Commentary

A Case For Partnering With The Best

The speed of innovation has changed the way we think about the age-old question do we partner or build it? We’re now convinced it’s much smarter to partner. How many old cellphones do you have in your drawer? We have quite a few. We also have old iPods, tape recorders, wireless phones, and even tablets.

When the speed of innovation was much slower, we could own a CD-player for several years before it was outdated, but no more. Now technology, including business and consumer technologies, quickly become outdated. The question is not when, but rather how fast.

Most small-to-midsize agencies have always gone the partner route. We had to. We couldn’t afford to build proprietary tools because we just didn’t have the tech and research budgets. Instead, small-to-midsized agencies spent their resources developing proprietary approaches to data management and inventing specific processes that led to insights. Insights become our competitive advantage.

Instead of trying to build tech, which requires focus, deep expertise, and capital, we hired data analysts to help make sense of all the output. We invested in traders who knew how to pull the levers and media technologists that could vet the marketplace.

advertisement

advertisement

We realized long ago we couldn’t compete with well-funded VC-backed start-ups filled with engineers to build our own tech stack. That’s a different kind of business. Instead our strategy is to partner with the best ad tech inventors. Instead of building it, we were going to service the hell out of our clients by recommending the best partners and providing the smartest insights.

Our stack includes solutions in data visualization, programmatic buying, attribution, fraud, viewability, and privacy.

Data Visualization:
We use Good Data as a data warehouse and custom dashboard tool, which allows us to more easily derive cross-channel insights and analysis. Because of the flexibility of the platform, we are able to incorporate marketing data from across a variety of different disparate sources (paid, earned, and owned). Additionally, this platform serves as a singular point of cross-channel reference for all agency and client historical reporting.

Attribution Modeling:
We’ve had a relationship with Visual IQ, the top-rated cross channel attribution platform (Forrester Wave, Q4 2014), since 2009. Visual IQ’s marketing attribution methodology starts by analyzing the entire stack of digital marketing touch points as well as the features (i.e. frequency, recency, contribution) of those touch points. This encompasses a variety of channels, including but not limited to: Paid Search, Programmatic Advertising, Publisher Direct, Email/CRM, Paid Social, and Organic Search. 

In addition, Visual IQ’s flexible technology stack has allowed TMK to port attribution data into buying platforms (i.e. Marin Search, Demand Side Platforms) for real-time optimizations. Utilizing Visual IQ’s attribution data, TMK has consistently outperformed the ambitious year-over-year cost per acquisition goals set by our largest ecommerce client. We also work with another third-party provider, Encore Media Metrics, a partner of Sizmek, for attribution modeling with other ecommerce clients.
 
Ad Server Based Attribution Models:
The DoubleClick and Sizmek Attribution Suites are currently used by several clients and supplement third-party attribution modeling with intra-ad server specific data including frequency, unique reach, path analysis, and creative sequencing.

Web site Analytics:
In collaboration with our Data and Analytics group, we use Web analytics platforms for both campaign optimization and reporting. Within this ecosystem, we have experience with Adobe Analytics and Google Analytics products. We believe analytics data is more valuable than simply measuring site-side performance. We utilize site analytics data to make landing page decisions (A/B and multivariate landing page testing), for creative development and targeting, and to discover new partnerships through referral traffic sources.

Viewability and Fraud Measurement:
We are committed to measuring viewability and ad fraud on behalf of our clients, holding publishing partners accountable for providing in-view ad impressions and verified human visitors.  We have partnered with MOAT and Integral Ad Science for viewability and with WhiteOps for ad fraud detection.

Cross-Device Measurement & Analytics:
As an agency, we believe that mobile and tablet measurement and attribution should reside within the same measurement ecosystem as desktop programs. Rather than maintain separate mobile silos, we’ve partnered with TapAd and others to integrate with existing Visual IQ implementations for cross-device measurement in cookie-less or cross-browser environments.

Demand Side Platforms:
We continue to invest in programmatic and have recently taken most of our programmatic in-house, where it resides inside our former search group, which we’ve re-branded the Programmatic Media Buying Group. As such, we need DSPs that are flexible, connected to numerous DMPs and supply sources, with a robust data visualization interface. Our solution was The Trade Desk.

They’ve developed a bidding solution that is both sophisticated and fully transparent and they’re a nimble partner.  Another partner we’ve come to rely on is MediaMath, especially for their capabilities in Asia, a region of growing importance to the agency.

It’s taken a long time and a lot of work to vet these partners, and we continue to meet and test new partners everyday. Our relationship with Ventures, the investment arm of kbs+, has also proven to be a huge advantage to surface new opportunities, as they see hundreds of deals every year, many of them in the adtech space.

One thing is certain: Today’s tech stack will not be the same in six months, which reflects the speed of innovation. In 2015, a special focus will be our mobile stack. The more we meet partners and the more we learn from our partners, the more we’re convinced it’s much better to partner than build.

Next story loading loading..