62% lower bounce rate through the use of intelligent search at NKD
The use of intelligent search in NKD's online shop had a positive effect on the performance of the online shop and significantly reduced the bounce rate by 62%.
An intelligent search function helps customers find their way around your online shop and reduces the bounce rate
Did you know that...
75%
of online store customers consider a good search function to be particularly important. (Source: Statista, 2017)
93%
of consumers report high customer satisfaction after using an intelligent shop search. (Source: E-Commerce Guide, 2016)
66%
of consumers want search functions and filter options in online shops. (Source: Statista, 2018)
80%
Users almost exclusively use the search function to find the product they want. (Source: E-Commerce Guide, 2015)
Intelligent search paves the direct path to the product in the awareness phase
An intelligent search function provides guidance during the awareness phase of your online shop and helps your customers find the right product quickly and easily. Search results and category listings are calculated in milliseconds and displayed in a personalized manner. In addition, an intelligent algorithm recognizes errors when entering search terms, corrects them, makes suggestions during the search entry, and also allows product characteristics to be entered. At the same time, targeted merchandising can be implemented flexibly in the online shop – for a more efficient start to the shopping process, a lower bounce rate, and faster checkout. That's exactly what Epoq Search offers.
Personalization plays an important role in relevant search results and category listings
Intelligent search is part of our personalization platform, which is powered by an AI engine. It uses artificial intelligence to calculate relevant search results for each individual customer. This allows it to quickly and easily guide your shop customers to the product they want and enable them to start the shopping process without any detours. The search results list shows exactly the products that match your shop customer's personal preferences at that moment. For example, after a customer has entered "pants" into the intelligent search, pants from the "Tom Tailor" brand in size 40 are displayed first. Why? Because the intelligent search has detected that your shop customer has already searched for the "Tom Tailor" brand, for example, using the faceted navigation of a category.
Watch our short explanatory video on intelligent search
You are currently viewing a placeholder content from Default. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More Information20% more sales through intelligent search in the online shop
"With the help of epoq, search results can be improved much more easily and quickly. This has enabled us to significantly increase our sales."
Dr. Marc Wagner | Data Mining Manager | Villeroy & Boch AG
Merchandising control: Your business logic in intelligent search
Our intelligent search is tailored not only to your users' interests, but also to your business requirements. With flexible merchandising options, you can specifically control which products, brands, or categories are displayed and highlighted in the search results. This allows you to strategically promote important items, boost sales of remaining stock, or optimally present new products – tailored to your product range strategy, current campaigns, and KPIs.
Prioritize
Determine which products, brands, or categories should be displayed preferentially and highlighted specifically.
Business Objectives
Define which business objectives – such as inventory clearance, margin strength, or new products – should be taken into account when displaying products.
Rules
Define which rules should apply to specific use cases or target groups.
Use our customer targeting for intelligent search campaigns
Our intelligent search thinks not only in terms of products, but also in terms of target groups. You decide who sees what: personally, segmented, or according to strategic priorities.
1:1 Personalization
Our core competence is 1:1 personalization – we offer every user tailored results in real time. From the very first click, we analyze behavior and immediately deliver relevant search results, without the need for an existing profile. For returning customers, we take previous search queries and interactions into account to tailor the results even more precisely to individual needs.
Segment-based Targeting
Segment-based targeting specifically addresses groups with similar interests, such as "outdoor enthusiasts" looking for hiking equipment. Thanks to 1:1 personalization, segments are automatically created from individual click and purchase data, so that search results are even more precisely tailored to customer needs.
Product-based Targeting
For a single product – such as a newly released book – as well as for a category or brand, users who are identified as potential buyers for that specific book based on their click and purchase behavior are targeted. Thanks to 1:1 personalization, these target groups can be automatically derived from individual data, ensuring that search results are tailored to individual needs with exceptional accuracy and effectiveness.
The most popular use cases for intelligent search
Don't let negative emotions arise and keep your customers in your online shop
Zero-hit pages lead to negative emotions, which can cause customers to leave the online shop. Typos are often the cause, leaving the searcher in the dark. Automatic detection of input errors and the opening of a preview window allow an intelligent search to efficiently guide the visitor and improve usability at the same time.
- Negative emotions are avoided
- User experience is improved
- Features in use: Error Tolerance and Autosuggest
Enable your shop customers to enter product attributes in the search field
If an online shopper searches for "small blue evening bag," this is a search term combination that includes characteristics in the search field. Intelligent search captures the meaning of the search term combination, also known as semantics, and returns the correct search results.
- Customers can enter product features as search terms.
- The semantics of the search term combination are recorded.
- The relevant search results are displayed.
Ensure the correct list of results for each search term
There are different names for certain products. Shop customers may search for a laptop and enter notebook in the search function. Both terms should lead to a list of results. That is why it is important to define synonyms for search terms, for example. However, it may also be the case that there are different interpretations of a search term. That is why it is equally important that the list of results allows for alternatives, for example. In our example, notebooks would also be displayed in the results list for the search query "notebook," since notebook is also used to refer to a notebook in everyday language. An intelligent search makes exactly that possible.
- Make synonymous terms searchable
- Expand results list to include alternatives
- Exclude certain search terms
- Feature in use: Thesaurus
Make it easier for your customers to complete their purchase
In addition to product information, content pages such as legal notices, payment terms, and delivery conditions play an important role in completing a purchase. With an intelligent search function, you can guide customers to important content pages that will ultimately convince them to make a purchase.
- In addition to product information, content pages can be made discoverable.
