Publishers have greater opportunities than ever to expand their advertising revenues by monetising their content on the internet. As they exploit these advantages, effective coordination between their internal teams (including ads ops, client services and account managers) and external clients (including advertisers) becomes increasingly important. But, at the same time, it is excessively complex due to the overwhelming availability of marketing information that is difficult to capture, analyse and distribute in an efficient manner.
By broadening their reach to include managing digital marketing campaigns, working with ad networks in the selling of remnant inventory and providing agency-type services, publishers find themselves having to trudge through a torrent of complex and voluminous marketing data in the day-to-day management of their campaigns. Accurate and timely reporting provides valuable and strategic insights into campaigns. But, gathering, aggregating, and analysing the data is a time consuming, resource sapping and costly undertaking for most publishers.
Different audiences and needs -- same goalAll parties involved in the publisher's world are reaching for the same brass ring -- accurate, relevant and timely data that can be used to improve ROI (return on investment). Who in the publisher organisations needs the detailed, timely reports?
- Advertising and agency clients
- Ad networksSales teams
- Search engine marketing teams
- Ad operations
- Financial staff
What are the businesses' needs for the data throughout the phases of a campaign, from pre-campaign planning to post-campaign analysis? The different stakeholders need information for business critical tasks including: forecasting, inventory management, under and over delivery statistics, insertion order discrepancy reporting and billing.
The types of available statistics that can be capitalised upon to improve ROI include:
- Site side
- Impressions
- Clicks
- Keywords
- Traffic source
- Demographics
- Geography
- Session visits
- Cost: CPM, eCPM, CPC
- Revenue
- Users
The problem is that marketing data originates from multiple sources with varying refresh times making it time consuming and costly to download, format, combine, analyse, and distribute. The solution is an effective media analysis platform that is holistic in its view of marketing data; easy and cost effective to acquire, learn and use; and able to provide accurate, timely and flexible reports.
Exploring data sourcesThe good news is the level of detail, frequency and amount of data available to publishers today is greater than ever. The bad news is that all this data is difficult and time consuming to collect, store and analyse. In general, all this data can be divided into two major categories: summarised data and granular level data.
While summarised data is smaller and easier to work with, granular level data captures more details about the publisher's campaigns providing unprecedented analysis capabilities for the publisher. For example, when reported in a timely fashion, data can be used to make decisions that optimise campaigns more quickly. But, as mentioned, reporting this key information is difficult because the data originates from multiple sources including:
- One or multiple ad servers or display and rich media ads
- Emails
- Search engine data
- Site analytics data
- Attitudinal survey data
- Billing and accounting data
The challenges include: a lack of time to generate reports from multiple sources and a negative impact on resources needed to analyse data.
Tackling data issues
A key challenge for the publisher is effectively gathering, combining and storing the marketing data from various sources. Several issues need to be addressed to ensure the accuracy and usefulness of the wealth of available data.
- Timing: Sources have different refresh times, making it difficult to compare data accurately
- Formats: Some data is tabulated by commas, others by tabs or pipes, making it difficult to combine data
- Automation: Downloading data in an automated fashion is not possible from some sources
A platform that aggregates, normalises, cleanses, and QAs data from disparate sources in an automated manner is key to solving a publisher's marketing analysis needs. Publishers use several methods for downloading data from the various sources:
- Manually
- Screen scraping
- Automatically through APIs such as DART and Google
Some companies are devising innovative ways of further automating downloads using ETL (extract transform load) tools especially designed for digital marketing platforms. Reporting is also difficult because extracting valuable information from large datasets is sometimes cumbersome or even complex.
Often the owner of the dataset is looking for 5 MB of relevant information in 1TB of data, especially when dealing with ad server log files. Sophisticated algorithms can help publishers to search for the desired information in a timely manner.
Marcus Harding is U.K. business development manager, Theorem.
