The Two Big Takeaways from Programmatic I/O

by | May 7, 2019 | Impressions

Last week I had the opportunity to attend AdExchanger’s Programmatic I/O conference in San Francisco. Programmatic professionals from agencies, brands, publishers, media companies and technology providers gathered for two days of presentations, panels and workshops.

It was a fantastic experience and from it, I had two key takeaways:

Setting the Right KPI Expert Guide Subhead: Educate

But There Is Still Plenty of Work to Do

The explosive growth and maturation of the industry over the last 10-15 years has been mind-boggling.

For example, TripleLift’s CEO and Chief Strategy Officer Ari Lewine noted that non-social native programmatic spending has eclipsed $10B annually even though the technology behind it has only been around for five years.

That’s just one of the many examples of innovation that’s seeing heavy adoption in ad tech today.

However, with every new inventory source, data stream, targeting capability, etc. comes myriad challenges that advertisers must overcome.

During his presentation, StubHub’s Sachin Puri outlined just how far cross-platform attribution measurement and optimization has come in the last few years.

But, in the face of ever-evolving attribution goals and privacy regulations, Mr. Puri urged the industry to come together to develop better cross-platform ROI measurement and optimization.

This situation sums up a theme that was apparent throughout the weekend:

There is no shortage of exciting innovation with technology, inventory, and data in the ad tech space.

Yet, despite all the strides the industry has made, there are plenty of problems we need to solve.

Multi-Level Complexity

Each individual component of the ad tech machine has its own intricate complexities that make it all the more difficult to bring all of those components together holistically (see LUMAscape).

Joanna O’Connell from Forrester Research kicked off the second day of the conference with a deep dive on the current state of one of these components: programmatic creative. She argued that tools to support teams building programmatic ads are stuck in the past.

Her argument was that the creative process must evolve to be faster and more:

+ Iterative

+ Data-infused

+ Collaborative

+ Omnichannel

She reasoned that improvements in these areas would allow advertisers to:

+ Use data to run more relevant ads

+ Enjoy faster, more efficient speed-to-market

+ Exercise centralized control over distributed creative tasks

Progress is happening and there is plenty of opportunity for the tech providers and marketers solving these problems on the front lines.

But just like the creative process has major wants and needs, so too does every single other component of the ad tech space—audience optimization, data activation, in-housing, cross-device tracking, buying, attribution, measurement… the list goes on.

Setting the Right KPI Expert Guide Subhead: Educate

O’Connell’s comments are in regards to the Creative Optimization component, but can apply to every component in this image, originally found in our A Beginner’s Guide to Programmatic Advertising article.

All that is to say that in such a complex, fragmented (buzzwords of the week at the conference) ecosystem, hundreds of speed bumps still lay between advertisers and their audiences.

And even when individual solutions are put in place, those disparate solutions need to work together for successful execution. For instance, what’s the point in implementing a creative management platform if you don’t have a tool to go buy ads?

In summary, the industry will continue to innovate at a fast pace and will bring new, unforeseen difficulties. As long as the industry as a whole stays agile, talks transparently about issues, and proactively collaborates to solve them, it will continue to innovate and prosper.

Setting the Right KPI Expert Guide Subhead: Educate

And Isn’t Clearing up Anytime Soon

As has been the case for the last five-plus years, bringing media in-house was a hot topic at Programmatic I/O.

There were five sessions dedicated to the topic run by some of the biggest brands in the world — Uber, HP, Wayfair, DraftKings and Williams-Sonoma.

You’d think that a major tech company like Uber would have their in-house programmatic capabilities dialed — and they do… for the most part.

Here’s the thing: “in-housing” isn’t an all-or-nothing affair.

The concept of standing up an “in-house programmatic team” means different things for each and every one of the brands that spoke about their in-house ad tech capabilities.

