We compiled background information on intelligent personalization in omnichannel retail for you. We explain central concepts and approaches and how our AI-based personalization solutions come into play.
LEARN MORE ABOUT INTELLIGENT PERSONALIZATION
Where would you like to start?
What is personalization?
What is real-time personalization?
How does personalization benefit retailers?
What does omnichannel retail mean?
What touchpoints can be personalized?
What content or elements can be personalized?
What types of recommendations are there?
Concerning AI in marketing: What does AI have to do with personalization?
How can I assess the quality of my personalization?
In just two minutes, learn how our personalization solution helps you provide customers with the right deal at the perfect time and the right place every time.
What is personalization?
In the retail business, personalization refers to the user-specific customization of content at every touchpoint. However, personalization can only succeed when you, the retailer, have enough data. Only then can you identify your customers’ individual preferences. You can gather this data, for example, from previous visits to the online shop, from customer master data or from current behavioral data. The personalization solution collects the data and evaluates it in real time. A typical application is the personalized adaptation of recommendations on the detailed product page in the online shop.
In our role as solution provider for omnichannel personalization in the retail business, we look at the topic as a whole: It is not simply the generation of product recommendations that is important but rather the entire personalization strategy across the various touchpoints.
Advantages for retailers und customers: From the retailer’s perspective, personalization aims to maximize customer value. You, the retailer, use personalization to optimize sales, turnover and earnings throughout the entire customer life cycle. But personalization also provides customers with significant advantages. Customized content increases customer-specific relevance. This relevance leads to awareness which in turn results in higher click and purchase rates. If customers feel good when they are with you, they will also look forward to coming back. Goal achieved: stronger customer loyalty, greater purchase frequency, bigger shopping baskets!
In other words: Personalization in retail means offering every customer the perfect deal at the perfect time and via the right channel.
What is real-time personalization?
Let’s start with a little example to illustrate what makes the real-time aspect so important. Let’s say Max buys a Smart TV in the online shop. What happens the next time he gets a customer newsletter or visits the online shop again? Look at the following two scenarios and decide which one makes more sense:
- In the newsletter and on the homepage of the online shop he sees recommendations for Smart TVs.
- He sees recommendations in line with his current buying behavior, e.g. a sound system or various cords.
Scenario 1 illustrates nicely that recommendations made on the basis of purchase history are often irrelevant. It is highly unlikely that Max is going to buy another fridge in such a short period of time. In Scenario 2, artificial intelligence (AI) identifies Max’s current needs.
Real-time personalization solutions analyze the search and buying behavior of customers starting at the first click, adjusting immediately to the new situation. Similar to a chess computer, AI calculates all of the possible “moves” of the visitor and generates the best suited recommendations.
Actual and expected buying behavior: If a female customer, for example, purchases a gift for her partner, the actual buying behavior deviates from the expected behavior. Real-time solutions respond immediately to each change in buying behavior with relevant recommendations. They also recognize that the female customer found the gift and has now turned to the next shopping goal.
How does personalization benefit retailers?
Personalization strengthens customer loyalty because it only generates content relevant to each individual customer. Ultimately, this leads to more sales for the retailer. En route to more sales, personalization impresses in three areas:
- Better usability of the online shop thanks to fewer clicks: Adapt the navigation of your online shop to the habits of your customers. This reduces the number and complexity of your navigation menu, offering a better overview. The customer finds his way around quicker and has to make fewer decisions. There have been scientific studies looking at how many clicks people will actually tolerate to achieve a specific goal. On top of the number itself, it takes into account the “effort” required for each individual click: “How much do I have to think? “Did I make the right decision?” You can find out more about this topic in Steve Krug’s “Don’t make me think”.
- Increased awareness leads to higher click and purchase rates: Appropriate deals that reach your customers at the right time and via the right channel are effective because of their relevance. It could be a product recommendation in the online shop or editorial content in the newsletter. Or, you can send a push notification with a personalized coupon to your customers’ cell phone. The range of applications is huge and there is no limit to your imagination.
- Automated personalized content results in lower costs: It goes without saying that with a good team you could manually create 100 relevant newsletters a week and send them to handpicked customers. The same goes for generating banners on the website or for digital signage solutions in the branch. However, automation saves enormously when it comes to both time and cost – and increases customer-specific relevance in the process. For example, AI automatically creates the right newsletter for each customer and sends it out at the time best suited to that customer.
What does omnichannel retail mean?
Wikipedia defines omnichannel as a “business model and content strategy that companies use to improve their user experience. Omnichannel is an integrated way of thinking about people’s relationships with organisations. Rather than working in parallel, communication channels are designed to cooperate and build a coherent, evolving, cross-channel experience.”
Omnichannel means that customers switch between different channels, e.g. store, online shop, mobile shop or app, call center, teleshopping, social media, catalog or even field staff, sometimes within a single buying process.
Channels are often also referred to as touchpoints. But touchpoints do not have to be sales channels, but “media” where customer and company “touch”. A good example may be the box, with which goods are delivered from company to customer. This ist clearly a touchpoint, which may be personalized, but no sales channel.
In omnichannel retail customers may choose their favored channel for every buying process. Omnichannel retailers are challenged to “talk” to their customers across all channels and personalize the complete customer journey.
Synonymously you often read the terms multichannel or crosschannel, but strictly speaking they are levels of development from separeted sales channels towards true omnichannel retail.
What touchpoints can be personalized?
Our personalization solution, the prudsys Realtime Decisioning Engine (prudsys RDE for short), is capable of maximizing customer value across all customer contact points – in real time. This presumes that all touchpoints are based on the same data basis. The more successful you are at approaching customers across channels, the more customer value goes up.
