Creating personalized customer experiences with one-to-one marketing in e-commerce

Two types of one-to-one marketing

The openPR wiki distinguishes between two main types of one-to-one marketing: adaptive marketing and personalized marketing.

Customization marketing encompasses measures that enable customers to customize products or similar items according to their own preferences. In this case, the shop operator offers its customers the opportunity to tailor a product to their specific tastes. The openPR Wiki cites the computer company Dell as an example of this, which meets specific customer requirements by offering customers the opportunity to customize the computers they purchase.²

Personalized marketing, on the other hand, involves personalization measures based on information about a customer or their behavior. In this case, the seller or shop operator finds out what the customer's tastes and personal preferences are and tailors their content (product suggestions, etc.) individually to each individual.

In the following, we primarily understand one-to-one marketing to mean personalized marketing.

Personalization as an important element of one-to-one marketing

One-to-one marketing is therefore about using personalization to display individual content to customers. But what exactly does the term personalization mean?

The Gabler Business Dictionary describes it as the adaptation of website content to the respective user (based on information collected about user behavior).³ Personalization therefore involves recording the visitor's previous behavior in order to display exactly the content they want at the right time. This allows online shoppers to be picked up at the stage of the digital customer journey they are currently at. Shop operators can thus use personalization to create a shopping experience for their online shoppers.

It is important that personalization is carried out in real time, so that content is only generated when the visitor requests it, e.g., by opening an email or clicking in the online shop. This allows current behavior and all up-to-date information (such as item availability) to be included in the calculation.

Technical background

But how does it all work? By using software services that record online shoppers' clicks and purchases with the help of self-learning algorithms and deliver personalized content—automatically and in real time. This reduces manual effort. To do this, the integrated algorithms use the concept of reinforcement learning, which originates from machine learning and enables self-learning optimization. This is based on a knowledge base that is generated individually for each online shop. This enables the algorithms to provide 1:1 communication at the various touchpoints in the online shop. In a sense, every online shopper finds their own individual shop. They are shown exactly the products and content that match their own preferences. This is achieved by making predictions about future interests based on a customer's past clicks and purchases.

Would you like to learn more about the knowledge base? Then we recommend watching our webinar recording on this topic.

Search function as a personalized guide

According to the E-Commerce Guide, 80 percent of online shoppers use search almost exclusively to find the product they want.⁴ This clearly demonstrates not only the relevance of suitable search functions, but also the need for a shopping experience that is as convenient and guided as possible.

A user-friendly onsite search function provides guidance and helps shop customers find the product they want quickly and easily without leaving the site. To do this, the search function should be error-tolerant, support the entry of search term combinations or item numbers, and take language usage into account. A type-ahead function also makes it easier for shop customers to find suitable products.

In the context of one-to-one marketing, search suggestions in Autosuggest and on search results pages can be personalized. This means that the results are tailored to the visitor's behavior and take their individual preferences into account. Standard product searches, which are not customizable, cannot implement these individual circumstances and, compared to an intelligent search, do not deliver one-to-one marketing.

Personalized search results for two customers

Advice as in brick-and-mortar retail

Not every shopper knows exactly what they want from the outset—some are unsure, while others want to explore different options. This is where the AI Shopping Assistant takes on the role of a personal advisor and guides the user through the selection process.

The dialogue-based product advisor analyzes visitor input and behavior in real time, identifies their needs, and filters out suitable products from the range. It processes natural text and voice input in the chat, accurately captures customer needs, and quickly and reliably guides each user to the right product. This conveys a positive feeling of "being well advised" and at the same time reduces the return rate.

The advice is tailored to the individual store and product range. Questions, filters, and recommendations adapt dynamically to the product selection as well as to the behavior and preferences of the users. This creates an advisory experience that is comparable to the personal service offered in brick-and-mortar stores.

As part of one-to-one marketing , the AI Shopping Assistant enables personalized product advice tailored to each individual customer, creating a customized, positive shopping experience.

AI shopping assistant in an example online wine shop (source: own illustration)

Recommendations for individual purchasing impulses

If the visitor is already in the purchase decision phase, a recommendation engine provides inspiration to stimulate impulse purchases for other products in the range. A recommendation engine can inspire the buyer with suggested additional products directly when they click on the shopping cart button, put together and recommend product sets, identify thematically matching products, and suggest bundle purchases or complete packages, thereby increasing the shopping cart value.

The impulse to buy is created by displaying recommendations that match previous behavior according to specific strategies. It is important that the strategy of the individual recommendation widgets (interactive elements in the online shop that contain recommendations) is personalized for the shop. To implement this, individual rules and filter settings are required, for example. An online shop may have a rule that only recommends items from the same or a similar brand for certain brands. However, if an online shop only offers one brand, this rule is meaningless. Customized recommendation engines can also integrate additional data sources, such as loyalty programs or CRM systems, to further optimize personalization. Personalization software that works exclusively with standard rules cannot therefore reflect the individual needs and circumstances of each individual customer and thus does not deliver one-to-one marketing.

One-to-one communication also arises when shop customers are only recommended products that are actually of interest to them. In addition to personal preferences (such as brands, colors, styles, patterns, etc.), availability should also be taken into account.

