During a recent webinar with Anthony Mills, we explored the hot topic of personalization and were given 5 strategies organizations can apply quickly to see measureable lift. We wrapped up this webinar with a live Q&A, and got to the heart of what many are curious about in the personalization space. Here are the questions you asked, expertly answered by Anthony.
How does Optimize 360 support personalization in real time website presentation. Meaning: how does it help identify a target user and present a unique experience on their website?
The real time capabilities of Optimize are still a bit limited. In Optimize, audiences can’t do real time delivery just yet. They’re more suited for return visitors. However, by using cookies and dataLayer elements, you can develop a next page delivery of content.
For example, let’s say a user clicks on a CTA on page 1 and we want to deliver a unique experience to them on page 2, based on that action. We’d need to cookie that person in the moment that they clicked on the CTA on page 1, so that when they arrive on page 2 that cookie is present. We can set that up in the targeting criteria of Optimize to deliver that content again.
How do you scale content for personalization?
If there is anything that slows testing velocity and learnings within an organization, it’s probably going to be content. Organizations can come up with great testing recommendations based on the data very quickly, but what tends to slow things down is considering which creative variations to test against and the hours required for content development. There are a couple of things that will significantly support scaling your content for personalization:
- It’s the job of the analyst and the UX team to work very closely together. As analysts, when we see data telling us something, we suggest one creative change. This is generally going to be considering some of the previous tests we’ve seen be successful. We’ll use UX heuristics like: motivation, value proposition, incentive, and friction anxiety. With the help of the UX team, we’re able to prioritize things. This helps focus the creative team, so they’re not trying to boil the ocean.
- Another really important element for scaling content is creating a testing backlog for your team. A testing backlog is an inventory of 5-10 test ideas ready to run at any moment.With this testing backlog built up, it allows you to project plan. And in project planning you can have creative and UX start working on problems in advance of them being executed.
Is there a best practice for introducing and piloting a new optimization tool (Google 360) when one is already in use in your organization (Adobe, Optimizely)? Is this something that you recommend?
Cardinal Path is currently working on a competitive vendor assessment, which is based on various use cases for testing and personalization programs. So be on the lookout for that!
When recommending one conversion rate optimization (CRO) tool over another, it’s important to identify a couple of things: 1) How well is the organization using their current tool, and 2) what are the pain points with their current tool.
We’ve had clients come to us using a very robust CRO tool and the team managing the tool hadn’t used it in the past, or maybe the people who had the most experience with the tool had left. Some clients come to us with issues in the way that their CRO tool may be passing information to their analytics tool. At the end of the day, it’s important to clearly identify the pain points keeping you from running the testing program that you want to have with that tool. Once pain points are identified, it takes our conversation to 1 of 2 places:
- We can discuss how to resolve what those pain points are with the current tool
- We can discuss moving to a new tool.
If you’re already well integrated into the Google Marketing Platform: you’re using Google Analytics regularly, you’re using Google Ads, the display network and some of the other media capabilities of Google – Optimize is a great option. Or if you have a new testing program or people who may be new within your testing program, Optimize is an easy add because you don’t have any 3rd party integration pieces to flesh out. The workflow is very short with Optimize. Whereas the workflow can be much longer on other CRO tools that have to add in 3rd party analytics.
Of the 5 personalization strategies that you shared, do you see a common/average KPI lift range?
The great thing about kick starting your testing program is you’re able to make a big splash quickly, which is great because it allows you to substantiate the resources you need or the budget that you need to mature your program. When you haven’t been doing testing in the past and you start doing it, its very easy to pinpoint where leakage is occuring. You can do this simply by a quick walkthrough of the customer journey on your site from a UX standpoint, or using a bottom up data-driven approach to find out where more specific leakage is occuring.
One thing that organization can get caught up in is trying to affect bottom line KPIs when just starting out your testing program. Things like: ROI, sales rate, or conversion rate of a purchase are ultimately the KPIs you want to impact, but those are harder to impact because you have less traffic volume reaching that point.
Start where your traffic is large and get big returns on that. Things like: your bounce rates, exit rates, and clickthrough rates along critical paths. These are the KPIs that will expand your funnel! With insights from those KPIS, you can apply those learnings to the more difficult, bottom-line KPIs. While the KPIs may not be as sexy as ROI, reducing bounce rates and exit rates will ultimately work towards those bottom line metrics.
With all of this in mind, when you create your testing backlog, one of the goals of that backlog is to estimate your test’s expected KPI lift. We usually back in to that expected lift per KPI by looking at what the current baseline is for the segment we’re targeting, and looking at similar pages with similar content objectives or similar segments and seeing if anybody is exceeding that baseline. Using those insights as a benchmark its easier to estimate your specific KPI lift.
Want to talk with one of our personalization and testing experts? Reach out, at: firstname.lastname@example.org