Recommendation tool and its benefits

Zia recommendation tool

Zia recommendation is a self-learning tool that recommends the right product for the right user at the right time by understanding the behavioral pattern of the entire customer base. We have outlined some common areas where you can use the recommendation tool to pitch the right product to your customers:

• Real-estate: The agents can suggest the most suitable property based on the customers' previous purchases and the attributes that they seek in a property. For example, a customer looking for an upscale condo with 4+ bedrooms, 150-200 Sqft with amenities like outdoor parking and courtyard, can be suggested similar apartments in and around the preferred location. The recommendation tool will understand the customer's preferences from the above specifications and suggest properties that have similar attributes, reducing the manual effort needed to filter the entire property listing and scouting the right one. Also, you can reduce the time and money invested in lead nurturing campaigns by sending audience-specific content, increase the chances of conversion, send personalized marketing content.
  
• Online learning platforms : Students who have subscribed for a particular course can be recommended similar courses based on the topic they have subscribed for, personal preferences and preferences of other students with similar characteristics. Based on the above results, you can suggest related materials such as interactive videos, assessments, supplementary education materials etc. to ensure better engagement. Zia's content-based filter lets you recommend more courses based on the current course content. For example, if a student has subscribed for Foundation Course on Machine Learning, you can suggest other courses that have related content like Deep Learning, Neural Networks, Regressions etc. 
  
• E-commerce : Customers are more likely to buy from you if they know you understand their requirements better. Therefore, in addition to suggesting products based on an individual's preference, buying patterns, and interests; Zia can use collaborative data filtering to suggest items that are often been bought alongside their product of interest. This will help you enhance the cross-selling and up-selling and definitely improve customer satisfaction and retention. 
  

Permission Required

Users with administrative profile can create recommendations.

Configuring the recommendation builder

• Selecting the module that should be recommended : Choose a module that contains information about what you want to recommend for example, deals, products, services etc. You can define criteria, if you want to set up recommendation for specific type of records.
  
• Selecting the target module : The module that contains end user or customer information is the target module. It can be a standard or a custom module. For example, if you want to recommend courses for customers, leads or contacts, the latter's information must be in the target module. However, keep in mind that only those modules that are interrelated to the source module through a lookup field or related list can be selected as a target module. If you choose a module that is unrelated to the source module, the system will show an error message prompting you to choose another module. Once the recommendation model is set up, the system will display the recommendations within the target module. here is an example of a recommended property displayed in the Properties or Apartments module.  You can create only one recommendation model per target module.
    
  
• Selecting the linking module : A module that contains information about source and target module is an interlinking module. It is mandatory to select a connecting module because it will help Zia determine the successful transactions that take place in your business.
  For example,  Deal  is the connecting module for  Product  and  Contact  modules, because it has all the information about the purchased products, date of purchase, type of customers who choose a particular product, other products that may have been purchased by them, type of products preferred by a group of customers.
• Indicating the completion state of a process:  The model can recommend the right product by analyzing the successful transactions that have taken place in your business. When a subscription is renewed, a deal is won, a product is purchased or a course is taken, they indicate successful completion of a sales process. These data are considered as positive values by Zia to make recommendations that best suit the customers with similar characteristics. These fields must be present in the  Connecting  module.
  In addition to this you must select the data/time field that indicates the completion state of these processes, for example, purchased on, subscription renewed on, deal closed on etc. The date and time field will help Zia learn the frequency of purchases and recommend those items for repeat purchase, next purchase or items to be bought together etc. 

To configure recommendations

1. Go to  Setup  >  Zia  >  Recommendation . 
   
2. In the  Builder  tab, click  New Recommendation . 
   
3. In  Create New Recommendation , enter the  Recommendation Name .
   
4. In What to Recommend? select a Module  from the drop-down list.
   For example, we want to recommend products.
5. In Criteria , select  All records  or specify the criteria.
   
6. In  Recommendation should be done to?   select a  Module  from the drop-down list.
   For example, we want to recommend it to the contacts.
7. In Criteria, select  All records  or or specify the criteria.
   
   
8. In  Linking module , select a module from the drop-down list that connects the above modules. 
   For example, Deals module can be a linking module between the Products and Contacts modules.
   
9. In  Criteria for completion state , choose the field(s) and the appropriate values.
   For example, the completion state is when the deal stage field is closed won.
10. In  Duration , select a field from the drop-down list, that determines the completion state.
    For example, the completion stage is determined from the deal closing date.
    
    
11. Click  Save , to save the recommendation model.
    

Viewing recommendation results

Provide direct feedback for the recommendations 

For every recommendation, to assess the usefulness of recommendation, Zia asks for your feedback.

• If you find the recommendation helpful, responding Yes will retain that recommendation until the completion state defined for that recommendation.

• If you do not find the recommendation helpful, by responding No , Zia will remove the recommendation from your record. 
  Furthermore, to investigate and to improve your experience in the future, Zia requests you to fill out the reason as to what went wrong.
  
  

Notes : 

1. Zia direct feedback for recommendation is not available for portal use.
   

1. Users with view recommendation permission in their profile, can only view and respond to this feedback.
   

View contact references for a recommendation

1. Contact's name and account associated with them.
   
2. Similarity score, an auto-computed value that is calculated by taking multiple factors into consideration that match between the customers. Higher score points at greater likelihood of a customer to purchase, as it is closely similar to the referred contact.
   
