Web Analytics & Metrics

Convert more visitors into customers, subscribers, or members

Over the past several years, the Web has matured as a channel for many businesses and is now a major revenue stream for most Fortune 1000 companies. These days, there is a call for the Web channel to be held accountable just as much as other, more traditional channels. Whether used for e-commerce, sales lead generation, or for product support, companies invest a significant amount of resources in their Web presence. Often, hundreds of Web sites compete against each other for the same traffic. But, how do corporations determine whether their Web presence is effectively meeting their business goals? How are they pinpointing what the potential problems are and where they those problems actually reside? How can they monitor the effectiveness of the solutions that they put into place?

Web analytics is the answer. Web analytics is the collection, measurement, analysis and reporting of Internet data for purposes of understanding and optimizing web site usage. In today’s competitive online environment, Web Analytics can help businesses convert more Web visitors into customers, subscribers, or members.

Clickstream

Not long ago, the main source of Web traffic data was clickstream (the recorded paths that visitors take through a Web site). Even now, many people feel that Web analytics = clickstream. However, clickstream is just a small piece of a much larger Web intelligence picture. Unfortunately, clickstream is still the major source of data used for Web decision-making by a significant number of professionals. Therefore, many companies are confounded by the lack of real, actionable insights that clickstream offers.

Key Performance Indicators (KPI’s) - Beyond Clickstream

Modern practitioners have adopted the concept of Key Performance Indicators (KPI’s). These enable professionals to look beyond the raw clickstream data, and illustrate how well a site is performing against the real-world business goals that have been set. Each company has different business models, and different companies will likely have a unique mix of KPI’s. Below are some common KPI’s that a typical e-commerce Web site might be most concerned about, and their definitions in terms of an e-commerce operation:

Conversion Rate

Conversion Rate is the percentage of visitors who complete a transaction, based on the marketer’s intended action and is calculated as:

(Number of Sales / Number of Visitors) X 100 = Conversion Rate

Conversion Rate is an important KPI in that it enables the enterprise to keep a pulse on the Web site’s overall ability to turn visitors into paying customers.  A rising Conversion Rate demonstrates that sales efforts are improving overall, although this KPI cannot indicate what specific strategies have worked.  If this KPI value is falling, it signals that action is needed, although it cannot diagnose what the specific problem might be.

Conversion Rate for a Specific Campaign

Similar to “Conversion Rate”, this is the percentage of visitors who complete a transaction for a specific marketing campaign. Conversion rates should not only be monitored against benchmarks monthly, but also before, during, and after campaigns, special seasons, etc. Landing pages are often used to make it easier to measure the response to a specific campaign.

Average Cost Per Conversion

The Average Cost Per Conversion describes the cost of acquiring a customer. In terms of e-commerce, the “conversion” refers to an actual sale. Here is the formula:

Sum of Acquisition Marketing Costs / Conversions = Average Cost per Conversion

Should a campaign’s Average Cost per Conversion rise markedly, it likely indicates that a costly program has been launched, and that it is failing to drive an appropriate number of conversion outcomes. Such a program should be monitored closely, and well thought out measures should be undertaken to remedy the problem if no improvement is seen.

Average Order Value

The Average Order Value or AOV is the average amount spent for a single checkout purchase on a retail site for a particular customer segment or group. It is calculated as follows:

Sum of Revenue Generated / Number of Orders Taken = Average Order Value

Working to increase the Average Order Value is an effective means of increasing revenues. However, it’s crucial to monitor this particular KPI in conjunction with the Conversion Rate. For example, consider a situation wherein most customers buy high-ticket items in small numbers. The AOV will be high, as most orders will be of a large monetary value. Outside of the context of Conversion Rate, these would appear to be good numbers, however the overall revenue may actually be down. This can happen when each customer has spent a lot of money, but there are not many actual customers. Also note that when the AOV falls, but the Order Conversion Rate rises, the revenue per visitor should stay roughly the same. However, when both AOV and Conversion Rate shrink, revenue per visitor will likely be strongly impacted in a negative fashion.

