How to plan tracking performance of e-commerce sites using Google Analytics
Translation of an article by Justin Courtoni. Original: http://cutroni.com/blog/2012/02/09/pimping-out-google-analytics-for-ecommerce-websites/
Once I was asked how to configure Google Analytics for an e-commerce site. And before diving into the explanations of the technical details, I wanted to talk about the basics of this process from a business point of view.
So this post is about how I plan to track electronic sales performance. Immediately make a reservation, e-commerce sites are different in size, some are more loaded, more advanced. This plan is good for medium sized sites. We will talk about more complex e-commerce dimensions, such as net profit and cohort, another time.
Analytical data on e-commerce can be divided into four categories, and for clarity, I created a small graph (everyone loves graphs):
Four types of data for analyzing e-commerce sites
For example, I will use the REI site . This is an e-commerce site combining a large number of online stores.
To increase sales, the site uses interesting and elegant tricks. I will also clarify the frequency of reporting to give an idea of how often I turn to certain data. Frankly, this is a huge amount of information, and it’s impossible to constantly monitor all this!
This section is completely devoted to data on how and from where clients get to the site. Marketers need to know which campaigns work and which go unnoticed. There are many different types of campaigns, some of which are aimed at disseminating brand information, while others urge consumers to take specific actions. And approaches to measuring campaign performance can vary.
Based on the foregoing, tracking the performance of each campaign is critical . It allows you to work with different types of segmentation, and we have the opportunity to correlate data on evaluating the effectiveness of campaigns with their goals.
By tracking the effectiveness of various campaigns, we can measure things like:
Reporting frequency: for marketers - daily. Less often - for the authorities.
Description of the process of user interaction with the site may not be an easy task. There are several key indicators, such as Bounce Rate, which are very easy to understand. But I want more - to measure the entire process of consumer interaction with the online store. And if you look at a site like REI.com, then make sure there are tons of different ways to interact.
These types of interactions are usually referred to as micro-conversions, since they do not generate revenue instantly, but they will possibly bring in the future. Most of these activities are related to strengthening relationships with potential customers.
Interaction metrics include:
Newsletter subscription and store search are also considered micro-covers
[Don’t worry, we’ll look at how to implement all this, another time.]
Frequency of reporting: to marketers - weekly. Less often - to the authorities.
So we got to the very core of the business! Deal! It focuses on it, right? Every employee wants to see a profit. Fortunately, we can do it!
But it is necessary to measure not just profit, but also many important metrics associated with it. And here is how I see them:
Reporting frequency: I review daily.
In addition to all the data related to the life cycle of the transaction itself, there is a lot of other information that can help us analyze business performance.
Reporting frequency: once every two weeks (for heavily loaded sites) or monthly.
Reporting frequency: depends on initiative.
Having studied the information that people searched on the site, you can get a deeper idea of the nature of user behavior, their desires and motives.
The keywords that users drive into the site search provide valuable information about what customers think about your products, for example, the terms with which they describe your products. If the search on the site has led to nothing, it means that there is no demanded product in the product leneka, or your search engine is not working. In any case, interesting information.
Reporting frequency: twice a week (for heavily loaded sites) or monthly.
This is what my overall plan looks like. The next article will provide more tactical explanations on how to keep track of all these things.
Once I was asked how to configure Google Analytics for an e-commerce site. And before diving into the explanations of the technical details, I wanted to talk about the basics of this process from a business point of view.
So this post is about how I plan to track electronic sales performance. Immediately make a reservation, e-commerce sites are different in size, some are more loaded, more advanced. This plan is good for medium sized sites. We will talk about more complex e-commerce dimensions, such as net profit and cohort, another time.
Analytical data on e-commerce can be divided into four categories, and for clarity, I created a small graph (everyone loves graphs):
Four types of data for analyzing e-commerce sites
- Attraction: information about received traffic,
- Interaction: information about what people do on the site,
- Conversion: liquidity information (profit, profit, profit ...),
- General information: other universal data that we must check every now and then.
