Today’s column is written by Tony Ralph, director of data and marketing technology at Intuit QuickBooks.
With at least one notable exception, nearly all researchers predict the inevitable convergence of the mar tech and ad tech ecosystems.
Predictions aside, as a technologist working within and between these two worlds, I can say definitively that it is the exception when we find solutions that gracefully blend the data and performance advantages of each, not the rule.
This is unfortunate since it should be our obligation to provide prospects and customers with digital experiences that are graceful, consistent and harmonious across both paid and owned media.
When we find that dollars, data and platforms are bifurcated across these two realms, how can brands present a holistic and unified user journey experience?
Bringing Ad Tech Data Science To Mar Tech
Ad tech spans the systems used to optimize a potentially global dialogue between businesses and their customers via paid media. With large budgets and vast volumes of data, the space has evolved to encompass a world of real-time bidding, deep algorithmic exploration and optimization at scale via machine learning.
I’ve personally participated in rigorous experimentation that has shown that for long-running campaigns, pure machine learning, on the smartest platforms, regularly outperforms manual targeting and optimization. In the ad tech world, today’s contest of man vs. machine learning is tipping in favor of the machine.
Mar tech platforms, on the other hand, have evolved in a manner that relies on defining targeting and segmentation rules in advance of initiating a campaign. These rules are often operationally challenging to setup, test and debug.
Thus, introducing machine learning into the mar tech realm has the potential to enhance optimization and streamline day-to-day operations. In fact, early returns for campaigns I’ve been involved with show significant engagement increases relative to static rules-based campaigns, as well as ease in configuration.
Challenges to this approach may include finding resources that are abundant in the ad tech world: data and data scientists. In the case of mar tech, we’ve found that there may not be enough scale or fidelity of data to drive modeling exercises efficiently. However, within the mar tech realm, one can often supplement basic interactivity data such as opens and clicks with information gleaned from customers via web and product interactions or even event-level data from the ad tech space.
It is not uncommon for mar tech teams to operate in-house, whereas ad tech’s historical complexity often compels teams to work with agencies. However, it is not as common to allocate data scientists to mar tech teams. I would recommend committing data science efforts to mar tech in the same way they have been deployed in ad tech space. Aside from the modeling exercises, brands will likely find that enhanced reporting and analytics is a fast follow with this mindset shift.
Bringing Mar Tech First-Party Data To Ad Tech
Mar tech entails working with a much smaller number of known customers, contacts or leads. Rather than paid media, mar tech generally involves first-party experiences such as email or mobile campaigns, web experiences or offer programs.
As opposed to mar tech, where I suggest an increase in data science investment, it’s time to confront the opposite trend in ad tech.
As an avid ad tech hacker, I say this with some sense of regret. I feel nostalgia for the rooms packed with armies of people optimizing keyword bids, building in-house bidders and conducting nuanced bid-optimization testing. Alas, we’re finding these approaches are becoming less and less justifiable from an ROI perspective due to the level of machine-learning investment by the big ad tech platforms.
We shouldn’t, however, just turn over the keys entirely to platform partners. The strategy evolves to ensuring brands are setting appropriate conversion goals and supplementing the modeling with high-fidelity feedback in the form of their first-party data. This mastery of first-party data and customer segmentation is precisely the domain that mar tech teams have been refining for years.
In short, just as mar tech tactics can be enhanced with the data science approach we find in ad tech, the data and systems that drive mar tech efficiency can be extended as feedback loops to optimize ad tech campaigns. More importantly, this cross-pollination compels marketing and advertising teams to integrate tactics and harmonize campaigns that were once partitioned.
Imagine traditional mar tech campaigns across web, email or product experiences not dictated by targeting founded on static rules but infused with modeling output from an ad tech data science team. At the same time, ad tech campaigns might be seeded with high-value audiences, bidding influenced by lifetime-value indices or optimization algorithms fueled by detailed conversion data supplied via your internal mar tech tools and teams.
These opportunities are available to brands today with no need to wait for vendors to travel down the arduous road toward convergence, which has many technical and governance challenges. What may be more intriguing is how many brands handcuff themselves organizationally by splitting budgets and teams across paid and owned media. When the focus is on tactics and channels, this outcome is easy to derive.
However, when the focus is on unifying the holistic user journey for customers, brands may arrive at a different outcome. With this approach, the most profound convergence may be breaking down silos within teams and organizations to create a world where paid and owned tactics work in unison to deliver a brand dialogue that is intelligent and unified across every digital touch point.
Follow Intuit QuickBooks (@QuickBooks) and AdExchanger (@adexchanger) on Twitter.
This post was syndicated from Ad Exchanger.