💁‍♀️ Merchandising and personalization for e-commerce: what works and what doesn't?

E-commerce merchandising is about the online store shelf layout: in other words, the presentation and promotion of products to influence online purchasing behavior, increase customer satisfaction and increase profits. Personalization - ensuring that the right products are shown to the right customers - is an important part of this.

In simple words: merchandising decides who sees which product and when. Sounds great, right?

How e-commerce merchandising works

Like traditional physical store merchandising, e-commerce merchandising is about shelf layout. The playing field of e-commerce merchandising is therefore the product lister page (PLP) and the search results page. And although the physical shelf layout is static, it is actually dynamic in (many) web shops. In other words: the shelves in the store look the same for everyone, but this does not have to be the case online.

This gives us an immediate grip on the mechanism of online merchandising. This consists of three parts:

1. Product data
2. Behavioral data
3. Rules

Let's discuss them briefly.

Product data

Product data refers to all information and details related to a particular product. Think about the product name, description and price. But also color, dimensions, margin, delivery time and a whole range of category-specific features (e.g. filling weight, speed, noise level, centrifugal energy consumption).

Behavioral data

A customer visit to an online store consists of viewing one or more pages, which can be products (product detail pages), product categories (product lister pages), inspiration pages (content pages), the homepage, checkout, search terms and results, and more. And on these pages, the customer also shows' behavior ': images are viewed, products are clicked on and compared. All of this results in first-party data: a digital crumb trail that a customer leaves behind when they visit a shop.

Rules

With rules, you link product data and behavioral data to each other, in order to fulfill the promise of a truly personal shelf layout. In other words: What behavior results in which products shown?

There are two kinds of rules (just ask your kids):

• The generally applicable rule: 'the PLP must always sort by the most recently added products” or “the search results should always be the products with the highest margin show first.”
• The exception: 'the PLP must always sort by the most recently added products, with the exception of Product X that I always want to show in first position”.

These are both examples of generic merchandising. Here, a rule is created based on only product data to arrive at the desired store shelf.

When a rule uses product data one behavioral data is there personal merchandising. For example:

• The generally applicable rule: 'the PLP products must be ranked based on brands previously visited by the customer'.
• The exception: 'the PLP products should be ranked based on brands previously visited by the customer, with the exception of my own brand, which I always want to show in first position'.

On the one hand, the goal is to create a personal e-commerce experience that helps customers move forward, while on the other hand, the commercial goals of the webshop are achieved more quickly.

Why personalized merchandising is ineffective

The promise of personal merchandising is appealing: an e-commerce experience for everyone that meets their needs. However, there are two irrefutable reasons why this does not work.

1. Customers don't know what they're looking for

Let's say you're planning to go camping with your partner and are looking for a tent. But you don't know about tents. So you land on the category page and click around: a tunnel tent, an inflatable tent, a dome tent, a family tent. So what am I looking for?

The doubting customer, who is the overwhelming majority, does not know which product suits him and what to buy. As a result, the data he produces (e.g. in the form of the products he visits) is an inaccurate source on which to base your merchandising. And garbage in, garbage out.

2. Good personalization requires a lot of data

To properly personalize a customer journey, you need as much data as possible. Twenty data points tell more than two. Two hundred data points, more than twenty.

However, this is a catch-22: to help customers properly (through personalization), you need a lot of customer data. But by definition, helping well also means efficient help: customers want from A as quickly as possible (” help, I need something!”) to B (” top, I want this product!”).

So this doesn't work.

Personalize by need, not behavior

Personalized merchandising is the wrong solution to the right problem.

382 benches, 1,992 headphones, 82 drills, 3,125 cases, 98 PVC floors, 394 curtains, 863 laptops... The product lister page (PLP) is indeed abundant, incomprehensible, and impersonal. This causes choice stress and uncertainty for customers, causing them to delay and cancel purchases. As a result, web shops lose turnover every day.

But why choose to personalize based on behavior? What is the added value of still letting go (read: leaving them to fend for themselves) in the store, 'secretly' monitoring their behavior and then making assumptions based on a crumb trail that consists of imperfect data?

Why choose this cumbersome and imprecise way of personalization? You're marginally improving an imperfect status quo. You're still pushing products, but in a so-called personal order (with the message still being: “figure it out for yourself”). The actual customer pain will not be addressed, let alone resolved.

Customers are not interested in your dozens, hundreds, or thousands of products.

The only number that matters to the customer is 1:

• The one product that suits me.
• The one product that matches my intended use, situation and wishes.
• The one product that suits me so well, so price is less important.
• So the one product that I don't have to return.
• The one product that makes me so happy that makes me come back again.

In other words: personalizations must respond to the actual need of the doubter. Not on their click behavior. And you can identify that need in a fairly simple and direct way: ask for it.

Start a conversation with the customer and offer them a short cut to his best choice. Exactly this piece of real personalization (what we still know from the offline store) is what e-commerce misses — and what merchandising doesn't solve. We call this guided selling, and you can still do it today get started.

‍