Actionable Web Analytics Advice
Clients who are using Web Analytics data to trigger email messages are seeing amazing results. So why aren't more marketers jumping on this opportunity?
. Is the data overwhelming?. Are there not enough resources to dedicate to the project?. Are marketers not sure where to start?
Visualizing the OpportunityI admit web analytics event data can be overwhelming if you look at all of the different data sources. But there are a few basic activities (or events) that I think create excellent opportunities for all marketers such as:. Abandoned activity (like shopping carts or web applications) . Browse activity (a customer searches web site content, but does not convert). Non-responder follow-up (email clickers who do not convert)
So where do you start? I've outlined 5 easy steps to help you jumpstart your Web Analytics initiative:1. Define Your Goals2. Quantify the Opportunity3. Considerations and Reality4. Test Planning5. Analysis and Reporting
Define Your GoalsThis may seem like a no-brainer, but I am always surprised by how many campaigns get kicked off without clearly defined goals and objectives. Take a few minutes to outline what you are trying to accomplish and to define your success metrics before launching your campaign.
Examples of goals may include:. Increase revenue. Increase site traffic. Increase engagement. Increase conversion rate
Success metrics may include:. Email response rates: Open, Click, Unsubscribe. Sales per message sent. Conversion rate. Average order valueActionable Web Analytics Advice
Quantify the OpportunityBefore committing to a Web Analytics program, it is a good idea to quantify the opportunity to make sure you have enough volume to support a recurring campaign. Depending on the size of your customer file, this could be a simple exercise conducted by a business analyst, or it may require support from your IT team.
Note: If you don't have the resources to conduct this research, Yesmail Professional Services can help!
Here are some examples of how you could analyze event data to determine if the volume supports program development:
. Quantify daily/weekly volumes. Account for duplicate data and multiple customer actions. Summarize counts by brand, product category or business line. Forecast response rates (if you mail to 100 subscribers per day, how many do you expect to open, click, convert?)
Considerations and RealityI have learned a few things along the way that are useful considerations to be aware of even if you decide not to take action on any of them.
1. Data Integration: Understand the data integration process between your Web Analytics provider and your ESP. Some clients chose to send summarized data to their ESP, while others chose to setup direct integration between their ESP and Web Analytics provider.
Note: Yesmail has direct integration with Omniture Genesis Network and Coremetrics LIVEmail
2. Message Frequency: Consider your message frequency rules and understand how the new campaign fits into your overall message Lifecycle. Determine whether your new message will have priority over standard messages. Is it ok if a subscriber receives multiple messages in one day or week? 3. Hierarchies: Create a hierarchy for multiple events. For example, if a customer abandons multiple products at one time, does this impact your communication?4. Creative: Do you want to feature specific products in your email design, or do you want to take a more subtle approach?
Test PlanningOne thing I cannot stress enough is the importance of creating a controlled test plan to ensure results are clean and actionable. Test plans do not need to be complex. Here are some basic guidelines:
1. Use A/B test methodology to ensure you have adequate random test groups for reporting 2. Determine if you will track a true control group (no message received)3. Test options include: - Subject line - Creative (subtle approach or product specific) - Offer - Timing/FrequencyActionable Web Analytics Advice
Analysis and ReportingDon't wait until after you launch your campaign to think about reporting. Define your reporting requirements up front and code your customer records to support reporting needs. Make sure your reporting will answer questions about your objectives, and track your success metrics.
... [download for more]
