ranges between $90 – $160. source: Monetate eCommerce quarterly
Contribution margins
Segment by: Product categories.
Why it matters: Contribution margins are some of the most important sales KPI because they tell you how different categories or products contribute to the bottom line.
Data sources: Sales, Logistics & Payment data is your shop’s database, Marketing Spend in your Ad platform. Join them in Excel or use a BI tool
Benchmark: Varies greatly from industry to industry, product categories and countries.
Honorable mentions
Product Affinity
Product Relationship
Days and visits to purchase
4. Loyalty / CRM
The fourth process phase for a successful eCommerce site involves managing relationships with customers to retain them and sustain business.
Key goals: Understand your customers and their behavior in order to maintain a sustainable business, increase repeat purchase and customer loyalty.
New to returning customers & First and repeated orders
Segment by: Acquisition sources.
Why it matters: Informs you on which share of the business is coming from repeat customers, and the potential you have for converting new business into repeat one. Look at these numbers in context with absolute value, as the ratio might fluctuate both due to an increased customer acquisitions, or drop of repeat purchase.
Data sources: Your shop’s database. Import and calculate in Excel or use a BI tool.
Repeat purchase rates
Segment by: Customer cohorts.
Why it matters: Repeat Purchase Rates shed a light on how your repeat business develops, when and which of your customers stop buying, as well as the right time to communicate with them. They are particularly useful when segmented by customer cohorts, when used to spot trends over time and compare the effects of changes to the business on different groups of customers.
Data sources: Order data lies in your shops’s database. Calculations are fairly more complex than other metrics. Setting up a cohort analysis in Excel requires importing data from the database, creating cohorts identifiers, calculating life cycle stages, and using a pivot table with graph. A BI tool can automatically perform this type of analysis.
Churn rates (customers & revenues)
For subscription based businesses:
For businesses not based on subscription:
Segment by: Customer cohorts.
Why it matters: Knowing which of your customers have churned, or which ones might be about to, allows you to target them for reacquisition or retention campaigns. Retaining a customer is cheaper than acquiring one. Measure churn rates, both in terms of absolute customers lost and in terms of revenue lost.
Data sources: Data is in your shop’s database, the calculation for non subscription businesses involves deciding a time frame outside which, if no repeat purchase, the customer is considered churned.
Benchmark: For subscription based monthly churn should be less than 1%
Customer lifetime value (CLV)
Customer lifetime value is the net present value of the sum of future revenues expected from each customer. A simplified formula assuming constant margins and retention rates, and an infinite time horizon:
![CLV=margin\times[\frac{\text{Retention rate}}{(1+\text{Interest rate}-\text{Retention Rate}}]](https://s0.wp.com/latex.php?latex=+CLV%3Dmargin%5Ctimes%5B%5Cfrac%7B%5Ctext%7BRetention+rate%7D%7D%7B%281%2B%5Ctext%7BInterest+rate%7D-%5Ctext%7BRetention+Rate%7D%7D%5D&bg=ffffff&fg=000000&s=2 "CLV=margin\times[\frac{\text{Retention rate}}{(1+\text{Interest rate}-\text{Retention Rate}}]")
Segment by: Customer registration cohorts, acquisition sources.
Why it matters: Customer lifetime value represent the net present value of a single customer for the shop. CLV allows you to optimize marketing and CRM by giving you an upper limit for acquisition, reactivation and customer service expenditures. It is also very useful to identify high and low potential customer segments. Once spotted you can focus on giving high potential customers a special treatment and inquire further on which acquisition sources they came from and which products they purchased.
Data sources: Data lies in your shop’s database. A simple version CLV can be done in Excel, though since statistical modeling is involved, a BI tool will be more accurate.
Benchmark: Greatly varies according to industry and single customers.
Honorable mentions
Net promoter Score
Average customer service response time
To sum up
The following image summarizes what we have seen so far:
Of course, as every business is unique, the appropriate KPIs to track, or the filters you will want to apply, might slightly vary, so this selection is by no means to be considered one-size-fits-all. It is though a comprehensive one for every online retailer, and these KPIs will reveal themselves to be precious metrics to follow.
Wunderdata can help you monitor and analyze all of these KPIs and more, and provide you with visualization and filtering tools to make analysis and exploration a breeze. Check out a live demo!
Sources & useful reads
The framework used in this article has been adapted from the excellent work of Avinash Kaushik’s on web metrics
http://www.kaushik.net/avinash/best-web-metrics-kpis-small-medium-large-business/
An useful read on the topic of loyalty marketing, and the right metrics to focus on, is presented by Roman Kirsh. Roman is the founder of Lesara, where he uses Wunderdata to track these metrics.
http://romankirsch.com/blog/managing-loyalty-and-maximizing-clv-in-e-commerce-today.html
In the article, for the sake of simplicity, we have used an approximation of the Customer lifetime value formula. If you want to learn more about the concept of CLV, and how Wunderdata calculates it, we recommend this paper from Sunil Gupta & al.