Creating an Ecommerce Tracking Plan for Google Analytics

Someone asked me how I would set up Google Analytics for an ecommerce website. And, before I get into all of the setup details, I wanted to lay the foundation from a business perspective. I’ll describe the Google Analytics features and how to configure them in another post.

So this post is all about my measurement plan. A quick caveat: there are lots of different sized ecommerce businesses, some more complex and advanced than others. This plan will work good for a mid-size business. We’ll get into more complex ecommerce measurement like lifetime value and cohorts some other time.

I’ve divided the ecommerce data into four categories, and created a small graphic (because everyone likes graphics):

The four types of ecommerce data.

The four types of ecommerce data.

* Acquisition Data: Information associated with getting traffic.
* Engagement Data: Info about how people interact with the website.
* Conversion Data: Info about business success (revenue, revenue, revenue…..)!
* Foundational Data: Un-sexy things that we should check every now and then.

The REI Ecommerce Website

The REI Ecommerce Website

And to make things even easier I’ll use REI as an example. They’re an ecommerce site with a substantial number of brick-and-mortar stores.

They use some neat selling features on their site. I’ve also included a reporting frequency to help describe how often I look at this data. Let’s face it, there’s a LOT of data and we can’t look at all of it all of the time!

Acquisition Data

Acquisition data is all about how we get customers to an ecommerce website. Marketers need to know which campaigns are working and which are not. There are lots of different types of campaigns, some are focused on spreading our brand while others are direct-response activities. How we measure these activities can differ.

With that said, Campaign Tracking is critical. This allows many types of segmentation so we can align our campaign measurement with the campaign objectives.

With a direct conversion campaign we might measure things like:

* Which marketing channels generate the most revenue?
* Which creative and messaging are most popular with consumers?
* Which marketing activities are effective at reaching people early in the buying process?
* Which marketing activities are effective at reaching people late in the buying process?
* What time of day are certain campaigns successful?

Reporting frequency: Daily for a tactical marketer. Less frequent for senior staff.

Engagement Data

Engagement can be a tricky thing to describe. There are some basic metrics, like Bounce Rate, that are very easy to understand. But I want to measure, at a deeper level, how people interact with an ecommerce site. And when you look at a site like there are lots of different ways to interact.

These interactions are commonly referred to as micro-conversions and, while they do not generate revenue immediately, they can lead to revenue in the future. Most of these activities are related to building a relationship with prospective customers.

Engagement metrics include:

* Bounce Rate: We’ve been talking about this one for years! Bounce rate measures the percentage of single-page visits. While it’s more-or-less useless at a site-wide level, it’s very useful when segmented by marketing campaign or channel.

* Newsletter signups: Email is still very important! Getting someone to sign up for a newsletter is HUGE.

* Store finder: This is a good one. Many sites have a store finder just in case you want to visit and physically examine an item. While measuring the number of times the store finder is used is not a direct indication of sales, it does yield interesting data. Especially when you collect which locations are most commonly searched for.

Micro Conversion for an Ecommerce website

Subscribing to the newsletter or using the store finder are micro conversions.

* RSS subscriptions: I love subscribers. While we can’t control if they’ll actually read what you put out there, a subscriber is still someone interested.

* Add to Carts: This is an important part of the buying process. If people do not add items to their cart then how will they purchase them? It’s a critical step in the buying process, so we’re going to measure it.

* Save to Wish List: This is like a predictor of the future! How many people save an item to a wish list or save their cart? And then how many people actually complete the transaction later? These are both important things to know. If we can measure both we can actually get a sense for future revenue from saved carts.

* On-site social interactions: (Tweets, Likes, +1): Are people clicking these things? And is there something special about the people that do? Segmenting by this group is important.

More ecommerce micro-conversions for an ecommerce website.

More ecommerce micro-conversions for an ecommerce website.

* Product Information Interactions: Many ecommerce sites have TONS of content to help convince consumers to buy. This information is usually separated into separate tags, like here, on the REI site. These tabs need to be tracked so we can segment data based on those that use the tabs.

* Product Ratings: This is another great way that ecommerce companies can generate interaction with customers. Interaction = engagement = future revenue. Reviews can be a great source of traffic (hello, free SEO) and insight into what your customers like and do not like.

* Product Video: I LOVE product videos. There is no better way to get a sense for a product. Unless you can pick it up and hold it. But that’s impossible on the web. But we need to measure video! Do the videos make a difference? If so, which type and how do people interact with the video? So if I’m a company like REI, I want to track product videos.

* Look to Book: This is a metrics that I learned from Bryan Eisenberg during a Google+ Hangout, and it is FANTASTIC. Bryan recommends measuring how many people look at a product or category and then do not buy that product. Think of how useful this is. If people are looking at a product, but not buying it, there is some friction there. Remove the friction.

