Best use of FastStats® 2014 Award

Revitalising an under-used tool to overhaul subscription email marketing


Haymarket Media Group creates award-winning specialist content for international audiences. The company has 72 market-leading brands in 20 offices across six countries, connecting people and communities across digital, mobile, print and live media.

Although Haymarket’s portfolio is diverse, its philosophy is the same across borders and markets: deliver brand experiences that truly meet the needs of audiences and clients.  

Haymarket’s mission is to be the best specialist media and information company in its chosen markets. Five key values underpin everything that Haymarket does: Creativity, Integrity, Dynamism, Independence, and Expertise.


Haymarket have historically had a ‘one size fits all’ approach to subscriptions email marketing with customers receiving the same message, and a focus on offer and creative, rather than the customer.

They promoted universal subscription offers to all customers, rather than providing more tailored offers for each type of customer. For instance, cheaper trial offers were often used extensively to deliver higher responses. 

Campaigns were managed in isolation to each other and sent out once every two months, and little consideration was given to the customer journey, resulting in more work, more time, and more costs.

The data selection request always targeted previously lapsed subscribers, competition entrants, and website registrants to that specific brand being promoted. From time-to-time, there would be cross-sell activity, but all customers would receive the same offer. The promotion and creative came first, and the data selection would always follow the lines of, ‘give me everyone we’ve got’. 

FastStats is the key in understanding a customer’s interests and building Haymarket’s new approach to subscription email marketing.


Until recently, this approach delivered good returns considering the low cost compared with other channels. But in 2013, a significant decline was noticed in response rates from email marketing, despite the increase of campaigns. What was traditionally a successful and cost-effective channel for subscriptions had now become more expensive and ineffective. Haymarket needed a new subscription email marketing approach.

The objective was to improve response rates, increase up-sell and cross-sell, decrease costs and reduce the overall time spent on email campaigns.

At the start of 2014, Alchemetrics completed the upgrade of Haymarket’s marketing database platform by rebuilding the informa® SCV. The solution included a complete re-engineering of the FastStats data model, together with ground-breaking integration between informa® and FastStats meant Haymarket were now able to add new variables to the SCV themselves which then automatically map through to FastStats and visible the following day. With such automated management of the data environment, meant FastStats was always up to date with the most recent customer data, and it allowed Haymarket to analyse and action the data within a day of a customer engaging.

With the new database in place, and the declining subscription sales, Haymarket strategically reviewed the entire approach to data driven marketing, with a priority given to addressing the declining subscription sales from the email marketing channel. 

The first step was to use FastStats to analyse all Haymarket’s data and explore different subscription marketing ideas and initiatives.

Selections were built to count the volume of customers within different data pots, split further into recency intervals. I.e. 0-6 month lapsed subscribers.

Using Cube Analysis, Haymarket could visualise the thousands of new registrants, competition entrants, and other prospect data that were available for marketing.

As a result of FastStats integrating with our email service provider (to provide daily updates of email event behaviour), Haymarket could view the customers who were opening and clicking on each email.

Using a simple counting wizard, an aggregation variable was created to show the number of active subscription accounts each individual had. This analysis validated the assumption that the majority of Haymarket’s customers only subscribed to 1 title and highlighted a potential for cross-sell opportunities.

Taking these factors into account, and through analysing the frequency, the recency, the number, and the type of customer activities, Haymarket could demonstrate the importance of brand engagement as a factor to predict a customer’s likely response to a subscription offer.


Haymarket decided to roll out a new portfolio email campaign, where the content was primarily driven from a customer’s interaction with eachMagazine email image brand. Each individual would receive a different email promoting different brands and different offers depending on their specific level of engagement with each brand.

A Flag Array was built containing a list of engagement activities with Expressions then used to score each activity depending on the level of engagement. Within the Expression, scores were then totalled up for each separate brand and arranged into descending order with the highest scoring brand at the top.

The scores were then used to ensure a customer receives promotions from a brand they were most interested in determined by the level of engagement they have with each brand. Haymarket now knew what brands their customers were interested in. 

Next they wanted their customers to receive targeted subscription offers, rather than everyone getting the same. Pack Attributes were created to allow Pack Codes representing different types of subscription offers to be entered manually.

Haymarket made use of their email service provider to deploy their new portfolio email containing all the different possibilities of brands and offers by creating a relational table of campaign offers and linking it to the customer data file exported from FastStats. 

When Haymarket uploads the customer data file, containing the brand and pack variables, it automatically links to the campaign offers table, and pulls in the relevant brand and offer content for that specific customer.

Increasing the number of brands, and adding or amending the campaign offers is very scalable and requires little work. It doesn't matter if hundreds of offers were used, as long as the brand and pack codes match up to the customer data file, the email will contain the correct targeted offers.

To date, the highest number of permutations within one email has been 5,829 different brands and offers.


In order to measure the effectiveness of the new portfolio email, Haymarket tested it against a traditional solus email for the 2014 Father’s Day campaign.

Fathers Day EmailThe Father’s Day campaign was made up of 4 separate emails over 4 weeks. The data was randomly split into 2 separate segments - SegA and SegB. 

The results were very positive.

The new portfolio email generated 65% more subscription orders than the traditional solus email. The new portfolio email generated 46% more revenue than the traditional solus email. 

Using FastStats to produce the results means Haymarket can identify the specific campaign codes customers used when purchasing a subscription. Furthermore, they can identify customers that went onto buy a subscription using a different campaign code. This insight can only be performed because Haymarket have all their most recent customer data available within FastStats.

Since the Father’s day test, a further 3 portfolio emails have been sent out (one a month). The volume of sales each month has continued to generate 40% more than what a traditional solus email would generate.