A Business Technology Place

Defining an eCommerce Operation – Metrics Management

UPDATE 10/27/10 – I posted a mind map of my eCommerce Operation on mindmeister that replaces the original map contained in this post. This includes the latest updates to my organizational thoughts on an eCommerce team.

This is my fifth post on the defining elements of an eCommerce Operation. Previously, I’ve written about management in the areas of solution ownership, content management, product management, and demand management. In this post I’ll explore elements related to the metrics management of the group.
The term metric in this case refers to a measurable unit of an item that is relevant to the success of the eCommerce operation.  Typically, management will also specify a few key performance metrics or key performance indicators to provide focus on those with direct correlation to business success.  The measurements are intended to be a measure of success against some threshold. That threshold might be a value set by previous baseline or it could be a stretch goal to show improvement in a certain area. Within the eCommerce group I’ll call out the areas where its important to have metrics.

Web Analytics

Web site analytics provide information about how many people are visiting your site, what actions they take, information about their purchase, and other metrics about their identity. This area is important because it assists with determining if your merchandising tactics are working to close sales or provide customers with answers to service needs. Common metrics in this area include:
  • number of visitors
  • number of cart abandonments
  • number of session abandonments
  • number of purchases
  • geographic location
  • referral source (organic, link, keyword, etc.)
  • keywords or adwords used for entry

Customer Involvement

Metrics for customer involvement include traditional survey results but have expanded in recent years with the advancement of social media. Marketers now have the ability to measure customer involvement through other techniques such as Twitter, Facebook, blog comments, and customer reviews. Results in customer involvement can be both qualitative and quantitative. Common metrics in this area include:
  • number of customer reviews
  • average ratings on customer reviews
  • customer feedback on survey questions
  • number of customers participating in discussion groups or fan pages
  • reach of customers as followers


Metrics related to sales of your products and services are also part of metric management. Depending on your business, you may look to measure revenue, average cart purchase amount, product up-sell statistics, or retail value.  When measuring the financial results of the operation be sure to specify if the metric is taken from the web site analytics or the back office reporting. It’s important to note that the numbers may vary slightly based on the ability of customers to cancel, modify, or return their order after its placed on your web site.


Metrics related to channel are used to measure the effectiveness of the Internet channel compared to other channels such as phone, paper, or retail store front. In some businesses, it’s important to get as many Internet orders as possible if it provides the lowest cost to service. In other cases, it may be more desirable to have channels work in tandem to capture a sale in anyway possible. Create goals on channel usage and then measure the percentage of orders that come in by each channel. You should also keep the financial metrics split by channel.


Metrics related to products will involve the type, attributes, and quantity of products sold on an eCommerce site. Cross-sell and up-sell metrics belong in this group. Common metrics in this area include:
  • items per order (IPO)
  • number of SKUs sold
  • optional attributes purchased
I added these elements of the eCommerce organization to the concept diagram that I’ve grown with each post.  This updated image shows the eCommerce operation that I have discussed so far.  I have a few other areas to add so stay tuned. ( Select the diagram to expand it.)

eCommerce Operation with metrics management included, more areas pending (merchantstand.com)

What we can learn from carrier pigeons tweeting the score

This morning I was reading a few pages in the Historical Baseball Abstract by Bill James. The book is full of baseball history “stuff”. The interesting find today was a story about using carrier pigeons to deliver baseball scores across Philadelphia in 1883. The account reads that zookeeper Jim Murray sent carrier pigeons with scores back to his fellow workers at the zoo after each half inning. Can you imagine a dude showing up at the ticket gate with a cage full of pigeons? Apparently the idea caught on and other fans requested to get on the message distribution (the first twitter follower?).

Carrier, or homing, pigeons had long since been used to deliver messages.  Their use in times of war is documented on many occasions. However, the use at an athletic event to deliver scores had me thinking about how we send business messages with performance metrics today.

twitter bird to carry messages

twitter bird to carry messages

Today we have stock tickers, sports tickers, web sites with score, and in-game tweets for message delivery. These are used for global message broadcast and don’t require a subscription. They are fast and provide up-to-the second updates.

But how can we better deliver business messages with performance metrics inside our company message systems? Today most key performance indicators are put in a spreadsheet and emailed or crafted to fit in a Power Point presentation and reviewed in a meeting. Would decision makers would benefit from having performance metrics available to them on regular intervals like the carrier pigeons used for baseball game scores? Absolutely. IT staffs do this today for monitoring the health of the systems that support the business. If a metric falls out of an expected range then the on-call person is alerted via page and/or email.

To do this for business metrics would require that the message be simple and in a format easily understood by the recipient. The carrier pigeon exampled worked well because they followed that principal. This is also a primary reason that Twitter is successful today. It forces the sender to keep their message simple and on point. More detailed reports could be sent via detailed spreadsheets, discussed in committees, or reviewed online. The format and delivery of the message should be determined by the primary device used by the decision maker (mobile device, browser, email client) Here are a few ideas about how to use short messages to send key business metrics or KPIs.

  • Use your web analytics tool to auto-generate a report with specific performance metrics and email it to the decision makers. For example: hourly orders, daily conversion rates, hourly email opens for email campaigns, etc.
  • Set the browser home page for decision makers as an internal dashboard showing the latest performance metrics. This requires the use of some software for displaying the metrics. It could be a web analytics tool, data warehouse tool, or custom programmed interface.
  • If the decision maker uses a mobile device as their primary means for business messages then use the text message feature to send performance metrics. Most analytics and reporting tools have an automated email function and all messaging services have an email address that converts to a text message. Yes, you could just as easily send the message as an email.  I like the text message option because it forces the sender to keep the message short and on point.
  • Setup a protected twitter account to publish information and links. I like this idea because it gets decision makers onto a platform that they can use to read other messaging about their business and industry.

Do you have automated messages sending key performance indicators to your decision makers on regular intervals? What type of messaging frame work do you use? Have you thought of using tweeting business messages to internal recipients?