For example, I will use the REI site . This is an e-commerce site combining a large number of online stores.
To increase sales, the site uses interesting and elegant tricks. I will also clarify the frequency of reporting to give an idea of how often I turn to certain data. Frankly, this is a huge amount of information, and it’s impossible to constantly monitor all this!
Attraction
This section is completely devoted to data on how and from where clients get to the site. Marketers need to know which campaigns work and which go unnoticed. There are many different types of campaigns, some of which are aimed at disseminating brand information, while others urge consumers to take specific actions. And approaches to measuring campaign performance can vary.
Based on the foregoing, tracking the performance of each campaign is critical . It allows you to work with different types of segmentation, and we have the opportunity to correlate data on evaluating the effectiveness of campaigns with their goals.
By tracking the effectiveness of various campaigns, we can measure things like:
- most revenue generating marketing channels
- the creative and message that resonates most with consumers
- marketing activities that are most effective for attracting customers at the initial stage of purchase
- time of day when one or another campaign is most successful
Reporting frequency: for marketers - daily. Less often - for the authorities.
Interaction
Description of the process of user interaction with the site may not be an easy task. There are several key indicators, such as Bounce Rate, which are very easy to understand. But I want more - to measure the entire process of consumer interaction with the online store. And if you look at a site like REI.com, then make sure there are tons of different ways to interact.
These types of interactions are usually referred to as micro-conversions, since they do not generate revenue instantly, but they will possibly bring in the future. Most of these activities are related to strengthening relationships with potential customers.
Interaction metrics include:
- Bounce Rate: We've been talking about it for years! The Bounce Rate determines the percentage of visitors who viewed only one page and immediately left the site. This metric is not particularly useful on a site scale, but is extremely curious as part of an advertising campaign.
- Newsletter: E-mail newsletters are still very important! Getting someone to sign up for the newsletter is awesome.
- Nearest store search form: Very good metric. Many sites provide users with the opportunity to find the nearest store, have a look and personally examine the goods. Measuring the number of times a shopping search button has been activated will not say anything about sales, but will produce interesting data. Especially if you analyze which points of sales you searched most often.
Newsletter subscription and store search are also considered micro-covers
- RSS Subscription: I love subscribers. Although we do not know for sure whether they are reading what we are sending, in any case, the subscriber is an interested person.
- Cart: This is an important part of the buying process. Until you add to the basket, you can’t buy! And since this is such an important stage in the buying process, we will measure it.
- Wish: This is a kind of predictor of the future! How many people add goods to the wish lists, and how many add to the basket? And how many people actually make a purchase? It is important to know the answers to all these questions. When we learn how to measure this, we will actually get a forecast of future income from deferred goods.
- Blocks of social networks on the site: (tweets, likes, +1): At least someone clicks on these buttons? And what do we know about those who do this? This group is an important segment of consumers.
- Product Information: Many electronic sales sites have a ton of information that encourages consumers to make a purchase. This information is usually divided by various tags, as on the REI website. Tags also need to be tracked, this will provide an opportunity to collect data as searched by tags.
- Product Rating: This is another great way to interact with customers. Interaction = acquaintance = future income. Reviews can be not only a great source of traffic (hello to free SEO), they will also give a deeper understanding of what your customers like and what not.
- Product Video: I LOVE the product video. There is no better way to feel the product, unless you hold it in your hands. But this is not possible online. We need to be able to measure video! After all, not all videos are the same? If not, what videos and how do consumers interact? So if I have a site like REI, then I want to track the video.
- Attitude of views to orders: I learned about this metric from Brian Eisenberg at the Google + party, and it is fantastically cool! Brian recommends measuring how many people view a product or category but don’t buy. Think about how important this is. If people look at a product but don’t buy, there is some kind of hindrance. Just eliminate it!
[Don’t worry, we’ll look at how to implement all this, another time.]
Frequency of reporting: to marketers - weekly. Less often - to the authorities.
Conversions
So we got to the very core of the business! Deal! It focuses on it, right? Every employee wants to see a profit. Fortunately, we can do it!
