05 April 2015 · By Nikhil Hunshikatti
Data gathering tool allows The Columbus Dispatch to step away from third-party data companies while retaining high-risk subscribers, optimising advertising revenue, and improving content delivery.
The Holy Grail in the publishing world is to harness our ability to maintain a direct relationship with our customers when using various third-party data providers and applications.
Publishers such as us – The Columbus Dispatch – that have a limited footprint or single market coverage find it difficult to invest in proprietary technology, unlike large publishing corporations such as Gannett, Tribune, etc.
We end up relying on multiple technology vendors with disparate data management platforms that don’t talk to each other and, as a result, miss out on the opportunity to harness the endless amounts of data to drive revenue, tailor content, remain competitive, and more strategically appeal to advertisers and our news consumers.
Defining our digital future
We know our future is increasingly digital. We also know that reliance on third-party data providers and multiple technology vendors is inevitable. However, as a single market publisher, we had to find a solution to harness the terabytes of first-party data that we sit on.
Enter Mather Economics and its new Listener™ tool. Listener would provide us with a “near real time” capture of user data across our Web and mobile sites and mobile applications and, through the process of entity resolution, combine that data with additional offline information about our customers and prospects.
For the past six-plus years, our organisation has been partnering with Mather Economics for pricing strategy and activation. As a result, Mather already had access to a significant amount of our print subscriber data.
It was a no-brainer to extend the relationship by launching a data science initiative – with Listener serving as the cornerstone of this effort.
Clearly defined goals
At the outset, we knew that we had to clearly define expectations from the initiative across all our platforms. That helped identify what data to capture. No more silos.
The goals we identified were to support both sides of the equation – subscribers as well as advertisers.
For subscribers, we identified developing and institutionalising customer lifetime value for our print and digital users, pricing strategy based on digital content usage and modeling for churn and thereby retaining “high-risk” subscribers as some of the key goals.
On the advertising side, we identified dynamic digital advertising rates, as well as dynamic metering to optimise advertising revenue while not cannibalising our digital subscription revenue opportunity.
Editorial also would benefit from the initiative through improved intelligence on how our core consumers were interacting with our content in the digital space, thereby enabling improved content delivery and formats across platforms, time of day and devices.
Notching early wins
While listener code was being introduced on our Web sites, institutionalising customer lifetime value (CLV) became the priority. We had CLV appended to our current subscribers as well as formers and never subscribers, which then got integrated into our marketing automation tool.
We built campaigns to optimise volume and revenue. We saw wins on multiple fronts – increased daily units and improved value of low CLV subscribers.
We also overlaid CLV data on top of key advertiser desired zip codes and saw increased engagement of high value CLV as measured by upgrades to high preprint volume days. CLV data is now deployed at our customer call center with scripting for stop-saves and a rate matrix for downgrades/upgrades.
The wins kept on coming. With Listener in place for about three months:
- We now have sufficient data to start looking at pricing our subscribers based on their usage of our digital content either via our metered Web site or the e-edition (Web and tablet). We are in the process of testing the target files and price points and expect to deploy on a weekly basis.
- We are able to establish a strong correlation between article viewership and subscriber retention, thereby triggering campaigns to improve engagement with our digital assets among our low and medium users.
- We now have the ability to pass through our proprietary segmentation data (at HH level) and assign that to individuals (established through the entity resolution process). This enables us to know how our core customer segments — media sophisticates and traditionalists — are interacting with our content. Not just descriptive statistics at an aggregated level, but right down to the most granular data points such as types of stories they are interested in and spend time with
What’s coming next?
As you can see, we are just scratching the surface with some of the examples shared above. The Listener tool will be playing a critical role in some of the following key initiatives to help retain our existing subscribers, grow digital subscription and advertising revenue and better engage with our consumers.
- Churn model: Using all the digital and offline data on subscriber behaviour, we will be able to apply predictive analytics to identify “at-risk” subscribers. The model also will provide us with clear direction as to which triggers would pre-empt churn — be it EZPay, loyalty membership, or increased article PVs in sports.
- Dynamic metering: Listener’s ability to integrate within multiple third-party plug-ins (for example, Syncronex/Newscycle) allows for quick implementation of dynamic metering — the ability to adapt meter counts based on content type, user or story-level.
- Segmentation: Improved ability to customise messaging and campaigns at an individual level. Listener data will help maximise digital interactions and drive engagement through modeled user behaviour and move potential subscribers along the path from casual reader of the site to a paid subscriber.
The twain shall meet
It would seem from reading up to this point that everything we touched turned to gold. I wish that were the case. We definitely had our share of growing pains.
But, as mentioned before, as we kept scoring those small wins, the organisation and the stakeholders truly got on board, and this became a priority for all involved.
We still have work to do to get Listener software development kit (SDK) integrated into our mobile applications offerings. We are working with multiple vendors to define timelines and level of effort and getting our mobile applications authenticated and metered.
This will alleviate our inability to make the “individual” connections between Web and mobile and will move us farther along in terms of establishing a future in an increasingly digital world.
And, the beauty of all this is: Going forward, we won’t have to pay a third party for data on how our news consumers use our content, and how we can serve them best. We — the publishers — have had that information all along. We had just not harnessed it.
Listener and the data science initiative are enabling us to move in exactly that direction.