Targeted Marketing using Behavioral Analysis
Akin Arikan is the author of Multichannel Marketing: Metrics and Methods for On and Offline Success. He is also an evangelist for Unica’s web analytics and interactive marketing solutions. Akin’s blogs at Multichannel Marketing Metrics
The following poll was taken at a web seminar with Search Marketing Now: Getting from “What Happened?” to “Why?”, sponsored by Unica. Our speaker, Evan LaPointe, asked the audience (mostly web analysts and marketers) to check which of the choices they perceived as challenges for their use of analytics. It is striking how many in the audience were in agreement that going from mere reporting to behavioral analysis was the most common challenge.
There are many interpretations and questions that we can draw from this anecdotal information. One of these is about the use of behavioral analysis for targeting purposes.
Two Uses of Web Analytics: Aggregate and Individual Level Insights
Conventionally, marketers use web analytics at an aggregate level. They seek to report on the performance of their web sites and advertising, so they can adjust their efforts to improve the results. This is an extremely worthwhile application that can deliver excellent return on investment. For example, Unica’s customer Citrix has a published case study with 1900% improvement in conversions.
However, marketers may be squandering a huge opportunity if they do not also leverage web analytics as a rich source of behavioral insights on individual prospects and customers. Using behavioral targeting, web analytics can play a far more direct role in engaging customers, improving customer experiences, and increasing sales, by enabling companies to deeply personalize their communications and interactions.
Identifying What Behavior Nuggets To Mine For
Individualized web analytics can reveal actionable insights into individual prospects and customers. For example:
- Individuals’ personal preferences and current product or content interests
- Where individuals stand within the buying or customer life cycle
- When they are most susceptible to being persuaded, converted, or up-sold
- When timely action must be taken to retain them
- Which offers will be seen as most relevant and persuasive
- How much each individual will be willing to spend
These insights can be transformed into targeted marketing initiatives at every stage of the customer life cycle (see graphic).
Figure: Mapping life cycle stages and business goals to initiatives that can be driven by personal web analytics
Targeted Marketing Initiative Example
Venn Diagrams are useful for many purposes (see my recent post on fun Venn diagrams that describe the challenging job that web analysts face). A very cool way that Venn diagrams can help in analytics is for segmentation and targeting. The screenshot below is an example from Unica’s Interactive Marketing OnDemand product which combines the web analytics product (NetInsight) with Unica’s products for email marketing, testing, and website personalization.
The Venn diagram shows the overlap of 3 segments:
- Received a recent email campaign (data from the email marketing component)
- Clicked-through on the email (date from the web analytics component)
- Converted (data from the web/customer analytics component)
The blue slice represents visitors who clicked-through but didn’t convert. They would be an ideal target segment for a re-marketing campaign. Their behavior suggests that they took an interest in the content of the email but weren’t quite persuaded. So we could send them an email reminder if they don’t return to the website on their own in coming days. Or if they are frequent visitors to the website we could create a dynamic banner that will be seen by these visitors to remind them.
Meanwhile the visitors in the Turquoise segment have clicked and converted. Since they seem open to receiving and responding to our marketing email we may want to target them with a cross-sales offer next.
This can be accomplished by saving and exporting the lists to email service providers. Or in the case of the Unica InteractiveMarketing OnDemand product the email and website personalization capabilities are built directly into the solution.
Takeaways
- There is a goldmine to be found in using web analytics data for behavior analysis at the individual level which most companies haven’t tapped yet.
- Usually, the reason for this lack of action is not negligence by analysts and marketers but the scarcity of integrated web analytics, email, and personalization solutions.
It is our mission at Unica to make behavioral targeting easier by providing better integrated web analytics solutions to our customers.
Posted in Targeting & Segmentation | 5 Comments » | Akin Arikan | July 9th, 2010
Tags: Articles | Behavioral Targeting | Featured | Marketing Measurement | Marketing Optimization | Personalization | Segmentation | Targeting Articles | Website Optimization
I must say that from now on you will be Mr. Venn… The Venn diagram presented in this post is a truly original way to visualize campaign attribution, very mind opening.
Great article! Would Post View Tracking also help in this process?
Good point, Holger! Let’s say your company is heavy into display ads. If person X has been exposed to your ad (or even exposed with multiple impressions over time) they might make a better candidate to target.
Not easy getting the ad serving data to match up with the web analytics data and contact info. It can be done but will require dealing with the large volumes of ad impression data.
Can you describe what the portions of the Venn diagram where just two segments overlap represents? I’m not sure I see what the orange, purple, and green segments represent, they seem to be non-entities.
Hi Greg,
Good question, i should have pointed out better.
Green: Received mail + clicked + converted.
Orange: Received the email – did NOT click – but did convert. As you can tell, this requires not just web analytics but also customer analytics. This slice is about attributing response back by matchback to the list of email recipients.
Purple: Yep, this one should be empty. After all, red are the people that received the email but didn’t click. Blue are the people that received the email and did click. The intersection would be empty since you can only be in either one or the other bucket if you received the email.