You are here

Guest Blog - Colin Carter from weather2travel on using data feeds for marketing travel

How structured data helps travel affiliates stand out from the crowd

Affiliate marketing has become even more powerful over the last few years. As more and more online companies (merchants) have developed datafeeds to distribute their product information to publishers (affiliates), they have identified the benefit of giving affiliates access to product level data along with vouchers and special offers. This enables their products to be promoted to an extended audience through long-tail search at relatively low cost.

Merchants distribute their product information to affiliates using datafeeds, which are typically XML or delimited text files via their affiliate network, agency, datafeed specialist or directly themselves.  Given that consumers today want ‘cool’ applications and ‘one-stop-shop’ websites, affiliates must look to combine product-level data with their own content to create a unique experience that is more likely to drive traffic and sales.

At for example, we combine our climate and weather content for specific destinations with travel products such as package holidays, hotels, flights and hire cars. This enables us to promote specific travel products for a given destination alongside our climate guide or weather forecast. Our experience is that showing destination-specific products with a price point and a landing page deep within the booking process will convert much better than using traditional banners, buttons and text links.

The Structured Data Challenge

The quality of product-level information provided by merchants within their datafeed is vital. It should be well structured, consistent in terms of hierarchy and labelling, as well as updated regularly, depending on the type of product. More fundamentally the quality of the data itself needs to be high; however this is where a lot of current datafeeds fall down.

All too often, key fields/attributes are missing or incorrect. Simple problems such as spelling mistakes, product names in caps, truncated fields, badly encoded text and broken links to products and thumbnail images makes the job of the affiliate considerably more difficult.

Furthermore, product attributes are often provided with long generic descriptions, which are all but useless to affiliates ever since Google clamped down on duplicate content with the Panda update. A better solution would be shorter attribute-based fields. In the travel sector, for example, a hotel’s datafeed could include categorised fields with the hotels facilities (i.e. Pool = yes, Restaurant = no, Car Park = yes) rather than long hotel descriptions. This reduces the size of the datafeed and enables affiliates to present this content how they wish.

One Size Doesn’t Fit All!

It’s also important to understand the different types of datafeeds available. Traditionally, a datafeed was at the ‘Lead Price’ product level where a product has a single landing page and the price was either indicative or static. The datafeed would simply have a single record for every product and would therefore be finite. When creating a datafeed of products with many variations in parameters, such as the ‘availability’ of package holidays, there will potentially be millions of permutations, with different departure dates, departure airports, board basis and holiday durations.

Because of the amount of data produced, only a limited number of affiliates are able to work with ‘availability’ datafeeds, therefore merchants need to think much more creatively in terms of how they offer their product data. Presenting their datafeeds through interactive widgets and dynamic banners is a clever way to engage with affiliates who don’t have the technical know-how to work with large datasets.

It is important to understand that there are different types of affiliates out there and if a merchant wishes to engage with as many affiliates as possible they may need to consider offering different levels of datafeeds and widgets. The most important point is that the merchant takes control of its product-level data and has a clear understanding of what affiliates need:


  • Include as much product/destination information as possible
  • Check URLs work and thumbnail images exist
  • Offer levels of data feed (lead price and availability)
  • Provide interactive widgets and dynamic banners
  • Normalise data to cut down file size and remove duplication
  • Include categorisation (e.g. hotel facilities) so affiliates can build their own descriptions


  • Include long descriptions (i.e. avoid duplicate content)
  • Use non-ASCII characters (where possible) or be consistent with encoding
  • Change the data feed structure without telling affiliates
  • Export millions of records that are unmanageable
  • Export products that are not available or where the prices are incorrect

Colin Carter is Technical Director of


Contact FusePump


Sign Up For The Newsletter