Customer Case Study: Smart Forecasting Tool | SmartAdServer

Customer Case Study: Smart Forecasting Tool

Customer Case Study: Smart Forecasting Tool

The client, currently operating in 40 countries, is one of Europe’s leading media companies. They’re not only one of the biggest newspaper publishers in terms of circulation and sales, but also one of the continent’s leading digital publishers.

In Q2 of 2015, digital media accounted for over 60% of its revenue. With the recent purchase of a major international online publication, the company now reaches almost 200 million unique visitors. It delivered over 10 billion ad impressions in Q1 of this year, making it one of the top worldwide in terms of reach.

The leading publisher began using Smart AdServer’s ad serving solution six years ago. Since then, the two have been working closely to build comprehensive ad strategies. The media conglomerate currently utilizes Smart AdServer technology for display, video and mobile, selling through both direct and programmatic sales channels. They were one of the first to adopt Smart AdServer’s forecasting tool in January of 2009.


Over the past few years, the company has invested in dynamic growth of its digital activities, resulting in an increase in revenue and earnings in 2014.Challenges for Smart AdServer:-Supply accurate forecasting for over 500 employees who are spread across 100 sites worldwide and who manage between 2,000 and 5,000 campaigns per month. The tool is particularly important to sales teams, who needed to sell the maximum volume of ad space without exceeding available inventory.-Provide the sales team with accurate information about available volume, giving them the ability to sell the appropriate amount at the correct time.-Offer insight about overbooked campaigns and their level of importance, allowing publishers to take necessary actions to ensure campaigns objectives are met.-Offer teams the possibility to combine multiple targeting criterion without sacrificing precision.-Provide the publisher with the ability to forecast fine-tuned audiences and then sell behavioral campaigns at a higher rate.


Predict future traffic volume and audience composition to increase overall sales performance
It was essential for the group to have access to a reliable tool that allowed them to accurately forecast available inventory for key audiences worldwide. Over 40% of all campaigns utilize gender targeting, and geotargeting campaigns are also popular.

Smart AdServer built a tool that allowed the combination of multiple targeting criterion to accurately estimate the number of available impressions and, in turn, optimize inventory sales.

Offer an easy-to-use and intuitive interface that provides accurate information about campaigns
The media group has 200 sales representatives, 10 yield managers, 13 traffickers and 260 online marketing managers who use the forecasting tool. At least 50% of these users have relied on forecasting tool information.

Smart AdServer ensured each user could effortlessly navigate the interface to gain access to crucial information without delay. Easy-to-read symbols, pie charts and lists keep this diverse group of users up-to-date with the progress of campaigns. The sheer number of campaigns obliges Smart AdServer to provide the publisher with a quick and easy way to see which campaigns are running at the same time, which are under delivery and which are overbooked. This overview of booked campaigns sorted by priority level helps campaign managers adjust ad strategies, which has resulted in the implementation of more ambitious and aggressive sales campaigns.

Create a forecast algorithm that takes into account changes in site traffic
Utilizing its machine learning algorithms, Smart AdServer first predicts site traffic. It then runs a full ad serving simulation, taking into account all campaign targeting criterion and diffusion rules. This methodology enables a precise forecast.

Inventory changes constantly, depending on the time of day, day of the week and season. Smart AdServer’s algorithm takes these variations into account, but there also are unforeseen spikes of activity due to special events. Including these variations in the calculation leads to wrong predictions. Smart AdServer has implemented a “spike detection” algorithm to remove the influence of such events. The result is an accurate estimate of available inventory.


-Easy-to-use interface provided crucial information at the click of a button

-The precise algorithm ensured an accurate estimate of available inventory and audiences

-Managers adjusted ad strategies using overview of booked campaigns

Download the Forecast Case Study in PDF format

Access the Forecast Case Study via Slideshare