Website optimization is a journey, not destination. So while many website optimization platforms claim to have similar capabilities, the real concern for a potential buyer is not only what a platform can do, but how easy that platform is to use. After all, in competitive industries, the true cost of doing business is measured in time.
In this post, we’ll look at the process of product badging (which is placing an emblem on a product image thumbnail) across the three dominant architectures in the website optimization space. The three product images shown below look quite similar, right?
All three might feature the same “On Sale” badge, but the methods by which these architectures (would attempt to) get there reveal the devil in the details. First, let’s define these architectures:
2) Server-Side: Server-side platforms serve each test element via HTTP request and response. Because each test element on the page requires a separate server call to return the content, server-side platforms tend to be among the slower methods for delivering content. In addition, IT resources are needed to code each test element and deploy it into production.
3) Proxy: Proxy solutions reside within the network, sitting between the client’s server and user’s browser. They work by grabbing a page after the host has produced it, altering the page, and then producing a new version for final delivery to the end user.
So how would each of these architectures go about badging product images?
Client-Side: Client-side platforms offer one-tag integration, so there’s no additional code needed for any campaigns, and no need to worry about website release schedules. It’s as IT-free as website optimization gets. Client-side platforms also dynamically overlay the badge on top of your existing image thumbnails. Because there’s no need to create alternate versions of each image, there’s no burden on Creative to modify your images.
But a strong client-side platform must combine IT-free implementation with real-time knowledge of your product catalog (from the category and brand down to the individual SKU), an understanding of the types of pages on your site (category, product, etc.), and the building blocks of each and every page.
These capabilities are what make the product badging process highly scalable. Want to badge all 600 products in your “Shirts” category, whose products might be featured on the homepage, category, product, and “top sellers” pages? Catalog intelligence (i.e., an understanding of what you sell from the category to the SKU level) and knowledge of your page/site structure make it possible to apply simple but powerful business rules to badge all relevant products in a single campaign.
One-tag integration, true catalog intelligence and knowledge of page/site
structure combine to make dynamic product badging easy and highly scalable.
Server-Side: The server-side platform can’t overlay the badge on top of your existing images; you’ll need to create new images so that the tool can swap the original for the new one. Creative resources—check. Next, you’ll need test-specific code to perform the swap and will have to fit test deployment into a future website release. IT resources—check. 600 shirts? That’s 600 separate pieces of throwaway code, useful for no other purpose on your website.
Proxy: The proxy solution fares no better. Like the server-side platform, your creative team will need to produce a new image and make it available for the proxy to swap in (dynamically rendering the badge on top of your existing image isn’t possible). Creative resources—check. And although deployment doesn’t require test-specific code, doing anything requires your knowledge of Regular Expressions. IT resources—check.
And neither architecture has catalog intelligence or an understanding of your page and site structure. That means the process of badging a product begins, first, with just trying to figure out all the places where that product actually appears.
When evaluating potential website optimization solutions, keep an eye out for each platform’s capabilities. You need business-relevant options, and shouldn’t settle for anything less. But your selection criteria should include questions about the processes that lead each platform to its results:
- Does the action require test-specific code?
- Does the action require me to duplicate or modify my creative assets?
- Can I repeat the action on multiple pages? Is additional code required?
- Do I need to involve IT to create and deploy tests?
- Does the tool have true catalog intelligence, understanding what I sell from the category to the SKU?
Why do these matter? In a sentence, the time it takes to set up and deploy tests is inversely related to the number of tests you’ll be able to run. More tests equals more insights, which means more revenue-generating opportunities for your company. It’s all about what conversion marketing expert Bryan Eisenberg calls “Return on Time Spent,” a concept which is related to the number of concurrent live campaigns (what we call CLiCs) you’re able to run. The easier a campaign is to create, the more of them you’ll have—which means more revenue opportunities for your business and fewer requests for your already-constrained resources.