Quick Tips: Experiments by Vertical – What to Test for Ecommerce Sites

Quick Tips: Experiments by Vertical – What to Test for Ecommerce Sites


Hi everyone, I’m Krista Seiden, Analytics Advocate at Google, and in this Quick Tip video today I’m going to give you some ideas of what you might want to test if you are an e-commerce website. So with that let’s go ahead and jump right in. On your screens right now you can see an example of an e-commerce website. So this is the Google Merchandise Store. This is where you can go to buy Google-branded gear online. It’s also what we use as our demo account for Google Analytics. So you can actually see all of the data for this account as well. So here’s our above-the-fold view, we have a big high bar selling winter jackets and as we scroll down the page you would see that there’s a lot of different boxes advertising different products on the Google Merchandise Store. So let’s go ahead and look at a few examples of what we want to test on this store. So first of all if you are an e-commerce site you very likely have a checkout flow on your website, and this is going to be a crucial element for your site in order to get people to actually go ahead and buy your merchandise. So it’s also going to be a crucial area that you want to test. Let’s look at a few different ideas of what you can test here. So first you can simplify texts or bullets, or try to remove a lot of extra text on that page making it a very simple-looking sign-up flow. You can change your checkout CTAs. Perhaps it says Continue To Checkout, or Buy Now, or something like that. You could reduce the fields required, you know, maybe you’re asking for a whole lot of information about their shipping, and their email, and their phone number, and all these things, and perhaps you don’t need all of that information to be able to go ahead and sell them the product. So you can try reducing those fields and see if it increases your checkout rate You could also reduce the number of pages or steps that is required in your checkout flow, and you can also try to adding or removing a progress bar. We’ve seen that a lot of times progress bars tend to have a high impact and flows like this if you let people know how many steps they still have to complete. So another example of the test is discounts and targeting these based on user behavior. So in this example here I’m actually targeting a test to people that we consider to be loyal customers. So these are people who have purchased from us in the past. That’s how I’ve set up the audience that I’m targeting this test to. And when they land on our website I want to give them a discount. I want to say thanks for being a loyal customer. Enjoy 20 percent off your next purchase with this promo code. That’s a nice way to engage with your users and thank them for their loyalty. So finally I want to walk through step-by-step one last test here that you can try out if you are an e-commerce website, and I’m calling this a Homepage Apparel Content, or Apparel Takeover. So the reason for this test is we want to provide a personalized experience for specific audience, and our hypothesis is that we will increase revenue by showing top-selling apparel products on the homepage for those who have previously shown a strong interest in the apparel section of our website. So we’re going to show this test just on the homepage of the Google Merchandise Store. We’re only going to have one variation. We’ll have the original plus our one variant, and we’re going to target this test to users who have visited the apparel category more than once before in prior visits. Our goal in Google Analytics is going to be revenue. So we actually have a couple of data points that support our hypothesis that we are pulling directly from Google Analytics to support building out this test, and they are that apparel is actually the best-selling category on the Google Merchandise Store, and in that category t-shirts are actually the best selling item within the apparel category. So to give you an idea of what this might look like, I have created a new version of the Google Merchandise Store. So you can see on the left the red boxes which I’ve replaced on the right with some t-shirts and a hat, and this is really getting at that data that t-shirts are about selling and that apparel overall is best-selling. So we’re going to go and see how that does. A few other things that you can always look for as an e-commerce site, conversion funnel drop off. So you have probably a couple different options for conversion funnels, getting somebody to add a product to their cart could be a funnel, getting people to check out from cart could be another funnel, and you’re going to want to analyze the different steps in those funnels to see if there’s one area that people might be dropping off more than others, and then try to optimize that step. Keywords, it’s great to use Adwords Keyword data to find really high converting keywords and then use that as messaging on your landing pages. So consistency between ads and website generally helps to drive people to your conversions. And finally, you can always look at high page views but low conversion rates and low page views but high conversion rates to see if there’s something else you can do there to optimize those pages. So there you go, I’ve given you a few different examples of what you might want to look at and test if you’re an e-commerce website.

Leave a Reply

Your email address will not be published. Required fields are marked *