Real-time data-driven personalization is key to marketing success. It’s top of mind for every marketer. So it is no wonder that Optimizely has become the latest A/B testing vendor (after Optimost, Maxymiser, and others) to add some personalization features to its A/B testing solution and enter the personalization market.
Optimizely has built an easy-to-use and widely adopted A/B testing solution. So, one might assume that a natural next step would be to add personalization to its product, right? Well, no, it is not.
Think of this car as Optimizely’s A/B testing solution. It runs well and gets you where you need to go.
Optimizely’s engineers have spent seven years honing their product’s functionality and ease of use. To their credit, they’ve done a great job.
Now, think of Optimizely’s executives coming to these engineers and telling them to build a personalization solution as soon as they can. Yet a robust personalization solution requires several key ingredients:
- Maintaining a deep visitor/user profile
- Collecting, storing, analyzing, and using orders of magnitude times more data per session than is needed for A/B testing
- Understanding the meaning of each site’s pages and custom events
- Delivering algorithmic as well as rule-based responses computed and rendered in real time
In other words, it’s far more involved than just splitting traffic between a few “candidates” and determining the winner.
Evergage has been focused on doing all of this for the same amount of time Optimizely has been evolving its A/B testing capabilities.
Optimizely’s engineers didn’t have time to just start over. They couldn’t completely re-architect their approach. They had only one choice: take what they had and extend it.
The crux of the matter is that personalization is a real-time big data problem, whereas A/B testing is not. Optimizely’s efforts to add personalization as an extension to its testing tool is not like adding special tires to the car. Rather, it’s more like adding wings to the car and then asking marketers to fly it — a rather scary proposition.
The contrast with Evergage’s approach is stark. We began with easy-to-use, big-data-powered, 1:1, real-time personalization as the goal, and developed a purpose-built solution very different from Optimizely’s. Understanding the difference is as simple as understanding The Power of 1.
Let’s take a look under the hood and compare Optimizely’s approach to Evergage’s.
The Power of 1
|Optimizely: The Limits of Extending an A/B Testing Architecture|
|1 Platform to use|
A single personalization platform purpose-built to provide a marketer with:
Without the need for coding by your developers, Evergage understands the full context of your site.
Evergage enables you to personalize across channels:
Evergage can easily connect to your existing datasources like CRM, Email Marketing, Data Warehouses, In-Store Purchasing, and more.
|Limited functionality. Limited behavioral insight. |
Bolted on behavior-tracking
Optimizely was not originally architected with a need to track visitor behavior. In other words, it wasn’t built with big data in mind. Originally, the only data that was needed was whether a visitor was in a test or not and whether the visitor completed a goal or not. Optimizely has added the ability to track custom events, but the way it has done so has dramatic limitations.
Limited customer analytics
With Optimizely, after all the work necessary to track a visitor’s behavior, none of it can be used to help with your analysis of campaign results and attribution. A custom event used to track a visitor’s behavior cannot be used for dimensional analysis of attribution reports. This same information must be sent as a dimension.
Unproven old-school not real-time recommendations capabilities
Optimizely recently announced recommendations. These recommendations use an old-school “wisdom of the crowd” approach, and do not make recommendations to the individual that are truly 1:1. The recommendations updated every 24 hours, a far cry from real-time. To date, there are no public examples of customers actually using Optimizely for recommendations.
Lack of cross-channel personalization
|1 Profile for each visitor/user driving 1:1 experiences|
In-depth behavioral data is captured, external attribute data is synchronized, and everything is available in a single unified profile for each individual visitor/customer/account.
The profile is stored centrally, accessible from anywhere, and updated in real time. Targeting rules and personalization algorithms run server-side, accessing this profile and returning the results to the browser, where the personalization occurs.
Personalization is unique to each person, truly 1:1.
|3 limited distributed profiles |
In contrast to Evergage, Optimizely built a lightweight shadow profile that is stored in the cloud, while most of the visitor profile is stored in a mix of the visitor’s browser and cookie. This approach was a natural extension of Optimizely’s A/B testing architecture, but not a sound architectural decision to support true personalization. Here are the key drawbacks:
|1 Moment to engage visitors|
All aspects of a visitor’s history is combined with all aspects of the current session in real time. All that data can be combined with everything you have learned from others, so you can deliver split-second personalization into web, mobile web, mobile app, email or search.
A marketer using Optimizely will always have to ask, do I have too many segments? Too many campaigns? Which should I deactivate? When do I deactivate?
In contrast, Evergage allows for unlimited personalization campaigns and unlimited segments, without these affecting performance.
|1 Marketer to manage the solution|
A single marketer can launch personalization campaigns in a matter of hours, without the need for developers or IT. Using built-in attribution analysis, the marketer can iterate rapidly to ensure delivery of the experiences and algorithms that get the best results.
|Requires guesswork and development cycles|
Optimizely’s A/B testing platform is easy to use. However, with the addition of personalization, Optimizely needs to be able to track and understand behavior. This addition radically decreases the product’s ease of use. It requires a developer to add custom events for everything of interest to the marketer. And a release cycle is required before any custom event can be tracked.
Additionally, it is impossible for a marketer using Optimizely to assess audience size and priority for targeting prior to actually launching the campaign.
|1 Company dedicated to your success|
Personalization experts, including a dedicated Client Success Manager, are passionate about helping you get the best results possible. They provide know-how, best practices, proven “playbooks” and more – all at no additional cost.
|Limited experience with personalization|
Optimizely is new to personalization and does not have the experience, knowledge, best practices and overall resources needed to effectively support a personalization program.
Let The Power of 1 Work for You
Marketers, I encourage you to dig into the capabilities of the personalization vendors you speak to. Don’t turn to an A/B testing platform to solve your personalization needs. That is not what it was built for, and you will run into limit after limit.
I do encourage you to check out Evergage. Our jet engine and wings are not an afterthought, because we started with the end in mind. We will help you fly. Download The Power of 1 whitepaper to learn more about how the platform can work for you.