Getting your recommendations from Quicklizard

This guide will walk you through how to fetch all your latest recommendations. 

The way to pull your new recommendations from Quicklizard is via API and it's a very simple process.
Our API endpoint api/v3/recommendations/log is the best way to pull your recommendations. 

The Power of Integration for Enhanced Recommendations
 

Prerequisites

First, you need to get your data to validate the request. You will need your API secret, API key, client key, and channel, follow this guide to get it on our Getting started with API guide

 

Building your request
 

After getting all the information needed, it's time to build the request. 
The endpoint requires 3 parameters to be passed in the query: 
Client_key, qts, and since. 
The client_key was fetched earlier. 
The qts is the timestamp of the request in UTC Unix timestamp in seconds.
The since is the time and date from which I want to get my recommendations. 

Other options to pass are channel (for omnichannel systems), last_id, changed, accepted, and only_new_accepted. 

To investigate the parameters please check the endpoint documentation here

 

Design considerations

• Each request to the API brings back up to 200 recommendations, and a value called last_id. 
  Each recommendation has a unique identifier, and in order to iterate over more than 200 products, we need to know what is the last recommendation ID of the current batch and send it in the query to the next, to know where we stopped. In order to do so, in each request you will get a last_id value to pass in the next iteration of the same batch. 
  In the first call, the last_id should not be sent, in the last call you will receive the last_id empty, and that's how you know you ran over all the recommendations in the results. 
• If a product was sent to pricing twice during the time we are looking at and fits the query, it will be seen in the results twice, make sure always to take the latest result. 
• Take into account your pricing timings agreed on when using this endpoint, to make sure optimal usage of the system. 
• For additional help, you can consult our dedicated support team.