10 Data Feeds To Integrate With Your ESP
Cara Olson details the essential data feeds that let you send relevant emails to your customers based on quality data.
The ability to send relevant emails is based on access to quality data. Data are usually integrated either in real-time (using an API call) or via a “feed” that posts a file to a secure FTP site for batch import into your email service provider (ESP) on a regular frequency.
Following are 10 key incoming data feeds every retailer should have integrated with their ESP.
This data feed should include all of the core information required when a customer creates an account, such as name, address, gender, customer ID, household ID, etc. Typically this would be a feed that originates from the CRM data warehouse, but this feed could also be as simple as the data collected when a subscriber updates her profile and preferences.
This allows pertinent data to be available within the ESP for segmentation and dynamic content. It also allows an email to be triggered to confirm the account setup and communicate the benefits of having an account. This email is often saved by a customer to refer to at a later date.
2. Purchase (If Not Real-Time)
If orders are not directly integrated via API to the ESP, then a daily batch feed of purchase details should be integrated. This feed should include all details relevant to the transaction, including product name, product price, image URL, coupon code and discount, shipping, order total, and so on. This feed originates from your e-commerce platform (such as Demandware, Magento, etc.)
If the purchase data is sent via API then an order confirmation email can be sent immediately. If not, and the data is delivered in a batch feed, then the order confirmation should be delivered via the e-commerce platform.
This feed should include details regarding a customer’s shipment, including a link to track their order. This feed can originate from the fulfillment or e-commerce platform or from the shipper (such as UPS, FedEx, USPS, etc.)
This allows a shipment notification email to be sent when the order is on its way. In addition, it could allow an email to be sent when the shipper indicates the order has been delivered. As part of the post-purchase email series, the delivered date in this data feed can also drive when the email should be sent, requesting the customer to write a review.
This feed should include details for items that customers returned and originates from the fulfillment or e-commerce platform.
This allows a return-received email to be delivered to the customer, which could also include alternative products the customer may want to try (depending on the reason for return). This data should also be factored into the logic for dynamic product recommendations. For example, if a product were returned because the customer disliked the style, you would not want to recommend the same style in a different color/pattern/finish.
This feed should include your product inventory, such as product image URLs, product names, SKUs, descriptions, price, and so on.
This information can be utilized in triggered messaging, such as abandoned shopping carts, order confirmation emails, dynamic content based on previous purchase history, and more. Depending on the size of the catalog, this feed may need to only include changes on a nightly basis, rather than a full sync each day. In addition, depending on how frequently the products and prices change, it may not be necessary to update on a daily basis.
6. Abandoned Cart
Abandoned shopping cart emails can be sent through a variety of integrations, including a third-party abandoned cart vendor like SeeWhy, or through a site analytics integration, such as Omniture or Core Metrics.
Unless you are using a third party to send these triggers for you, a data feed will need to be setup to indicate which subscribers should receive abandoned cart notifications and which products should be displayed. This data feed could only include the unique identifier for the product (such as SKU) and then a look-up for that product’s details could be coded to access the product data feed’s table to display the most current price.
7. Abandon Browse
In addition to abandoned cart, a feed for customers who only browsed products but did not add them to their cart should be setup as well. This audience will be limited because it requires users to have been to the site before and to have provided an email address at some point.
The user could have also clicked-through from an email and browsed online in order to know who to send the abandon browse email. At minimum this email will need to include the email address and then at least one type of unique identifier for the product(s) browsed, such as SKU, product name, or category like in this example below from J.Crew Factory. The subject line read, “Thinking about it?”
8. Point of Sale (POS)
If you have brick-and-mortar locations, you will need a data feed integration with your POS locations. Depending on your POS software, you may have real-time integrations to trigger a welcome email after a customer provides their email address to opt-in during checkout. Regardless, your POS data feed should also include the customer’s order details for their in-store transaction.
If you have a loyalty program, you may have an additional data feed, or this feed may replace one of your other feeds, like the Account/Profile.
The data in this feed allows you to display point balances and dynamic content for a customer to reach the next level/reward.
10. Store Locations
Depending on the number of locations and how often new stores open, you will likely have a store locations data feed that includes address, hours, and potentially contact information. Emails can then be setup to dynamically populate the closest store location to the subscriber. In some instances emails can also be sent on behalf of store location managers for specific events or sales.
You may notice that subscriber sign-ups is not listed as a feed. New opt-ins should almost always be sent via API to trigger a welcome email in real-time. Depending on your data structure and vendor partners, you may have more than these 10 feeds (for example, if you have a site satisfaction survey integration you may also have a feed indicating who should receive emails to complete a survey). Also, depending on the volume of your product catalog and orders, you may have multiple feeds to accommodate delta files as well as full syncs on a consistent basis.
Typically you would have additional tables that are the result of data joining across more than one of these feeds. In addition, you will have subscriber data inherent to the ESP and independent of a data feed, such as bounces and unsubscribes. Finally, there will be other integrations that are not necessarily data feed based, such as your analytics integration (Omniture, Core Metrics, Google Analytics, Web Trends), Rating and Review platform (such as Bazaarvoice, PowerReviews), or an email address validation tool (such as Fresh Address, QAS).
How many of these data feeds do you have integrated or are in the process of integrating? Which ones were more difficult to complete than others? Which have been the most valuable? Have you had to get creative in how to get the feeds integrated? I would love to hear what your experience has been in your quests to send the most relevant emails based on data.