Dynamic Targeting - Four key building blocks.
Our hosted solution is based on four key technology elements - profiling engine, business rules engine, content delivery, and optimization testing. Together these elements work together to deliver a personalized and optimized visitor experience.
The Kefta profiling engine automatically acquires individual visitor information and builds an ever-increasing level of understanding about user needs and expectations. It is the detailed nature of this data that allows your team to target both prospects and returning customers. Each customer is handled as an individual case that keeps track of user and reporting data. The case user profile data is updated in parallel as the user browses your website, submits forms and so on, and case historical data is updated as the engine receives event notifications from Kefta probes (Kefta code snippets you insert to affected pages).
Defined by business needs and deployed by Kefta, these profiles are key to understanding the differences between site visitors. It is the grouping of these differences that allow you to target highly relevant communications to site visitors. Some of the dimensions by which visitors can be profiled are detailed below.
Example inbound-click information available for tracking:
- Keywords typed through a search engine, paid and organic
- Referral sources such as affiliate sites, banner ads, text links, and third-party websites
- Cookies
- Connection speed
- Geographic location
- Browser type
- Operating system
- Native language
Example user behaviors tracked:
- Frequency of visit
- Products and offers viewed
- Purchases made
- Forms filled out
- Responsiveness to online and email offers
The engine continues to develop this profile throughout each customer's lifecycle. From the first click as a prospect through repeat visits, to purchasing, to customer loyalty, our profiling engine maintains the relevant information that you need to offer relevant communications.
The Kefta deployment team translates your online business objectives into a series of business rules that identify high value activities and initial visitor patterns to be aware of. As well, this engine is highly intelligent in that it automatically translates conflicting inputs, and scores them against desired outputs in the attempt to delivery the most relevant experience for each visitor.
The business rules engine has the ability to draw conclusions between profile data, and desired outputs, and from those form alternatives that are likely high-value options.
| Example actions the business rules engine recommends: |
|
|---|---|
| Example scenarios for an automotive site: |
Off-site targetingA prospect clicks on a car advertisement on a partner's site, Car Site A. The prospect is transferred to client's site, then leaves after viewing several pages. Three days later, the same prospect goes on a different partner's site, Car Site B. Using business rules to target key client objectives, the prospect will see an ad for the same car they viewed three days earlier. The message surrounding the image can also address a visitor's stage in the cycle of purchasing. Off-line / online targetingA customer bought a car at Dealership A. Leveraging business rules to cross-sell the customer. Eight weeks later they receive an email presenting a relevant service offer, directing them to go online. Once online, the customer will see the relevant service offer prominently displayed. Renditions of this ad continue to be displayed until the time that they purchase the targeted service, or business rules direct the customer experience to a new objective. Capturing a leadA couple visits the client site to gather information on a car, but does not complete the online lead application. Kefta's Dynamic Targeting system recognizes that the couple has looked at the lead form, but failed to complete it. Using business rules to trigger desired next steps, the visitors are invited to speak with a sales representative via a pop-up as they close their browser, exiting the site. The system can offer different messages and offers to drive the contact, as well as different options for each visitor to take advantage of. |
Using the information gathered by the profiling engine, as well as the functionality of the business rules engine, the content delivery system populates the appropriate, visitor targeted variables to your site. Variables can be comprised of any combination of the following: text, images, page layout, linking options, templates, and code.
Typically, the assets that are served can reside on the client side, within Kefta, or another third party.
Example delivery options across key online channels:
On-site:
- Images
- Links
- Page styles or templates
- Text
- Adjusting layout to screen/window size
- Pre-filling forms
Off-site:
- Banner ads on other sites
- Email messages
Extra-site:
- Layers on top of your pages
- Pop-ups
Our engines collect and monitor visitor results in real time and determine which variables to deliver to which visitors at which time.
The technology constantly measures and monitors the success each visitor has against other visitors, who although similar in their profile, were served a different experience. It then determines which paths and variables lead to the most success. Based on the business rules, the system evaluates which traffic should receive which variables in order to increase successful outcomes.
This form of optimization is unique in that it evaluates successful patterns in the context of those who are similar in profile. This is a critical point, as site visitors consistently demonstrate powerful patterns within visitor groups.
When these patterns can be identified, success becomes both dramatic and scalable. Using state-of-the art full factorial and fractional factorial models, such as A|B and multivariate testing and Taguchi orthogonal arrays, the Kefta optimization system is feature-rich and easy to deploy.
Beyond these typical models, the Kefta Dynamic Selling solution is unique in its able to learn from the activity of real visitors and proactively deliver high value experiences, not just identify those experiences.
