Despite the optimistic start to 2024, a recent study from Deloitte revealed digital marketing spend was down 45% in 2023. We commonly work with clients who are frustrated they don’t have the right platforms and systems in place (nor have the budgets to re-platform) or are wedded to martech decisions made at a global level, with each region forced to adopt tools they feel aren’t fit for purpose. 

So perhaps now more than ever, businesses are turning to CRO and experimentation.

So what are some of the basic tests to run to get started? CRO is a journey. While the most sophisticated of companies have deeply ingrained experimentation into the culture of their business and as a constant stream in their product roadmap, we believe that for businesses new to CRO, running a few tactical experiments is the best way to get started. 

Here’s a summary of some of the common ecommerce pain points we see and areas to consider testing around:

High cart abandonment rate

Many merchants experience high cart abandonment rates, where website and app visitors add products to their cart but leave without completing a purchase. This could be due to a number of reasons including unexpected costs, a frustrating checkout process, lack of trust or even security concerns.

This can be frustrating after all the effort (and spend) for customers to drop out at the final hurdle.

The types of test to run: 

• Showing customer testimonials and reviews as part of the checkout flow can be a great way to demonstrate social proof and ease the mind of an undecided shopper.
• Streamline the checkout process and reduce friction where possible e.g. remove unnecessary fields on form fills.
• Consider experimenting with upsell offers at checkout and the presentation of these across digital products.
• This one might sound counterintuitive, but reducing the options customers have at checkout – is it essential to have five different delivery options, for instance?
•  

Growing new customer acquisition costs

The cost of acquiring new customers via paid media is increasing. This creates more pressure on digital product owners to convert engaged customers, increase average order value and drive ongoing value from existing customers.

One of the biggest challenges is getting the right traffic to your site, yet the onsite experience thereafter is often neglected. The best performing experiments incorporate personalisation into their variants. A recent study by Optimizely found that personalised experiments generate 22% higher uplifts on average.

The types of test to run: 

• Consider testing a subscription or loyalty program for your customers – delivering value to them and recurring revenue for your business.
• Don’t neglect the ‘softer’ metrics that come before key conversion points. For instance, optimisations to the navigation and site search carry some of the highest expected impact (win rate multiplied by anticipated uplift) of any metric.
• Segment your audiences and personalise their experience not just on demographic data such as location or age, but by behavioural data to differentiate your experience for new versus returning users or through triggering a new experiment based on a previously viewed product category.
•  

Compete over online buying experience

Competition in ecommerce is fierce. There are always going to be merchants out there who offer products at a lower price or have a monopoly over key search terms. One area brands can compete over is online buying experience.

It’s important to get the foundations right first, regularly conducting performance audits to determine whether your site speed is up to

The types of test to run: 

• Do your research. Ok, so technically not a ‘test’ but conducting user research to get a head start on your customers’ needs and running user testing to understand how they actually interact with your products and their frustration points can be a vital source of new hypotheses to test.
• Be bold. This is the time to take a leap with a few riskier tests to a small group of users before rolling out to all users or committing time and resources into hard coding new features that aren’t useful to the user and don’t encourage conversion. Experimentation as a form of risk mitigation was a central topic at our recent CRO webinar with senior leaders at Currys, NatWest and Mazda.
• Test outside of pure revenue driving experiments. For instance, consider running tests that enable you to learn more about your customers, quantify risks and a data source to validate or debunk assumptions and to make more informed business decisions.
   

Discover how CRO can help your business to grow in our latest report, Getting the most out of CRO.