- Customers receive important information beyond the product itself
- Features in use: Link Finder
Draw your shop customers' attention to interesting items
The intelligent search function allows actions to be highlighted in the form of banners above the results list. The banner can, for example, highlight certain brands to boost sales or draw attention to discounts. Banners can also be displayed in the preview window, guiding customers to the right product before the results list is displayed.
- Visualization of campaigns and promotions
- Sales promotion of specific brands or products
- Use via result lists and in the preview window
- Features in use: Bannermanagement
Offer multiple ways to find the desired product
In the B2B sector, people often search for item numbers. Intelligent search can be designed for this purpose and can detect and correct input errors in the item number field. This allows customers to find exactly the products they are looking for.
- Expansion of the search function to include item numbers
- Detection of input errors
Make it easier for your shop customers to find products
Once the results list for a product search has been displayed, it can be narrowed down click by click. For example, by selecting the color, size, brand, etc. in the faceted filter. This means that online shoppers only see products that match the selected criteria. The faceted filter can also be positioned as desired and customized.
- Narrowing down search results click by click
- Various product features can be offered for this purpose.
- Only products with the respective features are displayed.
- Features in use: Faceted Navigation
Offer your shop customers the opportunity to incorporate their preferences
After the results list has been displayed and the faceted filter has been used, if applicable, online shoppers still have the option of sorting the results list according to their preferences, e.g., by relevance, ascending or descending price, or sale items. This makes it easier to find the right product quickly.
- Sorting according to your own preferences
- Various parameters can be defined
- Find the product you want faster
The AI engine of our personalization platform makes it possible
Before intelligent search is implemented in your shop, we first use our AI engine to generate a knowledge base for your online shop. To do this, our AI engine uses your product catalog, the click and purchase behavior of your shop customers, and our and your expert knowledge. It links the data to knowledge and thereby calculates the relevant search results for your shop customers and displays them. Since the intelligent search is located on our personalization platform, which is designed to personalize the entire customer journey of your shop customers, you have the option of activating additional touchpoints in your online shop and using the knowledge already generated by the intelligent search (modular structure) to further enrich the knowledge base for your online shop. Would you like to learn more about the AI engine? Then head over to E-commerce Technology.
These elements should be included in an intelligent search
Error Tolerance
Consideration of spelling errors and integrated "Did you mean" function.
Thesaurus
Language usage and colloquial terms are taken into account.
Autosuggest
Intelligent product and content suggestions for search entries via preview window.
Faceted Navigation
Filter options allow customers to narrow down products and make searching easier.
Semantics
Search term combinations with properties in the search field are enabled.
Results Management
Individual products can be promoted or hidden in the results list.
Discover our insights into intelligent search
01 / 3
4.23% higher click-through rate on product overview pages thanks to self-learning ranking
"epoq's intelligent ranking helps us to always show our customers the most relevant products, despite our heterogeneous product range. This was particularly evident in the significant increase in click rates during the test."
Claas Hemmje | Team Leader Data & Shop Management | A-Z Gartenhaus GmbH
Customization is a must-have for intelligent search
Product Data
Your product catalog is cleaned up, prepared, and aggregated in terms of the desired search functionality, e.g., the color designation is included in a separate column or missing attributes are added.
Search Algorithm
The search algorithm is adjusted, for example, with regard to error tolerance, field weighting, and ranking, e.g., selection of the fields to be searched, determination of the field weighting, and setting of the sensitivity level for error tolerance.
Configuration
Both functional and visual configuration takes place. Functionally, this includes selecting filters, sorting, and autosuggest previews. Visually, it includes the graphic design of your online shop.
This is how the integration of intelligent search works
Generation of a knowledge base
To generate a knowledge base, we need your product catalog in XML or CSV format and the click and purchase behavior of your shop customers. The latter is recorded in compliance with data protection regulations via a tracking code in your online shop. The shop system is irrelevant here.
01
Setup in the test system
We discuss your visual ideas together and send you the div element as a placeholder for the front end of the intelligent search. You integrate this into your test system. Now the search algorithm is fine-tuned and the necessary configuration is carried out.
02
Activation in the live system
After setup in the test system and successful functional testing, the intelligent search is activated in your live system and is available to your shop customers as a guide to the desired product. If necessary, we will perform fine-tuning during live operation.
03
Introduction to our backend
We will introduce you to our backend, the Control Desk, and train you in the various analysis and configuration options as well as feature usage. This will enable you to monitor and control the intelligent search.
04
Display of relevant search results
Your shop customers receive relevant search results and are directed to their desired product. This helps them navigate their product search and reduces the bounce rate, which has a positive effect on your conversion rate and sales.
05
Monitoring and control of intelligent search
The Epoq Control Desk gives you extensive options for monitoring and controlling the intelligent search.
Analysis
Checking the relevance of search results for specific search terms from the customer's perspective. Relevance here refers to the match between the content of the visited page and the product or information the user is actually looking for.
Key Figures
Review of various smart search metrics, such as click and revenue performance.
Configuration
Here you can configure the current settings and other search algorithm settings, such as error tolerance, weighting of individual search fields, facet filters, and sorting.
Features
Possible uses and configuration of various intelligent search features, such as thesaurus, link finder, and banner management.
When is it worthwhile for you to use intelligent search?
Basic Functionality
You already run an online store or similar and have all the basic features covered, and now you want to optimize your standard search.
Search Queries
You have more than 5,000 searches per month in your online store.
Product Range
Your product range includes more than 500 products.
Want to guide your online shoppers straight to the right product?
Our intelligent search ensures that your
online shoppers don't bounce so easily.
Popular blog articles on intelligent search
01 / 8