Wish lists and recently viewed product lists that are saved across visits (even for unregistered customers) should be employed to remind shoppers about items of interest, and make it easy to buy when they are ready. Enticing every customer to spend more with every purchase is an essential part of maximizing revenue.

Average Items Per Cart Completed

This is the measurement of the number of units or items in each successfully completed cart. Calculated as:

Sum of Products Purchased / Number of Completed Shopping Carts = Average Items per Cart

This KPI is a good indicator of the success of up-sell and cross-selling efforts. However, this metric may decrease upon the implementation of a successful promotion of a single item, thereby increasing the number of single-item orders. If no recent efforts have been made to improve up-sell and cross-selling opportunities, an increase in this KPI might show that better-qualified traffic has been attracted to the site, or that a very successful campaign has been established.

Shopping Cart Abandonment Rate

This is the percentage of visitors that start an order by initiating the check-out process, but depart before the purchase is completed. This KPI is calculated as:

(Number of transactions completed / Number of transactions started) X 100

Regardless of industry, e-commerce Web sites experience some rate of shopping cart abandonment. If this KPI begins to rise, it is time to ensure that the checkout process is as streamlined as possible. For example, it has been shown that streamlining the checkout process by removing even one Web page from the purchase routine lowers the abandonment rate by a significant margin. Another common problem is that some retailers do not disclose all additional shipping rates or handling fees on their products up front, only to disclose them later during the checkout. Additionally, it is best to offer as many payment options as possible. Enterprises should avoid placing any extra marketing content in the shopping cart area that might distract purchasers. Anything that retailers can do to make it quicker and easier for customers to make an informed decision to buy and check out quickly will lower their Shopping Cart Abandonment Rate KPI.

Key Insights Analysis

Both clickstream and KPI’s can offer a good grasp of the quantitative data (which indicates ‘What’ has happened). To become even more competitive, a new world of qualitative behavior analysis has arisen: the world of actionable Web insights. By listening to their customers, businesses can now begin to answer the questions of ‘Why” their customers do the things that they do by using Key Insights Analysis (KIA). The following are examples of KIA’s:

Click Density Analysis

Virtual “Heat Map”, and “Site Overlay/Click Density” overlays show where on the Web page visitors have actually clicked during the chosen time period. Using colors, heat maps clearly show the grouping of clicks on the page and their concentration (brighter colors indicate more clicks on a particular link or hot spot). In the case of Site Overlay/Click Density, it is also quite easy to segment the data in order to separate where the traffic actually came from. This gives some insight into what each segment was actually looking for on the site. By better targeting each segment, marketers can begin to maximize each visitor’s value based on the way that they arrived at the site, as different traffic sources often have different motivations, behaviors, and buying characteristics.

Task Completion Rates

While it is easy to measure success on an e-commerce site where it’s clear whether an actual transaction was completed or not, it’s not so simple in other cases. In the case of a support section of a Web site for example, it is important to have some kind of user feedback to gauge whether the user has found what they were looking for, or solved their problem. Surveys that ask, “Was this article helpful” with a box to click “Yes” or “No”, or an opportunity to rate the content from “1” to “10” are simple, but effective ways to gather qualitative user data. There is less guessing based on old-style clickstream page views. More sophisticated surveys can be conducted, or even lab usability studies can be performed (third-party companies can provide such services also).

Visitor Primary Purpose

Visitor primary purpose measures why people come to a Web site, not just how. Few professionals understand all of the reasons why people seek out their Web sites. Naturally, businesses cannot infer this from pages viewed, as visitors may arrive at the site seeking content that the business doesn’t have. A simple method to gain insight into the primary purpose of a visit is to observe the keywords that visitors use to find the Web site, and then attempt to determine whether the visitor succeeded in their goal. However, conducting online surveys and/or phone interviews are often the most effective way to make such a determination.

