2018 Finalist


Fairfax Media

North Sydney, Australia

Category Best Use of Data Analytics

Overview of this campaign

Australian Community Media (Fairfax Regional News) has a network of over 500 sales people and managers who require up to date audience information for each of their 300 news brands. Each news brand can have thousands of individual clients each year. Sales offices are distributed over every state in Australia without a centralised distributor of sales collateral.

Providing sales collateral based on research has historically been a difficult task. Audience metrics for print are provided by Ipsos (through emmatm), digital metrics by Nielsen and demographic metrics by census data (ABS). Because of the number of different data sources; provision of research was previously conducted by an analyst who provided ad hoc research for account managers who looked after the largest clients. Because of this, about 1% of sales people received data-driven sales collateral.

The project’s objectives were to create a self-serve audience research tool for all sales people across the business to use. The tool is intended to be used to create new revenue through creating ready-made presentations for sales people responding to briefs or pro-actively pitching for business. The tool provides brand specific data for each potential advertiser for specific newspapers. The tool also contains print/digital reach and audience demographic data for 300 newspapers. All newspapers are also plotted on to Google My Maps, providing a fully searchable (newspaper, postcode, suburb, Australian business) territory map.

A key objective was that each section of the presentation can be stand-alone and client ready. This meant that it had to be personalised to each possible client. For example, if the client is an airline then different metrics to a car manufacturing client will be shown. The objective was to use NLG text which names the selected client (eg x% of y newspaper readers have shopped at z in the last year) meaning that all presentations are personalised and simple to understand for a salesperson.

Results for this campaign

The result of the creation of a self-serve information tool is that 100% of sales people have access to reliable, up to date information. Previously in order for a sales person to receive collateral for a pitch they needed to send long email chains to insights analysts, graphic designers and management. This process took up to 5 hours in the past.

There are 350 brands included in the tool, spread over 50 industries. Brand level detail is provided for 90 news brands across all formats, available for 21 demographics. There are a maximum number of 750,000 audience combinations. For a human team to produce this number of personalised presentations would take approximately 425 years of continuous work, InsightsPlus completes this in a fraction of a second.

Since soft launch in September the tool now averages 50 unique users/month. Despite still being in beta phase the tool has already been used to win $70,000 revenue in sales pitches, this is forecast to increase at general release.

One result of the project was the creation of a fully scalable proprietary database that can be used for other BI functions. In order to permit a wide range of industry specific information for each presentation a key-value style non-relational database was designed which integrated Natural Language Generation. This allowed for greater flexibility in data visualisation as depending on the key selected; the screen would display different metrics. In traditional BI systems fed by relational databases the values change but the metrics don’t, in the non-relational system both the metrics and values change. Each industry and data type is a separate module that can be added to the database without inserting new columns, making the database massively scalable.


To contact a company representative about this campaign, click here for the INMA Member Directory