Every year, we speak with hundreds of e-Commerce program managers, analysts, and clients alike. In these conversations, we’re often asked a variety of questions, but one in particular always appears to withstand the test of time: “where do I get started?”
It’s easy to understand why – with so many layers of code, template language, corporate red-tape, and design complexity, just the idea of optimizing your e-Commerce site can seem like a daunting task. In reality, there are many small steps you can take to get started with your optimization program today.
- It’s okay to start small and scale as you grow
- Ensure that your results are statistically significant prior to taking action
- Full-funnel optimization creates more permanent and sustainable revenue growth when compared to single-page testing.
Picking a Solution
If you’ve already selected an e-Commerce optimization tool, implementation should be a straightforward process managed by either yourself or your vendor. Most eRPM (e-Commerce Revenue Performance Management) vendors offer fully-managed service plans, which provide you with an actionable strategy for campaign development that is then executed wholly by their services team.
If you need help finding the right fit for your e-Commerce optimization goals, be sure to check out our 2012 Solutions Guide, which contains a great deal of valuable information about the types of solutions available in the marketplace today.
Step 1: Start small
It’s important to understand that even small tweaks can have a significant performance impact. If you’re new to optimization, start by gaining familiarity with the technology you are using and communicate as much as possible with your vendor to find ways to take full advantage of the tool(s) at your disposal.
Start with a small experiment such as a single element change in order to fully understand how one campaign works within your optimization engine. Take note of how it allocates traffic and what happens when the “winning” (i.e., high-performance) version is identified. Some of our previous clients have started large, complex campaigns that yield great results simply by tweaking a banner or call-to-action first.
While you’re just getting started, I recommend staying away from running multiple, simultaneous campaigns. We find many instances where companies are running multiple, simultaneous A/B campaigns, ambitiously (yet incorrectly!) believing that they can simply combine the “winning” page elements together to produce a positive lift.
One new client wanted to start a small experiment by testing four free shipping options. Each had a different minimum purchase threshold:
‣ free shipping for all orders
‣ minimum purchase of $50, $75, and finally $100.
The expectation was that free shipping for all would outperform those with minimum purchase thresholds.
After the first week of this campaign, the client was surprised to learn that the highest revenue per visitor (RPV) came from the offer with a $50 minimum purchase threshold, followed by an offer with a $100 minimum.
Step 2: Be confident
The preferences of your online shoppers are constantly changing. In order to determine whether the results of your test(s) are significant – that is, the performance differences observed between test versions are unlikely to have occurred by chance, and also that the magnitude of change represents real outcomes – you need to wait until you reach your desired level of statistical confidence. This can take time, depending on the number of combinations you’re experimenting with and the amount of traffic you’re allocating to your optimization experiment.
Step 3: Start by focusing on one page template
Once you’ve succeeded with a single-variable experiment such as a banner or call-to-action, I’d recommend focusing on configuring your campaign to include multiple changes to one page template. If you’re unsure about where to begin, try starting with your product detail page or the shopping cart page (i.e., the first page in the checkout funnel).
By experimenting with a single page template, you will be able to understand how the platform handles multiple page variations. An experiment like this can help to demonstrate how each individual element on your page is interdependent with other elements, and how element interactions impact your conversion or RPV goals.
Example (follow-up from above)
As the campaign from the previous example continued, visitor reactions changed, and the winning offer began to perform poorly. In fact, the $50 minimum purchase threshold produced a -12% drop in revenue per visitor.
So, if this client had permanently implemented the $50 minimum purchase free shipping offer after the initial positive results, they would have missed out on significant revenue potential.
Step 4: Focus on the full-funnel process
Most online purchasing patterns involve going through a multi-page process. As such, the ideal optimization campaign would treat each of these multiple pages both independently and as a whole, with various elements that span across all pages simultaneously. Any solution that allows you to conduct full-funnel experiments also allows you to ensure a consistent user experience and contributes to more sustainable (permanent) revenue growth potential. By implementing an optimization campaign, you will glean actionable insights that help you brainstorm campaign refresh ideas based on actual behavior.
If you’ve chosen a solution capable of adaptive, full-funnel multivariate optimization (AMVO), this seemingly significant effort is actually a hands-free, continuous and automated process.
One key lesson of adaptive multivariate optimization is that many small changes tend to compound and produce noticeable, sustainable increases in web page revenue performance. If you’re new to e-Commerce optimization, you may want to try starting with small tweaks over the course of the first few weeks of your program. You may notice a modest uptick in revenue, but more importantly, you’ll be able to glean actionable insights that will guide you to better brainstorm optimization campaigns in the future.