What is Test&Target

[ADOBE TRAINING SERVICES For Omniture technology] Hi, everybody. This is Doug. Today we're going to talk about what is Omniture Test&Target. Well, the unofficial answer is that it's pure awesomeness. But the official answer is that Omniture Test&Target gives marketers the necessary capabilities to continually make their online content and offers more relevant to their customers, yielding greater conversion. Test&Target provides an intuitive interface for designing and executing tests, creating audience segments, and targeting content all from a single application. So Test&Target is part of our Omniture online marketing suite, of course. You'll see it right over here in the optimized section at the top. The Test&Target is very well named. It's going to give you two main abilities, to test and to run tests on your site to see what's working and what doesn't work, and also to be able to target segments of your audience. So, in this case, you have all these people out there that are visiting your site. You want to know which version of different content will increase conversion on your site. You want to be able to target people from different geographical areas or people that do different things on your site or that use different methods to land on your site and these kinds of things, targeting different kinds of segments, new visitors, returning customers, etc., so that you can really have that suite spot and know exactly which content will work for specific people, including being able to get reports on that and being able to provide them with the right content at the right time. So some of the challenges and needs that we have, we have lots and lots of content, but how do we match that to the right audience and the right visitor? We know we have lots of different kinds of people, that our audience is very diverse, but how do we segment that properly? And how do we identify these targetable segments? This is one of the most common questions, where you start. How do we decide what goes on our site? What works? How can we validate these proposed changes with confidence? How do we know? We can think that it works better for people, and we can think that it looks better, but how do we know and validate that with confidence? And, again, how do we set up that test to make sure that that test will show us exactly which version of content improved conversion on our site? Well, Test&Target has a very easy-to-use interface and that way marketers can just jump in and create tests and run them without having to get into the code. That's really one of the coolest things here. Once there is some basic code set up on your site, then, as far as changing content in and out and trying different html content or trying different images or just using different versions of your site, is all done in the interface, like you see here, rather than having to go into the code and switch out different copy. Okay, so very, very cool. And this is how we can run tests and be able to see exactly which items are working better. So you start again here with a control page. This is what our page has been looking like, but we're going to try experience B, here, and we're going to see if it is actually leading to more conversion. At the same time, maybe we actually have three versions that we want to test, so we can see exactly which version is going to produce better results. And with Test&Target, you can run these tests and then get an exact number on increasing conversion and you can also get statistical confidence on that so that you know that it is really a good number and that you can count on it. Now these kind of show three different versions of whole pages, but if we had, for example, experience C, and we said, okay, this layout works the best. Let's tweak this main picture, here. You can also, then, run the test to see if different little changes in this picture here on this page will increase conversion, as well. So, always a good idea to kind of refine your experience down. So we can see which layout works the best, and then we can refine it with different versions of copy, maybe a different paragraph down here, a different heading or whatever you want to do to see exactly which one works the best. So, never stop testing. We always want to see what we can do to get things to work better and better. In Test&Target you also have multivariate testing, which allows you to test more than one element at a time. Instead of just testing a couple versions of one item, like the main picture, we can test things like the picture itself or the message on the picture or the price or a phone number or the text in any of these areas and whether there's bullets or not. So you can have different versions of different elements at the same time and test them all in conjunction with each other to see which versions of the elements are going to drive the most conversion. That's done right in the interface, and you can test many different elements. You can see here that if, for example, you had seven elements and two versions, if you're going to test all those different variations, you would have to test 128 different versions of the page. Well, with multivariate testing and the magic that it is, you can end up only testing eight alternatives, and you actually just build it right in the interface. And Test&Target then predicts the winner by determining the individual element contribution and which elements will work best together, and then it will give you the winning alternative, here, so you can see exactly which ones you'll be able to push and drive the most conversion on your site. And it's actually very likely that a version that didn't even get tested will be the winning version because of understanding how the elements work together. In fact, you're really only testing eight versions out of 128 possibilities, so it's actually more likely that one of the other 120 versions would be the best, and that's really the magic of multivariate testing. So that's in Test&Target, as well, on the test side. Now let's look at targeting. Here you have right in the middle here, "Salt Lake City Take Control of Your Money. Join Dave Live on May 16th, etc." So we can go in, and Geo-Target offers to specific audience segments, and we're going to select a parameter, a Geo-Location parameter, here, whether it's city, country, etc. We're going to select a city. That's Salt Lake City, in this case. And what if we want to switch that out and see Dallas or Kansas City or Colorado Springs, etc, for the different places where this might pull up? So we can actually target the Geo-Location with the right message just by looking at where they are from. So that's a very powerful way, and just one small example of how targeting can give certain people coming to your site a certain message. So, in review, Test&Target, you want to be able to run tests to see which versions will work best overall on your site, or even for different groups of people, by targeting various segments of your population. [Adobe]