E-commerce is growing at an unprecedented rate all over the world and giving rise to new opportunities for brands to be competitive. Market research in the online domain is still in its early stages, with brands often opting for a much wider variety of research questions and a lack of focus in comparison to brick-and-mortar studies. Although all research questions are legitimate, clients tend to fall into the trap of wanting to test everything; from strategic to tactical, online path to purchase to specific content, and other. The result of which is too broad, only scratching the surface on many KPIs, with very little actual impact.

So, how should you approach structuring the questions for more effective research? In general, we find there are 3 levels of studies:

• Level 1 – online shopper strategy including motivations and full journeys (who, which sites visited, and why)
• Level 2 – online shopper behavior for specific website: how do shoppers navigate the site; which pages are visited; and what are key shopper marketing assets
• Level 3 – Tactical impact studies for optimized shopper marketing assets (that apply for most websites): what are the best practices for the shopper marketing assets (e.g. online package design) to maximize the impact on purchase decisions

Typically, the 3 levels demand a different research methodology framework. Surveys will suffice for Level 1 (online shopper strategy), while for Level 2 and Level 3 (shoppers’ behavior on a specific website and tactical impact studies), a combination with behavioral methods is required. Furthermore, the Level 3 study requires a design experiment (e.g., A/B/C test of package design or ads) to understand the impact on sales uplift.

Level 1: Testing online path-to-purchase and strategy

The key questions for online path-to-purchase studies boil down to who buys online, and why? Which websites did they visit? Which phases in the shopping journey did they go through? Where do they eventually purchase, and for which reasons? Within this level, it is also possible to explore who is close to buying and what would convert them.

To dig deeper into the research questions, here’s a breakdown for an example product:

Level 2: Testing online shopping behavior

The key question is how shoppers shop on specific online retailers and what are the opportunities to influence their purchase decision. Moreover, which pages do they visit, how do they interact with them, and what is essential in facilitating a purchase decision? Typically, shoppers are not good at recalling their behavior – therefore, to obtain good data, their behavior needs to be measured (clicks, navigation, gaze…) in context.

Most frequently, we test on the key online shops such as Amazon, Walmart, and Target; for pet food producers – Chewy; for electronics and tech – Best Buy, etc. We mainly focus on 4 areas: website navigation (search/navigation/filters), product list page, product detail page, and ads. Optionally, the check out as well.

Here is what you should explore to get the most out of online sales: 

Level 3: Tactical impact studies for optimized shopper marketing assets

The key goal is to maximize the impact of crucial shopper marketing assets such as online package designs, PLPs, and PDPs. We know that getting your product image and the PDP right is vital for conversion. Once you have that under control and leverage as much as you can, then start testing ways to increase the traffic to your category online.

So, what’s the best way to approach e-commerce testing?

Firstly, pick the level of your study: online strategy, shopper behavior, or tactical impact studies. There are two ways to approach this: top-bottom (strategy to tactical) or bottom-up (making sure your ads, product image, and PDP work first).

Although it might sound logical to follow a top-down strategy, the bottom-up approach might be the best: you will have quick wins that demonstrate impact to your stakeholders, and these studies are the easiest. Certain marketing assets are essential on all sites (e.g., online package design), so you don’t need to have the info of higher-level studies. Investing this way will ensure you get results faster and provide you with leverage to dig even deeper into strategic questions.

Secondly, depending on the research level you pick, combine the right conventional and behavioral methods. The complementary mixed-method approach will provide you with +30% higher predictive insights and an understanding of the full picture.

Lastly, pick an appropriate specific retail website to test on and make sure to test in context. Being able to expose respondents to testing environments that are visually and functionally the closest to the actual retail website ensures a higher correlation with real purchase behavior.

Interested in testing your e-commerce solutions? Reach out to us via [email protected]