Because personalized content can have such a large positive impact on site performance and conversion rates, the market has become flooded with tools and programs designed for making this process as easy and effective as possible for the marketer. The purpose of this post is not to catalogue and review each and every one of these tools but to provide an overview that helps you sort through the various offerings and better determine what tool or method fits your business goals. Note, as discussed in the previous post, there are 3 types of personalization (Provider-Centric, Consumer-Centric, and Market-Centric) each with their own set of tools and services. For the purposes of this post, this overview will focus on the most common of these types, provider-centric personalization.
Data: The Engine Powering Personalization
The personalization process hinges on consumer data. The type of data and how a particular tool processes it are the most critical elements involved in generating relevant and accurate personalized content. However, there’s a fine line to keep in mind when running a personalization program. If your tool uses algorithms that are refined and using the right type and amount of data to feed that algorithm, conversion rates will generally improve, often times quite dramatically. A recent study from Hubspot examined 93k CTAs over the course of a 12 month period and found that those that were accurately targeted to consumers via personalization saw an average lift on 42% in turning visitors into leads. On the flip side, if the content is not relevant, the impact can be equally dramatic in a negative way. Consumers often cite irrelevant content as one of the top most frustrating features of a site. Because of this importance, we will consider personalization tools based on the type and amount of data used in serving personalized content.
Four Categories of Personalization Data
The data used in crafting a personalized experience generally falls into 4 categories. Optimizely provides a good summary of the first 3 types:
- Contextual: Contextual attributes are those that indicate the current state of a visitor but they change over time.
- Time of day
- New/repeat visitor
- Device type (mobile, web, tablet)
- Behavioral: Behavioral attributes are groupings based on current or past behavior.
- Past browsing behavior: Sites and content consumed prior to arriving at your site.
- Interactions with current marketing campaign: Interactions with marketing materials prior to arriving at your site.
- Past Purchase Behavior: Similar to browsing behavior, patterns in purchase behavior can be used to understand what your visitors are looking for and their tastes and preferences.
- On-site Action: Often used to segment users once they have engaged with your site. This can be used to define custom behavior patterns that identify which flows one audience may prefer to another.
- Demographic: Based on individual attributes. These are the classic methods traditional marketing categorizes consumers and are available through first or third-party data sources.
- Marital Status
In addition to these standard 3 categories above, I would argue we should consider a 4th category of data that tools are leveraging in the personalization process:
- Social:Based on the user’s behavior within a social media application as well as the characteristics and preferences of a consumer’s connections.
Different tools use different methods to leverage the data and make personalized recommendations. We can bucket these tools into 3 categories.
These tools largely leverage user cookie data to generate refined segments based on the 4 data types listed above. These systems are especially strong at analyzing contextual data and provide marketers an in depth view of how personalized content impacts a user’s behavior across the site. Because these tools focus heavily on the data and insights, they are more limited in the complexity of creative options you can serve up as alternative content.
- Coremetrics Intelligent Offer: While IBM Coremetrics is widely used by large retailers for analytics, it is increasingly used for personalization and up-selling.
- Adobe Test and Target: One of the best established personalization engines evolving from the original Touch Clarity back in 2004. Users can set up A/B and Multivariate tests of content and couple those results with cookie data to serve personalized content to consumers.
Smaller Scale Heuristic Based Tools
While most of these tools can leverage the 4 types of data described above, this category of tool usually relies on one or two of the data categories to perform a specific function within a personalization strategy. These tools are often used within an ecommerce strategy to provide personalized product recommendations based on shopping and demographic data. Below is a sampling of these programs, compiled by Dave Chaffey at SmartInsights.com:
Barilliance – Saas Personalisation for Ecommerce
Bunting Website Personalisation
Magiq dynamic personalization software
Monetate Personalisation Software
One Stop Model Based Platforms
These tools leverage significant volumes of consumer behavioral data to develop complex consumer modeling to generate predictions for what content will be of particular interest for the consumer. Below are a few of the more popular solutions in this category.
- Conversant merges company data with a proprietary database of behavioral data to generate granular user profiles and serves personalized content with a real-time, predictive model.
- SalesForce has a number of tools within its platform to customize and target personalized content. It has a predictive intelligence engine that consumes explicit and implicit consumer behavior in real time to build individual customer profiles. Content is then served based on a series of business rules to predict the best piece of content (site, email, display etc.) for each consumer.
- Magnetic personalizes content via search retargeting based on an algorithms and an owned data source. The algorithm predicts a consumer’s intent and serves content based on a set of complex rules. It also uses its database of buying behavior and geo targeting to segment active and dormant costumers and provide relevant content accordingly.
As I indicated at the top of this post, this summary of companies is by no means comprehensive but intended to give a general overview and strategy to better understand the playing field of tools available for personalization. If you’re looking for a more hands on, in-depth view of the data, then an analytics-based solution might be what you should consider. If you need more of a strategic/focused solution, then the smaller scale solutions would be worth considering. Finally, if your organization has a larger budget and is looking for a more robust solution that touches on multiple business areas, one of the one stop model platforms are likely where you should invest. Whichever tool you choose, an essential part of any personalization program is understanding its impact on your ROI. The final post of this series will address this point and offer a similar overview of the methodologies and best practices for analyzing and optimizing the business impact of personalized content.