3. Time since the referred contact has been in business with you eg., first deal on 4th June 2015. This helps to understand the longevity and thus their loyalty to the business which may be an important factor for the new customer to consider doing business with you.
   
4. Number of times the contact has purchased the similar product.
   
5. The list of factors that were taken into consideration to draw similarities. For eg., Hubert is a referred contact for Leota because their mode of purchase is online, recent visit was yesterday for both, their geographic location and more.  
   

Note

1. Names and details of only 5 references contacts will be displayed.
   
2. The similarity score cannot be edited, it is auto-computed by Zia. This score can be used to identify the customers who would most likely be interested to purchase and improve cross-selling opportunities.
   
3. A maximum of 5 field values that majorly contributed towards computation of similarity score will be displayed in the detail page. 
   

Choosing recommendation models from the gallery

1. First time - Depending on the buyer's first purchase Zia will recommend the appropriate product.
   
2. Repeat - This pattern will identify the product or service that is repeatedly selected by the buyer. For example, a food and grocery chain can identify the customer who frequently purchases the same items and suggest the same to prevent discontinuity in the purchase cycle. The user can click ' Know more ' to find out why the recommendation is reliable. Details like when the item was last purchased and the trend is shown to the user. 
   
   
3. Relationship -  Zia will identify the buying pattern and give suggestions, for example, if other buyers have purchased a laptop and a laptop stand, then any buyer who purchases a laptop will be suggested to buy the stand as well.
   
4. Bundle - If any buyer purchases multiple items together such as mobile phone, screen guard, and mobile cover. Then, Zia will consider this as a bundle and recommend bulk purchase if anyone has selected either of the products from the bundle.
   
5. Sequence - Zia will identify the products that are purchased in a series and recommend the next items in the sequence if a similar purchase is made in the future. For example, if a student enrolment in AI courses is in this order: Fundamentals of Artificial Intelligence followed by Neural Networks, whenever a student enrols for a similar course the other courses in sequence will be recommended.  
   

What are available recommendation?

Setting advanced customization for each recommendation type

Note: Notifications and sharing recommendations with portals users can be configured for all the recommendation types. However, Workflow can be configured only for Repeat type of recommendation. 

Renaming the types of recommendations

Based on business relevance admins can rename the recommendation types.

1. Go to Setup > Zia > Zia Recommendation > Builder .
   
2. Select a Recommendation from the list view.
   
3. Scroll to Additional Information .
4. Hover on the name of any recommendation type and click the Edit icon next to it.
   
5. Enter a name for the recommendation type.
   
6. Click the tick icon to save it.
   

Selecting recommendation models for portal users

1. Go to Setup > Zia > Recommendation .
   
2. In the Builder tab, select a Recommendation from the list view.
   
3. In the recommendation model, scroll down to Additional Information .
   
4. Expand Advanced Customization , under Portals click Associate .
   
5. In Portal Association popup, select the portals from the list of available portals.
   
   
6. Click Save .
   

Create Workflow based on the recommendation

Note

1. A workflow can be created only for the target module, that is the module where your customers are or to whom the recommendation applies.
   
2. Recommendations that are shown for a product or deal that is already purchased or associated with the contact, are applicable for the Workflow. That is, if the contact has purchased a laptop then Zia will recommend the same product again; for this recommendation you can create a workflow. 
   

1. Worklfows configuration page
   
2. Zia Recommendation

1. Go to Setup > Automation > Workflow rules .
   
2. Click Create .
   Only the target modules or the ones for which recommendation is suggested will be available.  
3. In When do you want to execute the rule? select Based on recommendation .
   
4. In Execute , select the number of days and before/after/on from the drop-down list.
   
5. Click Next .
   
6. In Would you like to add some additional conditions to this rule? select Yes or No .
   If you select Ye s, enter the desired workflow conditions.
7. Click Done .
   
8. Select an Instant or Scheduled action .
   
9. Click Save .
   

1. Go to Setup > Zia > Zia Recommendation > Builder .
   
2. Select a Recommendation from the list view.
   
3. Scroll to Additional Information and select Repeat type recommendation .
   
4. Expand Advanced Customization , under Automation click + Trigger a workflow rule .
   You will be redirected to the WF configuration page.
5. Configure the Workflow rule as mentioned above in Steps 3 to 9.
   

Setting up notifications for recommendations

Note:

1. You can create upto 5 notifications.
   
2. Notifications can be edited or deleted from the recommendations page.
   
3. User fields added to the target module will be displayed while choosing to whom the notification should be sent.
   
4. Users can view the notifications under Zia notifications. They can also toggle them off from here. 
   

To set up auto-notification

1. Go to Setup > Zia > Zia Recommendation > Builder .

2. Select a Recommendation from the list view.

3. Scroll to Additional Information and go to one of the recommendation types.

4. Expand Advanced Customization and click +New under Notifications .

5. In the New Notification page, do the following:

1. Execute from: Select a date and time from when the notification must be sent.

2. Select Any record [eg., Deals] or Selected records.
   If you are recommending products, you can choose Any product or selected products.

3. Frequency: Select Every week or Every day.

4. Notify to: Select Record Owner or other users.
   All the user fields added to the target module will be displayed.

6. Click Save .
   

Editing, Deleting or Deactivating a model