It was notable that brands like Uber admitted that while they ultimately aim to own as much ad tech expertise as possible, they aren’t “fully in-house” and might not ever be. That’s music to the ears of agencies and tech partners alike who have been stressing about the in-housing trends for years now.

Ultimately, the brands chosen to represent the current and future state of in-house programmatic seemed to agree that in-housing means that brands can (or should) take control of the things that make sense for them at that time, and depend on partnerships for the rest.

For some brands that might mean they bring tech contracts in-house and depend on outsourced management of the platforms. Some brands might decide to build their own bidder— but others might find that the ROI doesn’t make sense to do so.

Ultimately, brands want to get closer to their owned data and better understand their ad strategy and audiences.

They’re also looking to get past proxy metrics and focus more on revenue and ROI. Agencies or tech partners concerned about the negative impacts of the in-housing trend would be smart to listen to that and adapt.

Services Ain’t Going Nowhere

Uber’s Bennett Rosenblatt talked about how Uber brought its bottom-funnel acquisition campaigns in-house first in order to “sell the dream” of running media in-house to its executive team.

Rosenblatt admitted that programmatic display ads aren’t best suited for bottom-funnel acquisition campaigns, but it was necessary to start with that proof of concept so he could ultimately start focusing on things like re-engagement campaigns and “rider to eater” conversion campaigns.  

Interestingly, Rosenblatt admitted that when they initially brought acquisition campaigns in-house, they weren’t meeting the performance benchmarks set by previous partners. He was okay with the short-term losses knowing that it improved not only Uber’s campaign performance but also its in-house expertise and data.

Uber has been successful in mid- and lower-funnel programmatic campaigns, but when they started taking on more upper-funnel branding efforts in 2017, they realized that it wasn’t as effective as they needed it to be. At that point, they decided to re-engage agency partners for some of the upper-funnel branding campaigns.

Rosenblatt used this example to illustrate the fact that getting help with things you’re not great at makes smart business sense.

He also noted that Uber was looking for a true partner in that situation, so they could continue to learn and grow and maybe even bring branding campaigns back in-house if desired.

HP’s Freddie Liversidge noted that his company’s reason to bring programmatic capabilities in-house was more about performance than cost-savings, although cost-savings was definitely a factor.

Liversidge felt that if brands only look at cost, they’d have a “lean team but not necessarily a great-performing team.”

His team’s goal was to get closer to the data to better match performance with revenue and allow his team to be more proactive.

That said, it doesn’t happen overnight, even for a $30 billion company like HP, which started the in-housing process two-and-a-half years ago by bringing ad server and DSP contracts in-house.

This helped them meet their objective of getting closer to the data, but they still used agencies for buying and activation. It wasn’t until the end of last month (April 2019) that they launched their first in-house campaign.

Similar examples permeated the various in-housing sessions. Williams-Sonoma is still leveraging its tech partnerships for strategy work.

Despite having 100 engineers on its marketing team, Wayfair has decided to buy DSP tech rather than build its own. And DraftKings onboarded a data science partner to help build bid algorithms.

At its surface, building an in-house programmatic team can seem like an all-or-nothing endeavor, but the truth is that agencies and brands need to first ask themselves what they want to accomplish.

Then they can decide (assuming in-housing even make sense for their business) how best to execute an in-housing strategy uniquely suited for their needs.

And despite the list of brands mentioned here, you don’t need a billion dollar market cap to make in-housing make sense!


Subscribe below for more of the latest news in ad tech and digital advertising from your friends at LumenAd.

Related Articles

Resources for Marketers: COVID-19

Updated: 4/24/2020 It’s April, and working from home and sheltering in place is now the norm for most Americans. (Thank you essential workforce for all that you are doing right now!) As we wrap our...

VIEW ALL Impressions resources

Subscribe to the Impressions Newsletter.

Get the latest news, tips and best practices delivered to your inbox. 

LumenAd logo      © Copyright LumenAd 2020
111 North Higgins Avenue, Missoula, MT 59802  -  406.552.1022  -