< All Topics

Control Your Feed With Filters | How to Guide

Set rules to determine which product or variant records are included or excluded from appearing in your feed.

The Basics

Your Feed Donkey product database that is in sync with your Shopify product database will by default or in other words ‘unfiltered’ form include every single SKU record no matter its publish status, data completeness or any other test. Therefore we will when configuring a feed usually need to apply some sort of filter to feeds that you are using in a live ‘production’ capacity.

In this case we will be wanting to make sure that unpublished/ incomplete are not included and we only send to our channels products that are ‘viable’ for promotional purposes.

When you create a feed from a channel template

This will come with a pre-configured filter

So the rule set here i.e. SKU record is Excluded If Products_OnlineStoreURL Is Blank

The presence of this data element with Shopify, directly relates to its publish status in the Online Store sales channel. So if this field is unpopulated, you can be sure that this is an unpublished product.

Now there may be exceptions to the requirement for this filter, but in the vast majority of cases for controlling viable product inclusion this is an essential one and why we include it as a default in the templates.

You may wish to further safeguard the quality of your feed by adding more viability type filters, for example most if not all channel feeds will have a minimum requirement that the product has at least one image URL, so a robust indicator that a product should be excluded is if it does not have a featured image (Products_featuredImage.src). So this is a perfectly valid filter to apply (although maybe most incomplete products should not be published and therefore would be caught by the first filter).

This all depends upon your circumstances, your businesses vertical, or the requirements of the channel. For example, it may be that GTIN/ barcode is absolutely essential for a product to meet the minimum data quality requirement – if it is not present it will be rejected as an error. If this is the case a filter on this would be advantageous.

Common Use Cases

As explained above not feeding unviable products is the most common use. Probably second to this is using filters to exclude products that cannot be fulfilled due to stock availlability.

There are generally a couple of methods that you can guard against ‘over-selling’ or in other words selling/ promoting stock that you cannot fulfil. One way is to pass all viable/ current but indicate with a variable.

Alternatively, you could choose to employ a harder border between promtoted and non-promoted product i.e. if it isnt in-stock dont send it. The suitability of doing this does depend on some factors though – i.e. you need to ensure that if a product drops out of your feed it has the same effect on the channel side & it doesnt persist with outdated information.

Another common use of filters might be to control the ‘type’ of products included – so for example a store might have a mixture of packaged/ barcoded items plus custom made/ personalised items. Only one of these being appropriate for the marketing channel. So Filters would be the method to ensure the product set was correct.

Filters can be stacked, there are no limits to the number that can be applied, mutliple conditions can be tested against a data element.

Previous Column Preview Function
Next Daisy-chaining Processes | Real-world Scenarios