Let Neocom’s Smart Algorithm become the perfect Co-Pilot for your Advisor.
Neocom has painstakingly crafted an intuitive algorithm that delivers you and your clients a shopping experience reminiscent of a real-world encounter. Neocom’s algorithm allows simple individualization that multiplies the background knowledge you already have on your topic, generating even better recommendations for your customers.
While creating your Neocom product advisor you will have plenty of levers to pull to maximize the algorithm, resulting in perfectly generated recommendations for your clients.
This Article is divided into two sections:
- Maximizing the algorithm during the questioning process
- Maximizing the algorithm across the result page
Questioning process
Filter and Ranking questions
Neocom allows you multiple options to individualize each of your advisors' questions. The most influential, apart from the Mapping, which we covered extensively in the Optimize Questioning article, is setting the importance of each question. You can choose between two levels of importance for each question, Filtering and Ranking.
Filtering Questions have the most impact on the result page, acting as a hard cut for the selection process. If a product doesn’t achieve the prerequisites of the answer, it is no longer available on the result page. Just like in real life, if you are being consulted in person, at a store, the salesperson will have certain questions which directly influence the product they are going to recommend. Think of Filtering Questions as deal-breaker questions.
Returning to our much-loved Bike Store example. Let's assume you are trying to find out if your customer wants a mountain bike or a road racing bike. Once the customer has decided on one of the two options, they are no longer interested in the other product. This allows the Neocom Smart Algorithm to cut out one whole category of bikes and direct the question flow accordingly.
Ranking Questions have less impact on the result page. Think of these questions as “Nice to Have”. There are some features that any customer, purchasing any product, will make concessions on. If a product doesn’t fulfill certain criteria for a ranking question, it still has the possibility to appear on the result page with users being notified which of the desires the product does not fulfill.
When a user goes on the Bike Store Website and completes the advisor for the perfect bike, they want to see recommendations that perfectly fit their needs. Those needs are things like Type of Bike and Size. These are things that are crucial to making a purchase decision, these needs therefore require Filtering Questions. Your customers do not want to be recommended bikes that do not fit or that aren’t designed for the type of riding they plan on doing.
These same users also have a long list of wants. These could include color, seat style, head and rear light, basket, and many more. These are things that are not crucial to their purchase decision but are nice to have and are therefore ideal ranking questions. A customer may want a red bike, but if our algorithm finds the perfect bike, in blue, it will still be shown to the customer, and maybe they make a concession.
Smart Ordering
Another individualization option you have when creating your advisor is the order in which questions are presented to customers. This is a key element to positive customer engagement. You can either set the order exactly the way you want or turn Smart Ordering on. Smart Ordering gives the algorithm free rein over certain blocks of questions. The algorithm then analyzes each customer journey individually and presents questions in a way that will get the client to the results that best meet their needs, the fastest.
Returning to our bike shop. You need to have answered a few questions to find an appropriate recommendation, mainly the type of bike and size of the rider. All the other questions you want answered by the customer could be asked in any order. If you were to put all the other questions in a group and turn Smart Ordering on, Neocom's intuitive algorithm would select the appropriate questions to display, creating a completely individual journey for each visitor.
Let’s say your customer is looking for a Road Racing bike and we know the size of the rider. The remainder of the questions are up to the Neocom algorithm to sort out. Questions like, do you need a basket? Do you need suspension? Does your new bike need a child seat? will be skipped, because no Road Racing bikes have a basket or suspension, or a child seat.
Smart Ordering can only be turned on if your questions are within a Group.
Results
Custom attribute boosting
Neocom provides the possibility to push products with certain values up the ladder of your result page directly from your feed. What seems complex is actually very straightforward. If there is any value in your Product Feed that you think needs to be highlighted for your potential customers, let the Neocom Support Team know which, and they will set up a background Custom Attribute Boosting that will push these products towards the top of your result page.
There are two common use cases for Custom attribute boosting.
Returning to our bike example. The first use case would be to push a certain Brand. Maybe the Bike Store also makes bikes, so of course, you, as a bike store owner, want to sell as many of your bikes as possible. Neocom allows you to rank your own brand higher than those of your competitors, meaning if the bike produced by you fulfills all the desires of a customer but so does a competitor’s, your brand will always appear before your competitors.
The second common use case is to push items according to actual stock levels. Let’s say you have a lot of stock of one certain bike. Neocom could create a rule that always pushed bikes to the top that fulfills the customer's requirements but also has a high quantity of articles in stock.
You can use Custom attribute boosting for a huge variety of attributes. Last Season items, products with higher margins, the list goes on.
Price Impact
Price impact is another method to individualize the result page. The Price Impact feature allows you to push products with lower, or higher sales prices to the top of your result page. If the Price Impact feature is not activated, the algorithm will automatically rank products in ascending price order.
Labels
Labels allow you to highlight certain products on the result page for whatever reason you deem as important. There are 3 default Labels available:
- Best Price-Value - Based on how many of the customers wishes the product fulfills, at the lowest price.
- Best Value - Best match according to the answers given by the customer.
- Cheapest Price - Lowest price but still fulfills a majority of the customer’s requests.
The Neocom Algorithm executes a variety of calculations behind the scenes and allocates these labels accordingly.
Every Neocom user has the ability to generate Custom Labels to highlight any products they wish, for any reason they deem as important. These Custom Labels work on the same mapping principle as the mapping found in any question within your advisor. These Custom Labels offer a fantastic visual queue to inform visitors of any special feature you wish, whether that is that the product is on sale, there is limited stock left, or it is a newly released product.
Returning to our bike shop example. Let’s say it is the beginning of the bike season and the bike shop wants to highlight last season's stock to make way for new inventory. The bike shop can create a Sale, or End of Season Label in bright red to draw the customer's attention. Through appropriate mapping, the bike shop can enable this tag on any product that has a sales price, or, has last year's release date. For each label created the bike shop has the option to show the label once on the first match, or more than once on any match of the result page. There is also the option to stack different labels on the same product, or just one.
These Custom Labels can be created to highlight products for infinite reasons. Some common uses are Sale, Seasonal Products, Low Stock, High Stock, In-House Brands, and many more. Just ensure you don’t clutter the result page too drastically with Custom Labels.
Think of Neocom’s Smart Algorithm as your personal Co-Pilot. It constantly makes decisions behind the scenes to ensure your customers are receiving the right recommendations after each completed advisor.