Brands have an ‘insight blind spot’ when it comes to understanding user behaviour on third party Marketplaces like Amazon, eBay and the plethora of online third party ecommerce sites where they could have a presence.

Sometimes the only data they get from the Marketplace is absolute sales or maybe a single conversion metric. Meaning they can’t answer those questions, that are crucial to increasing sales, such as:

  • Which product images are most appealing to shoppers?
  • How should we improve our product descriptions?
  • How can we improve the naming of our products to appear higher in search?
  • Why do customers choose competing products instead of ours?
  • Where should we advertise on this site, and with which creative?

The business impact of not having answers to these questions is three-fold:

  1. Brands revert to using guesswork, rather than insight, to optimise Marketplaces, wasting time and energy on initiatives that do not improve sales, or might even decrease them
  2. Budget is wasted promoting products on Marketplaces as a temporary band-aid to drive sales
  3. Poor sales on a Marketplace might lead to a premature withdrawal or scaling back where simple optimisation improvements could make this channel a success

But how can brands pinpoint what to change, when they have next to no insight on the customers’ needs, expectations and experiences?

They could start by following industry best practice. But these types of recommendations are made without the context of specific products and are generic in their nature. For example, Marketplaces often recommend that brands implement videos on product pages, but there’s little detail on what should be in those videos, and what creative treatment, style and content would maximise conversions.

Best practice recommendations are often too broad to be truly actionable.

So, with Marketplaces providing limited insight and best practice that’s too generic, brands are increasingly turning to UX research platforms to generate insight to inform their optimisation strategy. These platforms (like UserZoom) capture the data and insight on third party sites through ‘testing those sites with target customers who are set a specific mission.

There are numerous ways that UX research platforms can generate the insight that brands need – below are three simple steps that most brands follow when they get started with the UserZoom platform (with Amazon as the example Marketplace and Nutribullet as an example brand).

Step 1: Understand how people find products on Amazon

This simple but highly effective study generates data to identify how to optimise product names, descriptions and categories to ensure your products not only appear prominently, but are described in a way that makes them compelling.

Target customers are given a mission to find products on Amazon and their sessions recorded revealing:

  • The search terms people use
  • Why they select one search listing over another
  • The categories they expect to find your product in
  • Reactions to sponsored listings

Step 2: Observe people shopping on Amazon, and hear what they’re thinking

This qualitative study identifies those factors that influence customers during their Marketplace shopping journey, by allowing you to get a deep understanding of customers’ thoughts.

Target customers are given a mission to find a product on Amazon and add it to their basket, while saying what they are thinking as they do so (recorded on video via the UserZoom platform). This reveals:

  • The product information they need to make a purchase decision
  • What they think of product images
  • What trust factors influence them when selecting a product
  • Reactions to reviews.

Listen to a consumer giving spoken feedback on Nutribullet product pages as they decide which one to buy:

Step 3: AB Test product pages before they go live to find which is most effective

Insight from Steps 1 and 2 will generate a series of ideas to improve product imagery and descriptions on Amazon. But, before investing time into wholesale changes it’s prudent to test new variants with real customers to find which us most compelling.

To test different copy approaches and/or alternative images, run an AB Test on two (or more) mocked up product pages and ask target market users to

  • Click on what’s compelling
  • Click on what’s confusing
  • Say what’s missing
  • Rate each variant.

Running this type of study is like a pre-live AB Test that can be run at scale before investing time and effort into changing product pages. It’s a sure-fire and very efficient way to identify which variant of copy and images is most effective.

Version A:

Version B:

How to get started?

We recommend starting with one Marketplace and one product category at a time since customer behaviour can vary hugely between categories. Get in touch if you would like to see a sample study.