NNA (Nine Network Australia) has a challenging task of managing pricing and sales for its finite volume of advertising space on its free to air television stations. EasternMining has worked with NNA to create a business intelligence toolset, on Microsoft’s Analysis Services, to help NNA manage sales, pricing, and tune its business strategy. These business intelligence tools allow NNA to analyze all aspects of sales, programming, budgeting and rating.
NNA (Nine Network Australia) is in the business of providing free to air television in Australia. Its business is about attracting viewers with locally produced and bought content and selling advertising space. Its business model is relatively simple, there are three main variables
1. Content
2. Ratings
3. Sales
These variables are managed by
1. Managing content, which it can purchase or produce locally
2. Attracting viewers through content, programming and promotions
3. Managing sales demand and revenue through price structures
NNA balances all three variables in its business strategy. It cannot manage content, ratings or sales in isolation. Buying (or creating) better content is only valuable to the extent that it attracts more viewers of desired demographics, which it can turn into increased sales revenue. These three variables (sales, ratings and content) are inter-related, like 3 sides of a triangle. One cannot be changed without impacting the other two.
The aim of NNA’s business intelligence tools is to assist in managing these three variables and to understand their inter-dependencies.
Australia’s free to air television ratings data is collected by ATR (Advanced Television Research) a company that monitors television viewing by demographic in about 5,000 homes throughout Australia. Hence ratings are very competitive, monitored closely, and effectively underpin pricing structures.
Pricing advertising spots is a complex task. The amount of saleable time is finite. Every program effectively has its own price based upon the expected ratings, time of day and day of week. The challenge is to get the market price for every spot, and maintain high fill rates. Setting prices too high will lead to low fill rates, and setting prices too low will miss revenue.
NNA has a base rate card (pricing structure) that it uses to set prices for its advertising spots. There are many discounts, bonuses and premiums that can be applied to the base rate card, making the pricing of any one spot a very complex function. Break number and position in break also contribute to pricing. There are opportunities to discount prices, give substitutes and offer bonus spots to assist in fill rates that can affect NNA’s revenue.
Programs with low fill rates can be discounted, used as program promotions or filled with bonus spots. NNA needs to keep track of pricing, bonuses and discounting to ensure sales persons are reaching quotas and without over discounting. Early warnings are needed for low and high fill rates so that prices and sales activities can be adjusted.
NNA’s main business application, “9 Vision”, has been tailor made to help with the operational aspects of NNA programming and sales. It is based on a Microsoft SQL Server 2000 database. This is a complex system with over 300 relational tables.
The second main source of data is the ratings information from Australia’s independent ratings agency, ATR, which provides ratings by demographic for all of Australia’s free to air television.
The data warehouse and business intelligence services are on a dedicated server. Every hour updates are taken from the 9 Vision application and incrementally added, or updated, to the data warehouse and the OLAP cubes. This way, business intelligence users will see data within 60 minutes of the business transaction. Ratings data is incrementally added each day. Business intelligence users can use Excel, ThinSlicer (a web cube browser) or Reporting Services (report portal) to view and analyze NNA’s information.
The value of NNA’s business intelligence tools is immense. It provides executives at NNA with up to date information on sales, budget, fill rates, ratings, discounts, customer activity, sales person activity, station, date, time etc. In essence it helps them to manage the strategic triangle displayed on the first page of this document.
A bespoke data warehouse and set of business intelligence tools were created for NNA and its business strategy. This is why NNA’s business intelligence tools are so effective and full of salient information. Modern business intelligence tools do not need to be a huge development effort. Jason McIntyre (NNA’s Chief Architect) Warwick Sharp (NNA’s Director of Marketing and Sales) and Richard Lees (contracted from EasternMining.com.au) worked together to build these business intelligence tools iteratively. The first iteration was 4 weeks, in which time they made available a basic sales cube with easy slice and dice capability. Interacting with NNA’s sales and marketing department, this cube evolved into a data warehouse that covered all aspects of revenue, budgeting, fill rates, programs, customer activity, sales persons, stations, ratings, viewer demographics etc. Warwick, Jason and Richard worked over a period of 5 months to bring the business intelligence tools to where they are today. The dollar cost of this development was just a commodity Intel server, 2 months of Richard’s time, and a copy of Microsoft SQL Server. No other software needed to be purchased. NNA is very happy with Microsoft’s OLAP technology. According to McIntyre; “Microsoft’s OLAP technology is the fastest and most scalable on the market. It’s the crown jewel of Microsoft’s server suite!”