- Consultant-tablets at the point of sale (POS)
- Call centers
- Digital signage (e.g. smart fitting rooms or in-store TV)
- E-readers (e.g. Tolino)
- Receipts (e.g. through coupons)
- Newsletters
- Online shops
- Package inserts (e.g. catalogs)
- Shopping apps
- Smart TVs
What content or elements can be personalized?
Personalization in the retail business is often reduced to product recommendations. A personalized sales approach goes far beyond that:
- Navigation: Show each customer the navigation elements best suited to their behavior. Your visitors will reach their goal faster and more easily.
- Banners: Energize the display of banners that adapt to each customer’s behavior and always show exactly what interests the customer at that moment.
- Rebates & coupons: Generate strategic rebates for customers to whom those rebates are currently relevant. This increases the probability of purchase and avoids scattering losses.
- Content: Generate editorial content such as guidebooks, checklists or videos that match each customer’s behavior perfectly. Your customers will consider suitable information, especially online, to be a service. It could be styling tips for a specific outfit or a tip about washing instructions – use relevance to score big here too.
- Brands: When it comes to customers who are particularly loyal to brands, show them products from their favorite brands first or invite them to discover similar products from other brands. Some shops use so-called product feeds to do this. Customers themselves control the content of such feeds by marking their favorite brands and excluding unwanted brands.
- Products: Recommend the right product to each customer at the perfect time and via the right channel.
What types of recommendations are there?
Recommendations can be classified using so-called recommendation logics. prudsys RDE differentiates between five recommendation logics:
- Product-specific: Recommendations that are related to one or more products in which the customer is interested.
- Category-specific: Recommendations related to the category in which the customer is interested.
- User-specific: Recommendations related to the user’s actual buying behavior as well as to “historical” user data e.g. purchase history.
- Search term-specific: Recommendations related to the search term the user entered to initiate the search (e.g. in the online shop).
- Global: Recommendations related to the entire range of products.
The recommendation logic describes the recommendation’s frame of reference. You often see different types of recommendations at certain points in the online shop. “Other customers also purchased” shows product-related recommendations while the title “your personal recommendations” refers to user-specific recommendations.
Recommendation logic or recommendation type: Sometimes these terms are used synonymously. Strictly speaking, however, they are not the same thing. The prudsys RDE offers more than 20 types of recommendation. Each recommendation type is assigned to one of the above mentioned recommendation logics, which are different depending on their degree of personalization.
Three examples:
| Recommendation Type | Description | Recommendation Logic | Degree of Personalization |
|---|---|---|---|
| „Ordered together“ | Recommendations of products purchased together with the respective product | Product-specific | + |
| „Global top seller“ | Recommendations of top sellers from the entire product range | Global | – |
| „User to products“ | Recommendations based on the transactions of the respective user | User-specific | +++ |
Concerning AI in marketing: What does AI have to do with personalization?
Think about chess for a minute. This game is perfect for explaining how the prudsys RDE functions in simple terms: The software recognizes the customer’s “moves,” predicts the chain of all of the possible subsequent events and acts accordingly (e.g. product recommendations) so that it wins “the game” (e.g. sales optimization target parameter).
The more intelligent the personalization solution, the more successful it is. The prudsys RDE learns new personalization rules automatically and in real time. This takes place based on clicks, shopping baskets, purchases, external and internal search queries, clicked on categories and banners as well as self-selected events (= machine learning). The prudsys RDE continuously and independently collects the necessary data. The AI recognizes patterns and regularities in the data and can apply its knowledge to unknown data as well. That means that personalization is possible right from the first click.
The prudsys RDE algorithms recognize not only the customer behavior and respond to it (e.g. with individual recommendations), but they also measure the acceptance of the generated content at the same time (= reinforcement learning). When a customer accepts (i.e. clicks on) a generated recommendation, the system rewards itself. If the recommendation is not accepted, there is no reward. This way, the prudsys RDE gets to know its customers better with each move and can tap into the individual preferences of each customer in a more strategic way.
There’s more: The prudsys RDE can deviate from learned patterns by way of so-called exploit and explore mechanisms, e.g. to recommend products or content that has never been recommended or has not been recommended for a very long time. The AI ensures that its scope of learning and action remains flexible by trying out alternatives or new variants. For your customers, that means that they will always see new content and products. They are not stuck in a filter bubble. This is a great advantage, especially for quickly changing product ranges: Products from new collections or brands flow automatically and immediately into the recommendations.
An intelligent personalization solution evaluates large volumes of data in real time, recognizes patterns and performs the “right” action in order to reach the corporate goals of the retailer.
How can I assess the quality of my personalization?
There are different approaches when it comes to expressing the quality of your personalization with a KPI:
- Sales through recommendations
- Acceptance of recommendations (clicks)
- Return on Marketing Invest (ROMI)
None of these KPIs measures the quality of personalization independent of external influencing factors. The product itself, the personalized channel or even seasonal effects influence the KPIs mentioned and introduce disruptive factors that make it difficult to perform a valid evaluation.
The “prudsys personalization index” (PPI) is an independent maturity model that evaluates the quality and completeness of your personalization strategy. Let us calculate your index value together to establish a base value for your KPIs. Implement various measures of your personalization strategy and calculate your index value regularly. This gives you a statistically valid figure. You can also use the PPI value to measure the growth of your personalization strategy.
Read above to learn about the advantages of a well-functioning personalization strategy. Take a look at our website for a detailed explanation of the “prudsys personalization index”.
White paper personalization in fashion retail
This white paper shows you the benefits of a digitally managed customer journey, describes successful application scenarios, and explains how to use the Diderot effect in your sales strategy. You will also learn about the technological fundamentals of real-time personalization and how to determine the status quo of your personalization activities.
Mission
We automate personalization and pricing processes in retail by means of artificial intelligence (AI).