But it is necessary to measure not just profit, but also many important metrics associated with it. And here is how I see them:
- Revenues: Well, what is there to explain! You sell and calculate your profit!
- Return on investment: ROI is a great metric, it helps us understand how much money we earned on what we invested in. Cool, yeah? But not always Google Analytics makes it possible to track ROI. With it, you can only measure the profitability of AdWords. Why is that? Because Google Analytics does not have data on your investments. That is, if you spend $ 1,500 on a mailing list / marketing campaign, Google Analytics does not know anything about it. But this should not be an obstacle to assessing the profitability of your investment.
- Average order value: what is the average check for a purchase on a site? We often try to encourage customers to spend more money on one transaction, using, for example, cross-selling. We can also use the average order value indicator to identify the most cost-effective channels for attracting visitors, for example, electronic mailing, search or social networks.
- Profit from regular customers: Technically, this is segmentation, but I decided to add this metric to this list. Almost every business wants to have regular customers. Why? The incentive to constantly make a purchase for a customer requires much less effort than it takes to attract a new one. It is very important to segment both groups and study their behavior.
- Profit per visit or value per visit: The value of each visit is an excellent metric, because it makes it possible to compare the cost of traffic from various sources. And this is a great way to compare the return on investment in various traffic sources.
- Efficiency of internal campaigns: many enterprises promote certain products directly on their website. For example, they can hang on the main page an advertising banner with information on the elimination of seasonal goods. The ability to segment profits and calculate the effectiveness of such campaigns is extremely useful.
- Visitors Intention and Satisfaction: Here we have data from the category of "quality". Are the needs of those who visited your site satisfied? Did they log in successfully? And what's more, it’s important for what purpose they came in? It is impossible to get the right answer based on only qualitative data. Quantitative data, usually from reviews, can help you.
Reporting frequency: I review daily.
General Metrics
In addition to all the data related to the life cycle of the transaction itself, there is a lot of other information that can help us analyze business performance.
- Time to purchase: It takes a week or maybe a month to bring the consumer to the purchase? It is very important to understand this aspect in order to customize the marketing campaign for your customers.
- Visiting a website before purchasing: How many times on average does a visitor visit a website before making a purchase?
- Valuable customer behavior: How do valuable customers behave? By customer prices, I mean people who spend more than a certain amount of money. How do they find a site (where does geography come from, and with what marketing campaign)? Almost everything that we talked about so far is directly related to daily business management. But there is a whole world of metrics that help us evaluate the technical aspects of the site.
- The quality of the site: If the site is not fast enough, no one will linger on it. Seriously. Website performance metrics also affect revenue and should be monitored, although not every day.
- What is the average page page load speed?
- In particular, which page loads the longest? How does this affect Bounce Rate and earnings?
- What are the most common site errors (404, etc.)?
- What are the technical parameters of the devices that visitors use (operating system, screen resolution) and how this affects the business.
Reporting frequency: once every two weeks (for heavily loaded sites) or monthly.
- Mobile phone applications: device information also refers to site performance data, but there are still some differences. I usually consider mobile devices as a separate category, since the mobile sphere is developing very quickly. Not only do you need to check the operation of the website on a specific device, you also need to think about usability when using applications or a mobile site.
- Which devices are the most popular?
- Which versions of the device are the most popular?
- What media are more commonly used?
Reporting frequency: depends on initiative.
- Site Search: I have long exalted the importance of site search, it's just an amazing thing. Again, this is not the data set that I work with every day, but that I look at it once or twice a week or monthly.
Having studied the information that people searched on the site, you can get a deeper idea of the nature of user behavior, their desires and motives.
The keywords that users drive into the site search provide valuable information about what customers think about your products, for example, the terms with which they describe your products. If the search on the site has led to nothing, it means that there is no demanded product in the product leneka, or your search engine is not working. In any case, interesting information.
Reporting frequency: twice a week (for heavily loaded sites) or monthly.
This is what my overall plan looks like. The next article will provide more tactical explanations on how to keep track of all these things.