There are two classes of business intelligence users at NNA
· Expert users
These users are highly skilled, understand the information and know NNA’s business model inside out. These users tend to rely on a rich cube browser such as Microsoft’s Excel PivotTable for OLAP. These users can slice and dice any combination of dimensions, nesting and filtering by multiple dimensions and reporting on multiple measures. The reports can be in Excel, so users are able to play with these numbers within Excel.
· Line of business users
This class of user is not technical. They do not need to be familiar with cube browsing, but they want easy access to information in their own context. These users have reporting tools created for them, where they can select a combination of parameters (such as date, client, program etc) and see a standard report, fresh from the cube. These users effectively have information delivered to them in their context. They don’t need to be technical experts, or even know what a cube is. The information is delivered to them through a web browser (or canned Excel report). These users are seeing fresh information as each report they request is fresh from an OLAP cube.
NNA are very happy with the value their business intelligence tools are providing. Within 6 months of commencing this business intelligence project, it has paid for itself. However, NNA can see further gains with incremental improvements to their tools. For example, they would like to have content costs (production and purchase) in the data warehouse. This would complete the strategic triangle and provide NNA with more sophisticated analysis before bidding for new content. Other areas NNA are interested in pursuing are price sensitivity analysis and data mining. Price sensitivity analysis is a relatively easy extension, as all the fill rates are already in the warehouse, it is just a matter of storing these rates over a “time to air” dimension. Data mining is also interesting. It might help with questions such as “who are the likely buyers for spots in program x?” and “how much revenue are we likely to gain from switching program y to Thursday at 8PM?”
These evolutionary improvements to the data warehouse do not have big costs associated with them. They only take a few days to prototype, and a few more days to integrate into the data warehouse. According to Lees, that’s the way a data warehouses should be; continually enhanced. “In a sense, any data warehouse or business intelligence application is never finished. The business is continually changing, and the tools it needs to drive its business are continually adapting.”
Keith Roscarel (NNA’s CIO) had the vision for a data warehouse. “
“ NNA’s requirement for business information has been constant and evolving over the last decade.
In the past we have constructed countless reports to keep up with the demands of various business units, especially our sales team. Our delivery cycle could not keep up with the business’ appetite for information.
So we set out to create a dynamic platform that would enable our sales team to answer their own information needs.
IT’s role moved from being ‘report writers’ to ‘data infrastructure providers’ and key Sales staff can utilise the platform, without IT intervention, to answer their information needs.
Information access is via Excel using Pivot Table Services, so training time is minimal.
… One of the critical elements was to achieve this without a large capital investment.
“ We wanted to leverage existing toolsets and platforms as much as possible to minimize NNA’s Cost of Ownership. Like many operating environments using Microsoft technologies, we have already invested in SQL Server and Office. With minimal hardware and services investment, we had the 1st major stage implemented in 5 months.
Future stages include extending this platform to all key areas of our enterprise and to investigate data mining opportunities”
Vance Lothringer (NNA’s National Director of Sales) is excited about the business insights available to him via the data warehouse. “It has allowed us to quickly & accurately analyse simple things such as program fill levels & yield. It allows us to monitor the performance of each Client, Sales Person & Sales Office in terms of total volume of sales & average cost of sale. It has given us the facilities to set program prices with more science, which has led to significant revenue increases on some of our major telecast events. In 2005 the data warehouse will give us the facilities to adjust our pricing in a dynamic ratecard environment, whereby program rates are adjusted on an episodic basis, based on the fill level of each episode of a program.”