Personalized product recommendations in the babymarkt.de online shop as an example (source: screenshot from babymarkt.de)

Emails as wake-up reminders

By combining relevant recommendations with newsletters, one-to-one marketing can be sent out in the form of personalized emails, which increase traffic to the online store. This means that online shoppers can also be reached on other channels outside the online store. Personalized emails deliver individual purchasing incentives via email, guiding the recipient back to the online store.

Customers receive products specially selected for them and are informed about current trends and news. In addition, customers who have been inactive for a long time can be encouraged to return to the shop via a "We miss you" email. The content is tailored to the individual customer, enabling one-to-one marketing at this touchpoint.

Stay up to date on personalization: Sign up for the epoq newsletter. Register now!

In addition, emails can be sent on specific occasions, such as the customer's birthday. A child's life cycle can also be incorporated to offer clothing items that are appropriate for their age and size. For industries where prices change quickly, such as airlines or booking portals, prices can also be dynamically displayed in the email when it is opened. You can find more information on this in the blog article Dynamic Pricing: Curse or Blessing for E-Commerce.

Example of a newsletter from babymarkt.de with personalized recommendations for various categories (source: screenshot from newsletter from babymarkt.de)

All of these are measures for personally addressing customers through one-to-one marketing, based on the clicks and purchases of online shoppers. Overall, both the customer, who feels well taken care of in the personalized shopping experience through 1:1 communication and finds exactly what they are looking for, and the shop operator, who can achieve a significant increase in the conversion rate, benefit from the unique customer experience throughout the entire customer journey and automated one-to-one marketing.

...Include time span for repeat purchases

The likelihood of repeat purchases or the time span until the same or similar products are purchased again plays a central role. It can vary greatly depending on the industry or product range. Here are a few examples:

• Technical devices and furniture usually have a long lifespan. In this case, care must be taken to ensure that the same or similar items are not recommended for a long time in order to avoid recommendations such as "Customers who bought this washing machine also bought this washing machine." If a customer has ordered a garden shed, for example, it makes little sense to offer them more garden sheds in a follow-up email. Instead, additional items such as wood paint, a barbecue for the next garden party, or similar items are more appropriate at this point.
• In the area of clothing, books, and other media, there are basically no limits here. If a shop customer has purchased a book by an author, for example, they are likely to be interested in other books by the same author (i.e., similar products). However, it is important to note that in this example, the same book (hardcover) should not be recommended in a different format (e.g., as an e-book or paperback).
• For consumer goods such as drugstore items or food, the duration is quite short. In this case, customers develop preferences for certain products, so it can make sense to recommend these products again and again. Although no new items are recommended, the recommendation increases the likelihood that the customer will buy the product in your shop and not from a competitor.

...take into account the specific characteristics of the industry and product range

Every industry and its products have their own challenges when it comes to sales and personalization. Depending on the industry, there are specific characteristics that should be taken into account for personalized communication. Here are a few examples:

• In the fashion sector, size is of crucial importance. This should be taken into account in all product suggestions so that customers are not recommended products that are not available in their size. The current season should also be taken into account so that, for example, swimwear is not recommended in winter.
• Online shops that sell children's products should take into account not only the size but also the age of the child. These characteristics change continuously (age) and sometimes rapidly (size). To avoid a mother of a 5-year-old being shown baby products, these factors must be included in the personalization.
• When making cross-selling recommendations for technical items, you must take special care to ensure technical compatibility. So make sure that any additional items offered (covers, charging cables, etc.) are compatible with the specific smartphone model.
• In online shops that specialize in gift items, shop operators must be aware that customers usually do not shop for themselves, but for other people (sometimes several different people). In addition, the interests expressed through purchases are subject to cycles (Easter, Christmas, birthdays). To ensure that the right products are displayed at the right time, this must be taken into account for one-to-one marketing.

...align strategy with the respective customer journey phase

The customer journey phase and the touchpoints at which one-to-one marketing is used are also of great importance for the strategy.

As already described, an intelligent search function is helpful in the product research phase, for example. This is useful for every shop, as it provides orientation and guides customers quickly and easily to the desired product. During the product selection phase, however, good advice is crucial. Depending on the type of product (products that require explanation, such as suitcases, running shoes, or bathtubs, vs. self-explanatory products, such as clothing, books, or furniture), different strategies can be pursued.

In the purchase readiness phase, you can provide inspiration by using recommendations on various shop pages. Here, too, very different strategies can be applied. While it makes sense to display alternative products on product detail pages to help customers choose the right product, you should definitely refrain from offering alternatives in the shopping cart so as not to unsettle shop customers in their choice at this point. Here, additional items that can fill the shopping cart are suitable. But there are also differences in the recommendation strategy. For example, you can simply display cross-sell items in the shopping cart of a fashion store. Alternatively, however, it would be better to select cross-sell items from specific categories that go well with the outfit, taking into account the visitor's clothing size. A sales assistant in a brick-and-mortar store would not recommend an item that is no longer available in the right size.

Good personalization therefore requires customizable software that can be adapted to the various conditions of the online store. This is the only way to achieve successful one-to-one marketing for every store.