“You’re Just an Ad Tech Person” — How Advertising Produces Great Product Managers

Demanding Users, Rapid Innovation, and Data Saturation in a Polarizing Field

Adam Hecht
4 min readNov 10, 2020

I don’t use an ad blocker on my work laptop. Why would I? I have to make sure I can view our test creatives that target only my user_id (we built an internal tool to autogenerate a user_id and inject it into the database our campaigns use for targeting) to confirm that our bidding technology is receiving the requests from our test publisher site with the right parameters and responding with the right creative. Then I need to do it a few more times to confirm we are storing the information related to frequency and recency and responding appropriately. Then I need to switch locations on my VPN and reload the test creative a few more times to confirm that the geographic IDs we use (at the zip, city, DMA, state, and country levels) are being read and targeted correctly. Then, of course, I need to adjust the bid amount and make sure my user_id is being priced appropriately. Once all that is done, I switch over to the B version of our testing UI, and run all of the above again with a slight tweak to how our future Users may add geotargeting to their campaigns.

I don’t think anyone would deny that ad tech is complicated — “needlessly complicated” is a phrase that isn’t an unfair way to describe the revenue model powering a large portion of the internet — but I do think many folks are guilty of assuming that ad tech product managers are not in the same league as PMs who build consumer software or more “enterprise” B2B software. I spent the first 7 years of my product management career building enterprise B2B tools using a SaaS model, and the skills I need to succeed in ad tech are extensions of everything I learned there, with a smattering of consumer product management techniques as well. Some of the most talented folks I’ve ever met in the PM field, and many of the most famous names as well, have significant ad tech experience on their resume (they just might not call what they did at FB/Google/Amazon “ad tech”). It boils down to four main concepts:

  • Users of ad tech are technical, analytical, and skeptical, with many software options at their fingertips. Core to my vision of Product Management as a discipline is the well-known idea that users want and will pay for software to make their lives/jobs/tasks easier, and with ad tech the key is being able to build tools for each of the personae that will engage. Buyers, Traders, Campaign Managers (all coming in both internal and external flavors for ad tech with a managed service and self service business), along with the usual SaaS requirements for executive sponsorship, are all part and parcel with user engagement for ad tech. Ad tech product managers deal with some of the toughest users AND buyers around, and they need to have a significant amount of empathy while knowing how to push back against each persona. It isn’t an easy tight rope to walk.
  • Because ad tech users have so many options for their software, every company needs to innovate rapidly. By the time you release your latest features, your competitor puts out their v2. The pace of innovation is breakneck. Your customers’ customers operate quarterly (vs the multi-year contracts of typical enterprise SaaS), so you can’t afford to have a 12 month roadmap that is immutable within 6 months. This forces ad tech product managers into all the experimentation techniques and metrics analysis that typically accompany high growth consumer products, with all of the stability, security, and business requirements of steady enterprise technology.
  • Ad tech is full of data. Billions of queries per second flow through bidders, and each one needs to be parsed, checked against campaign parameters, discarded if mismatched, bid on if matched, and it all needs to happen in ideally under 50ms. How do we store this data? How does it inform our bidding logic? How much control do we give our users over targeting? What happens if a parameter is missing — skip or use machine learning to predict? Can we use machine learning to forecast availability? What about pricing? Ad tech product managers get a crash course in data science and machine learning when starting to look at precisely what makes the software easy to use and innovative for our technical, analytical, and skeptical stakeholders.
  • Advertising is a polarizing field. There are a whole lot of people who think that advertising has destroyed the internet, and whether that is true is out of scope of this discussion. But working in a semi-controversial field gives product managers the opportunity to step outside of themselves and understand how what they’re building impacts the world beyond their users and their company. It is scary and humbling, and makes for more thoughtful decision making when building future products and features.

Ultimately, domain knowledge reigns supreme when evaluating product managers for a given role. But there are experiences you get (in every domain) that are translatable to others. I believe that ad tech is a field that every product manager should experience at least once, especially if the product manager hates advertising. Learn it and change it from within, then take the significant learnings you’ll earn and go forth to the next adventure.



Adam Hecht

Lover of Product Management as a discipline, software as a service, data science as a hobby, and Iron Maiden as a band.