The mere mention of a log (outside the context of web logs, or blogs) could instil reader boredom. Logs are a fundamental, functional part of IT infrastructure and are used for everything. According to Phillip Hayman (pictured below), head of technical services, sales engineering, Integral Ad Science, when you refer to logs in the context of ad tech, they start to become interesting. Here, Hayman explains to ExchangeWire how the use of logs in ad verification allow for transformative digital advertising.
There’s no denying that a log, in the context of computing, is not something that’s easy to get excited about.
As an automatically produced, time-stamped documentation of data events that relate to a particular system, logs offer precise records. They are also prolific – created by nearly all software applications that provide reports.
Useful? Yes, undoubtedly. Exciting? Not particularly. But the humble log suddenly becomes much more interesting if viewed in a different context: that of ad tech.
In the digital advertising sphere, a log is possibly one of the most important snippets of information you’re ever likely to receive from a technology platform, regardless of your position in the ecosystem. For years, technology businesses have been working to deliver a way of aggregating logs, and the vital insights they contain, but only recently have we started to see widely available logs via all distribution channels.
The opportunities are wide-ranging, and the benefits are clear, but the human skill, technical architecture, size, and weight of the commitment are often common blockers to squeezing out the true value on offer.
If you’re interested in ad verification, I could list 101 things a log file can do for you – but let me outline five here to whet your appetite.
1) Fighting fraud & debugging
Joining verification logs and programmatic logs allows you to monitor information such as ‘declared bid URL’ versus ‘actual served URL’. In other words, you can conduct your own domain spoofing analysis and catch fraudulent impersonators in the act.
They are also great for debugging. Aggregated reporting is always useful when debugging technical issues. Individual log records in effect are – and can significantly speed up – the isolation and resolution of issues.
2) Streamlining decision-making
Log data forms an accurate, comprehensive foundation for a raft of decision-making areas. These include, attribution — where marketers can use viewability and fraud data to make attribution decisions on an impression-by-impression basis — and programmatic trading — where logs can be employed as a layer of data collection for real-time bidding (RTB) at a later time.
The latter application is especially effective when used in combination with pre-bid segments, often based around predictive modelling and direct packages. Using logs for programmatic delivery, marketers are not only able to build rules around which domains to target to reach desirable segments, but also supplement the pre-validation process.
3) Creating audience maps & detailed reports
Want to know what your audience looks like from an operational perspective? Logs can provide all of the information you need. Operating platform, version, rendering engine, browser versioning, you name it.
You can also track audience activity with adjustable reporting options, such as using impression-level time-in-view data to build time-based reporting and buying options.
4) Customising metrics & data access
All advertisers operate differently, with varying goals and needs. By combining certain log file attributes, you can build quality metrics based specifically on individual unique advertising targets, and produce quality indexes. What’s more, you can ensure the way information is accessed is personalised too — log data can be entered directly into business intelligence platforms and used to create bespoke dashboards and utilities.
5) Amalgamating & mining data
By passing the same unique impression or user ID macro across multiple ad tech players, you can join ad server, verification, device and attribution logs at a single record level — consolidating all core insight into one easily accessible place.
And that’s not all. While a purpose-built user interface (UI) doesn’t always allow you to interrogate data in the way you might want, accessing and working with logs will give you the means to define trends and business analytics that matter to you.
So, although a log may not set your pulses racing, for anyone working in ad tech they can cause quite a stir. Providing a wealth of information and the ability to transform data models used across the entire digital advertising ecosystem into efficiently and effectively targeted advertising campaigns; logs can help boost revenues, reduce wastage, and delight audiences. And that’s pretty exciting.