[ Don’t worry, we’ll cover the implementation in another post. ]

Reporting Frequency: Weekly for a tactical marketer. Less frequent for senior staff.

Conversion Data

Now we get to the heart of the business! Transactions! That’s the main focus, right? Everyone, from the top of the organization to the bottom, will want to see revenue. Luckily we can do that.

But it’s not just revenue, there are many revenue-related metrics that are important to an ecommerce business. Here are the ones that I put together:

* Revenue: Not much to explain here! You sell stuff, measure how much money you make!

* Return on Investment: ROI is a great metric, it helps us understand how much money we make based on how much money we spend. Cool, right? But Google Analytics is limited in it’s ability to track true ROI. IT can only track the ROI of AdWords. Why? Because it does not have any of the investment data. So if you spend $1500 on an email marketing campaign, that data is not in Google Analytics. Regardless, you should be thinking about ROI.

* Average Order Value: On average, what do people spend per transaction? We often try to get our customers to spend more per transaction and employ various techniques, like cross-selling, to increase AOV. We can also use AOV to identify high-value channels, like email, search and social.

* Revenue by Repeat Customers: Technically this is a segment, but I’m putting it here. Almost every business wants repeat customers. Why? It takes less effort to attract a repeat customer versus a first-time customer. It’ important to segment these two groups of customers and study their behavior.

* Revenue Per Visit or Per Visits Value: Per visit value is a great metric because it normalizes the value of traffic from different sources. It creates a good way let’s me compare the performance of different traffic sources.

* Internal Campaign Performance: Many businesses will run some type of promotion or campaign directly on their site. For example, they might put a banner on the homepage to liquidate seasonal merchandise. The ability to segment revenue and measure the effectiveness of these campaigns is performance.

* Visitor satisfaction and intent: Here we have some qualitative data. Are the people coming to the site happy with their experience? Is their visit a success? Even more important, why are they coming to the site? We can’t get a good answer using qualitative data. Quantitative data, usually from a survey, is the way to go here.

Reporting Frequency: I look at most of these daily.

“Foundational” Metrics
In addition to all of the data related to the purchase life-cycle, there are many other pieces of information that can help us understand the business performance.

* Time Before Purchase: Does it take one week to sell a customer a product or one month? It’s really important to understand this behavior so you can tailor your marketing campaigns to your customers.

* Website Visits before Purchase: How many times do we need to interact with the person (on the website) to drive a transaction?

* High-value Customer Behavior: What do my high-value customers do? By high-value I mean people that spend more than a certain amount of money. Where do the come from (i.e. geo-graphic location and marketing campaign) Almost everything we talked about so far has to do with running the business day to day. But there is an entire world of metrics that help us understand the technical health of our website.

* Site Performance Metrics: If your website is not fast then people will not use it. Seriously. While site performance metrics are not as sexy as some of the above, they’re still important. These are things that can impact revenue and should be monitored, just not every day.

* What is the average page load time for the site?
* Specifically, which pages take a long time to load? How does this impact bounce rate and revenue?
* What are the most common error pages on the site (404’s, etc.)?
* What is the technical profile for visitors (operating system, screen resolution) and how does this impact the business

Reporting frequency: Bi-weekly (for heavily trafficked sites) or monthly.

* Mobile App/Website Data: Related to site performance, but slightly different is device data. I usually put mobile devices into a separate category as the mobile space is evolving very quickly. Not only do we need to consider how the website works on a specific device, but potentially usability information about how people might use an app or the mobile site.

* Which devices are most popular?
* Which versions of the device are most popular?
* What carriers are people using?

Reporting frequency: depends on initiatives

* Site Search: I’ve been long extolling the value of site search, it’s just amazing data. Again, this isn’t a data set that I’m working with every day, but something that I look at weekly, bi-weekly or monthly.

A site search box

You can gain a deep understanding of user behavior using Site Search data.

The search terms used in site search provide valuable insight into how customers think about products specifically the keywords they use to describe your products. Site searches that return zero results usually indicate a missing product, or that your site search is broken. Either way, interesting data.

Reporting frequency: Bi-weekly (for heavily trafficked sites) or monthly.

So that’s my general plan. In the next post we’ll get more tactical and talk about how to exactly track all of this stuff.

Did I miss anything?

I’d love to hear how others are tracking ecommerce.

Be Sociable, Share!

    Like this post? Sign up to get posts delivered to your inbox.


    1. says

      It’s really interesting to see how you’d have tackled this, Justin. I think I already use most of these, apart from the “look to book” metric which I’ll need to give a try.

