This article is the second in a two-part series. In the first part we looked at how reporting and visualization tools hold some businesses back. Here, we'll lay out a path for something better.
It used to be that you were ahead of the game just by having your data aggregated and normalized. Then the goalposts moved and you needed to have the data available and at your fingertips on any device at any time. Today, you need to be able to read and react to the story your data is telling the moment it shows up.
Protecting Revenue with Deep Data Monitoring
To achieve that level of data-responsiveness, publishers need to move from dashboard management to proactive business monitoring. That's a tall task, but it's not impossible. In fact, with the right technologies and methodologies in place, it's probably a lot easier than you'd think. In its simplest form, it's a recipe that requires just four basic ingredients.
These are the 4 key capabilities that publishers must posses in order to level up their data monitoring practices and assert real-time supervision and control over a complex business environment.
The ability to process data at a breakneck pace
Data — often from disparate systems — will need to be ingested, normalized, organized, and processed with staggering speed at extraordinary scale. This speed must extend not only to preparatory processing, but to all the other capabilities outlined on this list.
Since there are so many contributing factors that fluctuate so frequently in ad monetization, you need to be able to spot and make sense of issues as soon as they appear. That fundamental time-sensitivity is why automation is a vital ingredient in smart, proactive monitoring.
Publishers need a way to keep a vigilant and intelligent eye on their datascapes at all times — without having to wait on human operators or lengthy manual reviews.
The ability to precisely distinguish between normal & abnormal data
Your data can tell different stories depending on how you look at it. At the same time, not all those stories will be equally true. Which is why it’s so important that the data be approached with a deft hand and an experienced eye.
Suppose, for example, you have a digital property that produces a lot of content on sports news.1 Interest in each sport you cover would naturally wax and wane with the coming and going of its season. Even more granularly than that, interest in each team would ebb and flow with the particulars of their schedule and matchups. This is what’s known as seasonality.
Even if you have a good awareness of seasonal impact — say if it’s currently offseason for three of the four main sports you cover — that awareness will mostly show in your expectations for directional shifts. When it comes to the magnitude of those shifts though, the awareness is rarely very specific.
If you expect a medium-sized drop in traffic to have a knock-on effect on revenue over the offseason, you won’t be surprised when it happens. But that doesn’t guarantee that everything is working as it should be. To the contrary, vague expectations can provide cover for problems to fester.
Smart monitoring, by contrast, would be much more precise. Rather than a general awareness, the system have a very specific expectation for the impact of seasonality on your key metrics. It would know, for example, to expect a traffic drop of 12-18% through July and August — with a knock-on effect on revenue between 8-14%.
In such a case, if you experience a 19% decline in revenue, you’d know — despite the familiar look and feel of the trend — that something is actually very wrong. Somewhere along your value chain, there is an inefficiency or mis-calibration that's turned into a 5-11% revenue leak . Without smart data monitoring, this leak would likely go unnoticed.
This is just one example, but it does a good job showing how revenue is lost when publishers fail to precisely distinguish between normal and abnormal data fluctuations.
The ability to tell if anomalous data events are significant to the business
Even when data is abnormal (anomalous) and statistically significant, it won’t necessarily impact the business in a way that matters. And even when it does impact the business, it won’t necessarily be actionable.
This is where an understanding of business and operational context comes into play and why it’s so important that data insights be translated to business problems and opportunities.
This capability renders obvious the gap between data dashboards and true monitoring. Reporting and visualization tools simply cannot match the data to the business context as needed to understand the practical implications.
True monitoring solutions, on the other hand, apply your operational logic to the situation and present the data along with its business context. That ensures the user gets a clear view of business problems and opportunities — without false alarms.
The ability to trace issues to a root cause.
A suitably advanced and operationally intelligent monitoring system would not only detect problems and opportunities, but reverse-engineer them to understand what brought them about and what actions should be taken.
Such a solution would interpret your data around-the-clock — summoning the situational intelligence and subject matter expertise needed to pull the story from the statistics.
It would scan the entire datascape from end-to-end in view of all interconnections and dependencies — distilling the mess of information into genuine improvement imperatives.
Only with such a proactive monitoring system in place can managers be sure to see and understand the things that matter in a timely manner.
On Top of Your Data and Ahead of the Curve
Data reporting and visualization is good, but when it comes to real-time problem detection and correction, it’s not good enough. Without a true monitoring tool, even the most capable professional will be at a profound disadvantage trying to oversee and optimize a large and complex revenue operation.
Rather than serving as the springboard for open-ended investigations, a good monitoring system will empower its users to act quickly and surgically. Smart data monitoring analyzes each data point in the context of all related monetization factors and the whole ad stack — seeing everything, missing nothing — keeping your operation running smoothly and your revenue flowing freely.
- Of course, seasonality exists in every industry, not just sports. That said, sports offer a particularly strong example of the acuity required in establishing precise and well-defined data baselines for all relevant metrics and dimensions. This is because sports seasons do not follow a fixed yearly calendar. As such, the arbitrary and erratic nature of their seasons make it much more difficult to accurately account for and model seasonal impact.