Overview

Consumer Clickstream Analysis is a consulting engagement that helps companies better understand how consumers use their services or web sites.

In this engagement, ConsumerEase consultants use sophisticated tools and analytics to learn consumer behavior patterns, navigational patterns, and feature usage patterns—patterns that can’t be obtained from normal reporting tools or other forms of transactional analytics. The results of the engagement include both a detailed analysis and an executive-level presentation. All results are presented in simple terms and graphical views, like the Consumer Behavior Bubble Chart displayed here. Key results are highlighted with actionable recommendations for improving performance with consumers.

This service is designed for large consumer product companies. To perform this engagement, the company must have a service that generates transactions as users interact with it. Usually this means analyzing web sites or web portals (web servers automatically store every user transaction). However, ConsumerEase can also work with other types of transactions, such as mobile Internet WAP servers, and may be able to work with your product’s data as well.

ConsumerEase performs the following steps in a Consumer Clickstream Analysis:

• A representative set of data is collected and loaded into the ConsumerEase data lab
• The data are converted into easily recognizable phrases, such as categories, products, and features
• The data are partitioned into user categories (e.g., new users, regular users, heavy users)
• Behavioral metrics are calculated (e.g., attraction, usage, stickiness)
• Navigational patterns, such as the paths users tend to take to get to different features, are detected
• A detailed report that analyzes the results is prepared
• An executive presentation that recommends the top ways to improve the performance is prepared
• The executive results are presented to employees and executives

Results

Most companies already receive daily or weekly reports from their web servers or IT department to show the day-to-day performance of its services and web sites. These reports are good for tracking daily or weekly performance, but often don’t provide enough information to make service design decisions. Instead, reporting tools tend to focus on counting service metrics, like how many users use a service or feature, or how many pages are viewed in a period of time. They don’t tell the paths users are using and when users are failing to move through the product successfully. Reporting tools also count URLs or IP addresses, which are computer terms, not easily recognizable feature terms (for example, marketers and executives have to know how to turn “www.myproduct/xyz-22/tfield.aspx?tsr=02X7B” into a meaningful feature name if they want to review the report). Finally, and most importantly, reporting tools tend to hide the big wins for future improvements, which are often the small features that greatly outperform their expected results for their location in the product. A hidden feature with a strong following of repeat users can be capitalized on or improved for other users, thus increasing the consumer base and, often, improving the revenue.

Companies also benefit when they can understand the paths consumers take within products. For example, it helps to know that 90% of users always abandon the product in the 3rd step of a 6-step process; or that the stickiest feature—the one with the highest percentage of users who visit and then return again—is 7 levels down, where only 5% of the users find it. These data points bring to light how to advance the service to the next performance level.

In the end, ConsumerEase delivers the information companies need to make good decisions. ConsumerEase doesn’t fix the services, but provides the facts on where the fixes should be focused. It is up to the company to decide which recommendations it wants to implement.

The results of the engagement are strongest when customers implement the recommendations. In prior projects we have seen 200, 350—even 600%—improvements in consumers usage. Most of the improvement comes from the new understanding of where “churn” and “stickiness” occur within a service. Companies reduce the locations where churn occurs and increase the visibility and quality of the features where stickiness occurs.

Beneficiaries

The key beneficiaries of Consumer Clickstream Analysis are the executives and marketing management responsible for a product’s consumer success. Although ConsumerEase works with the IT department to collect and translate the data, the real focus is on preparing results that help marketing departments make future product design decisions. Executives like the results for their ability to drive good investment decisions. Even engineers like the results, as they provide factual, actionable data.

Timeline

A full Consumer Clickstream Analysis on standard web-log data requires anywhere from 2 to 3 months to deliver. Most of the time is spent off-site, out of the way of the corporation’s day to day business. The corporation is only involved in the beginning (to help collect the original data) and at the end (to attend the final presentations). Below is a typical timeline:

ConsumerEase will perform the following steps in a Consumer Clickstream Analysis:

Week 1

Week 2

Week 3

Week 4

Week 5

Week 6

Week 7

Week 8

Week 9

Week 10

Week 11

Week 12

Sample data collection

Data transformations

Decipher & encrypt

Final data collection

Data processing

Analysis

Executive report and presentation

More time is required if the analysis is performed on a new data format from past projects or if custom analysis is required. The duration is often reduced in future projects because fewer data transformations are required to match the data to the product’s flow.

Pricing

The pricing of the service depends on several factors:

• Number of transactions
• Number of consumers
• Complexity of the product offering
• Data format of the transactions

A typical project involves analysis of 50 million transactions across 5 million consumers on a system with fewer than 50–100 features (and of a standard server data format for web or WAP). These projects typically are billed at a fixed price of US$150,000. If custom work is requested or if there are more product complexities, more transactions, or more consumers, the first analysis will run as high as US$300,000. Follow-on analysis tends to be less expensive if the system complexity remains constant, as past work can be leveraged.

Final pricing is determined once the engagement has been properly sized against the workloads. A proposal will be created only if ConsumerEase can accept the project.

Example results

Consumer Clickstream Analysis evaluates very private data for corporations, and therefore the results cannot be easily shared. However, most of the processes and details of how these analyses are done can be found in the book, Five Myths of Consumer Behavior, by Paul Allen Smethers and Alastair France, who have performed many of these analyses in the past.

Scheduling

A Consumer Clickstream Analysis usually requires 90 days’ advance notice to schedule time in the data lab for the project. Scheduling is done on a first-come, first-served basis timed at the signing of the proposal for the work (the appointment and travel will not be reserved for unsigned proposals).

Interested?

If you are interested in scheduling a project, please contact ConsumerEase with information about your product or service. We will work with you to define a project suited to your budget and needs.