Seven Secrets of Highly Successful Vendors
How ad technology vendors capture, monitor and use data can have a tremendous impact on their success. Let’s look at seven practices that can transform a vendor’s data collection processes and business model for the better.
1. Invest in Data Operations Talent
In my last blog entry on the metamorphosis of the ad trafficker, I focused on the notion of marketers and agencies investing in talent to manage and optimize the flow of data to the third-party vendors they work with. Vendors, however, have just as much to gain by more effectively receiving and utilizing data.
Today, advertisers are making their data more widely available at very granular levels of detail. That introduces great opportunities for vendors to provide a wider array of services and strategic insights, but making all that data actionable requires management of abstract concepts like relationships between disparate data elements. A vendor’s “data operations” team would, for example, understand when to aggregate data instead of keeping it at very granular levels in relation to the client’s particular use cases.
The human brain’s ability to act on the best interest of both the client and the vendor by applying semantic logic to abstractions is more powerful than what a plug-and-play computer model can provide. Vendors need talented data management experts to effectively guide their businesses.
2. Invest in the Data Collection API
BrightTag’s “Road to Here” chart illustrates that the history of the tracking tag is messy. In most cases, the tracking tag is just a lightweight API that uses a cookie to establish a user ID and collect some data. In the past, these tracking tags were very limited in scope, because they were implemented on small numbers of pages and collected little data.
As proven by the proliferation of hundreds of distinct services that marketers are enthusiastically testing today, a vendor’s data collection efforts have had to rapidly expand in scope as a response.
Thus, when thinking about how to ingest the data that clients make available, vendors should always assume that valuable data could come from any experience the client provides and with any level of detail. Data could exist on any page, within a Flash or AJAX experience, within a mobile application, or even from in-store, kiosk, or offline experiences. Data could be very granular, and on any given page, there may be hundreds of unique data elements available. Finally, the very notion of the tracking tag itself is going away and being quickly replaced with newer forms of data collection that are much more agile and responsive to evolving customer experiences.
With these assumptions in mind, vendors must consider the very mechanics of data collection and how to transform it into meaningful actions.
3. Invest in Architecture
The API, or “tag”, is the front-end of the data collection experience. As data comes in, the back-end supporting architecture and storage should be robust enough to make it actionable immediately in order to provide truly real-time services and insights for the client. I’ve observed a few important architectural designs that have positioned vendors really well as their clients continue to make more data available. Some best practice examples include:
Use of a server-side data store. Relying on a browser cookie as the primary means of data storage, while simple, creates a number of challenges and inefficiencies and is certainly not future-proof. In recent years, we have observed many vendors invest in server-side data stores to facilitate ad serving (in the form of RTB). That model has proven to be a highly efficient means to enable different technologies to speak to each other, and that same concept should be expanded to all aspects of a vendor’s technology stack.
Replicate data real-time across multiple data centers. As data comes in from any source, it should be distributed to all data centers in real-time. This is a key design consideration in making the data actionable for anyone in the world at any moment.
Scale it from the beginning. With this evolution in data availability, there is simply much more data to ingest. Thus capacity, redundancy, failover procedures, and all the usual things that go into managing high-volume, low-latency systems should be applied specifically to inbound data collection. Queries Per Second (QPS) should never be a gating factor for getting started on a creative new program with a client.
4. Integrate the Stack
I often view our ecosystem as one that fragments just as quickly as it contracts. For every new product release announcement, start-up capitalization, or pivot, there is an acquisition or merger. But one consistent thing I’ve heard from vendors is that they want to solve a broader set of problems that builds on initial core services. This is often done by introducing new layers of technology ostensibly integrated into a core application.
As vendors build new solutions or acquire technologies, they should seamlessly integrate the front-end and the back-end simultaneously. On the front-end, a single data collection point should suffice to collect data for the entire stack. For example, marketers shouldn’t have to implement multiple tags with different IDs and cookies, to enable different services. Meanwhile, on the back-end, all aspects of the stack should work on the same user IDs and the same data centers without the need to retrieve any information stored in a cookie or wait for a batch file to be delivered.
5. Keep track of what competition is doing
We’ve established that marketers want to test out new and innovative programs, and their ability to do so is largely driven by their ability to better control and optimize the flow of data. A very valuable asset for the vendor, that data enables them to build on their existing relationships by adding new services, creating new technologies or providing new insights. So it’s important for vendors to frequently step back and assess how competing vendors make use of data. Doing so may establish a much deeper, much more strategic relationship with clients and shape a very strong product roadmap for years to come.
6. Talk to Clients About Their Data
The concept of data being affixed to a tag that is affixed to a client’s webpage is undergoing a rapid paradigm shift. That shift causes data to become available any time, across all touchpoints, at all levels of depth and under complete direction by the client. In other words, advertisers become much more empowered with their data and want to make it available if it helps drive optimization, insights or anything else that adds value.
Vendors have the opportunity to create a highly strategic dialog with clients about leveraging their own data. I frequently browse a client’s site and get a sense of how it works before speaking with them. A simple exercise like that, without worrying about tactical execution of any potential strategy, often leads to very productive, value-added discussions focused on testing and validation.
7. Invest in Business Development: Partner with Key Enabling Technologies
Data is very complicated, fragmented, brittle, and often difficult to scale or translate. It is also the lifeblood of a marketer’s campaigns and the backbone of technology innovation. Fortunately, a vendor doesn’t have to solve all the problems on its own and risk shifting focus away from core service offerings in order to work with a marketer. Today, many technologies are emerging which specialize in operational, integration and connectivity pain points, allowing a vendor to just focus on strategic deliverables and implementation. Investment in exploring partnerships with enabling technologies can pay off in the form of a future-proof, integrated technology stack that provides a robust set of meaningful solutions for a client.
Vendors can leverage these best practices in data management to run their businesses more effectively and efficiently. In doing so, they will likely transform some of their relationships with marketers and other clients and grow their reputation as a trusted partner.
This article is reprinted with permision from BrightTag.
Pondering the future of TV and digital video? OPS TV will bring digital advertising leaders and ops professionals together to discuss the intersection of digital video and TV advertising. Register today for OPS TV, which will be held July 11, 2012, in New York.
SVP Partner Development, BrightTag
Todd Chu has proven expertise in the areas of data strategies, tag management, server-side data distribution, systems integration, attribution, ad serving and web analytics. As SVP of Partner Development at BrightTag, he works closely with global enterprise brand leaders and their digital solution providers to drive online marketing results. Todd is recognized by clients and partners alike as a creative problem solver and an invaluable resource who can seamlessly bridge the gap between business and technology. He has held senior sales engineering, client services and business development roles at a number of leading advertising technology companies including Doubleclick and Atlas. Todd has a B.S.E. in Industrial and Operations Engineering from the University of Michigan.