There are several reasons people may visit a Web site. They may be there to actually buy a product, or research a product. However, they may also arrive at the site to find company information, register a product, return an item under warranty, look for a job, and more. Generally for a retail site, 15 percent of visitors are there to buy, and another 20 percent are there to research a product. This combined 35 percent of all visitors are the ones that you can expect to be able to actually convert into customers.

Segmented Visitor Trends

In the online world, “Visitor Segmentation” has been a term used to refer to filtering tools used by online professionals to categorize the behavior of visitors, or groups of visitors meeting a selected set of behavioral parameters. While this is a kind of “segmentation”, it falls short of what many marketers think of as traditional segmentation, which is a complete classification of customers and/or prospects derived from a combination of demographic and psychographic information based on primary research.

Traffic data is easy to collect, but it is difficult to map visitors whom we often know little about (except their online behavior) to traditional marketing segmentations. Most Web analytic tools do not provide the required functionality to build Web data driven segmentations. In addition, a great deal of the data transformation and aggregation must be performed before segment building is performed. Therefore, the segmentations must be created outside of the web analytic tool and then imported it into a true analysis application.

Multichannel Impact Analysis

Because we live in a multichannel world, it’s essential to recognize that customers are everywhere: online, mobile, email, TV, print, billboards, and more. As such, it is important to measure the impact of such channels on Web traffic, as some (or even a large amount) of the company’s traffic may be visiting as a result of offline activities. Conversely, it’s also critical to note that Web marketing also drives sales in other more traditional channels.

Vanity URL’s

For offline advertising, it is smart practice for businesses to employ a “Vanity URL”. This is a domain name that is easy for a prospect to remember, and is creatively linked to the branding, product or service. The company in question would use the URL in media such as TV, print, etc. Customers that type in the URL are redirected to the proper Web site, with a tracking code attached. The online response to the TV, print ad, etc, can now be measured.

Making the redirect “permanent” will ensure that the referrer’s data will be sent to the Web site. This enables the “online referrals” (referrals from a link posted on the Web) to be separated from the “offline referrals” (those referrals generated by the offline advertising). This can be done because offline referrals will indicate a “blank referrer”.

Redeemable coupons/offer codes

Many merchants are using unique redeemable coupons/offer codes that are very effective for offline advertising mediums such as TV, magazines, radio, catalogs, direct mail, and many more. The offline advertising carries a unique “offer” or “value” code for a unique promotion. Some corporations use codes that are easy to remember, but this is not always the case. The customer that places an order online would then type in the unique code to receive the promotional offer (often a discounted price, a bonus item, or even free shipping). This enables such visits to be tracked as being motivated from a specific offline advertising medium/campaign.

Online surveys

Surveys can be used to flesh out the impact of the various channels in your marketing mix. Surveys are best kept as simple as possible in order to obtain a good response. An effective survey might ask two simple questions such as, “What best describes the motivation of your visit to the Web site”, and “What is the likelihood that you’ll make a purchase today?” The first question could be answered from a drop-down list (choices might be TV, print ad, billboard, etc), and the other could be acknowledged by clicking a “radio button” (“1-5” with 1 being “low” and 5 being “high” likelihood of a purchase).

Similarly, the question of “What best describes the motivation of your visit to the Web site” could be asked at the time checkout on a retail e-commerce site. It’s a short question and unlikely to interrupt the checkout process. The benefits are that the actual marketing channel that brought the customers to the site can be correlated to where the customers reside, and how much each customer spent.

Conclusion

Clearly, Web Analytics is key to creating, growing, and maintaining online revenue, as businesses now have a greatly enhanced ability to actually listen to customers. Without Web Analytics, corporations are only able to guess at what’s working and what’s not on their Web sites. With the very good free applications available, powerful Web Analytics functionality is now a reality for more enterprises than ever before.

Copyright 2009 Shawn Kanciruk
Updated May, 2009