      I’ve added quite a few hacks to my sites to record other things including:

      * Telephone revenue
      * Voucher codes used
      * Service recovery vouchers issued
      * Checkout method (guest or registered)
      * Shipping method
      * Payment method (card, Paypal)
      * Shipping contribution
      * Currency and exchange rate
      * Item profit and margin
      * Basket profit and margin
      * Purchase latency
      * Total purchases
      * Total lifetime spend
      * Annual, monthly and quarterly cohorts
      * Newsletter preferences
      * Cost price
      * Discount given
      * Complaints received
      * Product returns data
      * Contact type (ie. sales query, order tracking)
      * Errors from form validation
      * Out of stock items
      * Errors from payment gateway
      * Category and product viewed
      * Promotions viewed
      * Products and SKUs added

      And probably quite a few more…

      • says

        @Matt – Thanks for the great comment! Many of those metrics that you list are perfect! One thing I try to do is balance the amount of implementation with the business needs. For organizations that have resources and time, I think your list is a must. But smaller organizations need to prioritize. But, hopefully, after a while, they’ll be able to track everything we have listed.

        Thanks again!

    2. says

      Not much left to say here – awesome article Justin, and with the list from Matt as well I may repeat something of what you wrote, but hopefully with a different “touch”.

      I track Add to Cart on a product level, and if there are different Add to Cart buttons I track them as well (Quick Buy Button vs. Ordinary Button). It has surpriced me on some sites how many who Add to Cart from a category page or from related products without visiting the product. I also track the price of the product added to the cart so I can analyze for example price sensitivity.

      Products removed from shopping cart is also something I track if I can.

      Since it sometimes can be difficult to track Errors from form validation, in these occasions I track Form field changes since this can be done with a simple jQuery/javascript using Events (it’s better than nothing). If Unique Event numbers vs. Total Event numbers for a form field are very different, it can be a result of a problem with that particular field.

      And a final thing that you have covered, but with a little “twist” from me about why it can be useful.
      How long time does take from a product is added to cart until it’s bought? Too many shopping carts expires to early.

    3. says

      It’s a great post that i’m just looking forward to.

      As an SEMer for an e-commerce website, I found there’re so many things need to be tracked in order to improve ROI and stand out in the competition.

      However,things get complicated when potential clients coming from different channels buy in the long term. It’s hard to analyze the ROI of the individual funnels. We’re using the custom variables as well as e-commerce tracking to solve this problem, also using the search funnel in adwords and multi funnel in ga, but i think there’are still some limitations, even this is already a challenge to get actionable data.

      So, i’d like to know more about how to solve the problem of multi channel & long term client tracking and getting most from these kind of data.

      Thanks in advance,sincerely.

      • says

        @xiaoq – Thank you for the comment. You’re right, there are some limitations. And there are lots of reports and it can get very confusing. One major drawback of GA is that it will only track the online channel. So true multi-channel measurement can not happen. And until GA can track offline transactions we’ll be stuck.

    4. says


      Need your help to fix my e-commerce tracking code… Please email me if you have the solution/suggestion.

      Today I added e-commerce tracking in my analytics code on the order receipt page. Everything looks correct to me but it is not tracking the data.
      However, wehen i remove the code, regenerate it again and put it again, it tracks just once and then in subsequent tries it does not track anything. All other pages are being tracked correctly where the ecommerce code is not present.

      Please see the code below and if anyone can let me know what am i doing wrong, i will be greatful.

      var _gaq = _gaq || [];
      _gaq.push([‘_setAccount’, ‘UA-24606731-1′]);
      _gaq.push([‘_setSiteSpeedSampleRate’, 10]);
      _gaq.push([‘_setCustomVar’,2, ‘PaymentType’,’Cash on delivery (COD)’,3]);
      ‘419’, // order ID – required
      ‘’, // affiliation or store name
      ‘1232’, // total – required
      ‘0’, // tax
      ’50’, // shipping
      ‘New Delhi’, // city
      ‘Delhi’, // state or province
      ‘India’ // country
      // add item might be called for every item in the shopping cart
      // where your ecommerce engine loops through each item in the cart and
      // prints out _addItem for each
      ‘419’, // order ID – required
      ‘2413’, // SKU/code – required
      ‘English Learner’, // product name
      ‘Educational Laptops’, // category or variation
      ‘799’, // unit price – required
      ‘1’ // quantity – required
      ‘419’, // order ID – required
      ‘428’, // SKU/code – required
      ‘Monkey Expressions’, // product name
      ‘Learning’, // category or variation
      ‘332’, // unit price – required
      ‘2’ // quantity – required
      _gaq.push([‘_trackTrans’]); //submits transaction to the Analytics servers
      (function() {
      var ga = document.createElement(‘script’); ga.type = ‘text/javascript'; ga.async = true;
      ga.src = (‘https:’ == document.location.protocol ? ‘https://ssl’ : ‘http://www’) + ‘';
      var s = document.getElementsByTagName(‘script’)[0]; s.parentNode.insertBefore(ga, s);


    Leave a Reply

    Your email address will not be published. Required fields are marked *

    You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>