Make sure you have the login given to you by Avail. The customer id is in the format xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.

Upload the filter data to Avail. See tab Data for instructions on how to do this. Inform Avail about the names and data types of the Product Data you are going to upload.
To make the recommendations based on a product, products, a session or a user work well from day one, you can upload an initial feed of transaction data (i.e. order rows).

Access the Control Panel which can be found under the tab Resources. The Control Panel lets you configure and test the recommendations and inspect statistics on the recommendations. For an in-depth guide on how to use the Control Panel download the Control Panel Guide which can be found under the tab Resources.

For the recommendations based on a product, products or a user, Avail BEM needs order data. When an order is placed use the method logPurchase to collect new data for recommendations. See tab API for details.

For the search recommendations Avail BEM needs data on searches made. Use saveSearch:
1) When a search has been made and a search result is added to the cart from the search result list
2) When a search is made, user clicks on a search result to go to product detail page and adds that specific product to the cart without any intermediate clicks.
See tab API for more details.

For the landing page recommendations Avail BEM needs data on the search phrase used on the external search engine and the first product added to the cart during the session. Use saveLandingPageData to collect data for landing page recommendations. See tab API for details.

To enable detailed metrics and a set of features like A/B Testing, Avail BEM needs information on clicked-on and added-to-carts products. Use logClickedOn, logAddedToCart and logRemovedFromCart to collect data. See tab API for details.
note: logClickedOn and logAddedToCart uses a server generated TrackingCode. You get a TrackingCode returned from the Avail server by committing a predictions request.

To make recommendations use the method getRecommendations. This is a generic method that excepts inputs such as product id, phrase and more. The type of recommendation returned is governed by the template. The See tabs API and Interfaces for details on how to implement recommendations.