AdTech AI Files

AdTech AI Files

By Steven OhrnsteinBusinessMarketing
Download on the App Store

AdTech AI Files episodes

  • #25 | Google or Bust
    Will Google’s Ad Empire Be Broken Up?
    When I first started in programmatic, 2nd price auctions ensured advertisers paid a fair market rate. You could bid confidently knowing you’d only pay a cent more than the next-highest bid.
    Then, about 8-9 years ago, Google led the shift to 1st price auctions (where you pay exactly what you bid) & enabling their publishers to create dynamic bid floors (to artifically increase the clearing price). This fundamentally changed how media was bought, increasing advertiser costs while boosting Google’s profits.
    My First Exposure to Googles dominance was in 2016.
    A programmatic trader in Google’s DSP meant to raise their bid from $5.00 to $6.25 but accidentally entered $625. No big deal, they caught and fixed the mistake just 5 seconds later.
    In that 5 seconds, the remainder of their entire 200,000 daily impression budget was spent at a CPM of over $200. What should have cost $1,200 ended up costing more than $40k.
    And where did that ad spend go? Straight into Google’s ecosystem.
    Keep in mind Google has Unmatched Control
    Today, Google dominates every part of digital advertising:🔎 Search ads (Google Search)📺 Video ads (YouTube)📈 Ad buying (DV360, the largest DSP)📡 Ad selling (Google AdX, the largest SSP)✅ Verification & measurement (Google Analytics, Campaign Manager 360)🌐 The largest browser in Chrome in the US and the largest Global smart phone operating system in Android.
    They own the buyer, the seller, the marketplace, devices, browser, and the referee, a level of control no other company has ever had.
    Will Google’s Dominance End? We’ll get an answer this summer…
    In the meantime, what can you do about it?
    PSA for the day: if you’re a marketer in any DSP, don’t think twice about using AI Bidding, if you can get screwed 9 years ago in 5 seconds, just imagine what the technology is today across any entity representing publishers. Pick a DSP that only represents the buyer, and let them work for you.🤔
    Anyone else have a similar experience?
    #GoogleAds #AdTech #ProgrammaticAdvertising #Antitrust
    6 min
  • #24 | Lift Off: The Truth About Ad Impact
    The Illusion of High Lift: Why Longer Lookback Windows Matter in Incrementality Testing
    Marketers love seeing high lift in incrementality studies, it feels like proof that campaigns are driving real impact. But if you’re only looking at short lookback windows (1-7 days), you might be chasing last-touch credit and overestimating effectiveness.
    The Problem: Short Lookback Windows Inflate Lift
    When we measure lift between 1-7 days, results look strong because ads do work, they drive awareness and may accelerate conversions. But here’s the catch:
    ✅ Many buyers were already on the path to purchase.✅ Your ad may have just nudged them to buy sooner.✅ If you extend the lookback window, the control group converts too.
    Cutting off analysis too early makes it seem like your ad drove the entire conversion, when in reality, it just influenced timing.
    Real-Life Example: Travel & CTV
    A major travel brand ran a CTV and audio, with display retargeting campaign to drive bookings. Their 1-7 day lift study showed a big conversion increase. But at 90 days, they saw:
    •People go on summer vacations, the control group converted too, just at a later time.•Many “incremental” conversions in the short window were people who would have booked anyway.•The true lift, those booking only because of the ad, was much lower than first thought.
    This insight changed how they measured success, shifting to 60-90 day lookback windows to separate actual lift from simple purchase acceleration.
    Why Longer Lookback Windows Give a More Honest View
    This is critical for CTV, audio, and upper-to-mid funnel media, where the goal is awareness and long-term consideration, not instant conversion.
    In high-ticket industries like travel, auto, and luxury retail, brands sometimes use 60, 90, or even 180-day windows for true impact measurement. If you only measure at 1-7 days, you’re likely capturing last-touch credit rather than real incremental lift.
    What This Means for Your Marketing Strategy
    🔹 Short lookback = overestimated lift. Your ad worked, but not as much as you think.🔹 Longer lookback = a more accurate read. It shows how many would have converted anyway.🔹 Test across different time horizons. Compare short and long-term effects.
    🚀 Key Takeaway: Marketing isn’t just about driving faster conversions, it’s about driving incremental ones. Ensure your lookback windows are long enough to separate the two.
    #MarketingMeasurement #Incrementality #Attribution #MediaStrategy
    6 min
  • #23 | Relative vs Incremental Lift | Pizza Talk
    Relative vs. Incremental Lift: How you can Leverage Both
    Let’s say you own a pizza shop, and you’re running an ad campaign to get more people ordering your delicious pies. You set up a test:
    •Exposed Group: They see your ad and start craving pizza.•Control Group: No ad seen
    At the end of the campaign, let’s say your control group only sold 10 pizzas and the exposed group sold 25, for a total of 35 pizzas sold.
    🍕 What can Relative vs. Incremental Lift tell you?
    ➡ Incremental Lift answers: “How much of my total sales actually came from the ad?”
    📊 (Exposed Sales - Control Sales) / (Exposed Sales + Control Sales)
    •This number is always between 0 and 1. In this case 15 incremental pizzas because of your ad out of 35 total pizzas sold, which means 43% of your total sales were actually influenced by your ad.
    ➡ Relative Lift answers: “How much more did the exposed group buy compared to the control?”
    📊 (Exposed Sales - Control Sales) / Control Sales
    •This number can be over 100%, in this case your control group only sold 10 pizzas and the exposed group sold 25, your relative lift is 150%!
    🎯 The Purpose of Both Metrics
    🔍 Use Relative Lift to optimize performance between tactics:✅ Which audience, targeting, geo, or creative messaging works better?✅ Which channels are more effective within my media mix?✅ How do I allocate spend efficiently?
    📢 Use Incremental Lift to measure true impact:✅ Is this campaign actually driving new demand?✅ What % of sales came because of my marketing efforts?✅ Would these sales still happen if I didn’t run this ad?
    📌 Why It MattersIn the very first iteration of measurement many years ago within Viant’s Advanced Reporting, we focused solely on Relative Lift. We learned that if you focus only on relative lift, you might be optimizing for efficiency, without knowing total impact. We then added incremental lift, which if you focus only on incremental lift, you might miss ways to fine-tune performance. We found our customers appreciate both in the same UI. Your DSP should have both as well.
    📢 Bottom LineGreat marketing isn’t just about improving results, it’s about ensuring those results actually matter.
    #MarketingAnalytics #Incrementality #LiftAnalysis #DigitalMarketing #AdTech #GrowthMarketing #Advertising #MarketingStrategy #Lift #AdTechAIFiles
    5 min
  • #21 | SHOUT OUT: Amanda
    In honor of International Women's Day, I wanted to recognize 2 women that have had a profound impact on my life. One of them personal, another professional.
    My mom, a lifelong nurse, specializing in end-of-life care (hospice) and mid-to-late-stage cancer care, always taught me that the most important thing anyone can do in their life is to CARE, whether for a person or a cause. With 8 billion people on this planet and countless causes to #care about, we each have a limited capacity to give our energy. While I’m not always great at expressing my gratitude, one of the things I care most about is the people I surround myself with, both personally and professionally.
    One practice we’ve implemented in our team meetings is a “shout-out” session. We all may think positive things about our colleagues, but we don’t always communicate them. And I’ve never met someone who doesn’t appreciate a genuine compliment. A simple gesture of appreciation can go a long way, and this few minutes is the best part of any of my meetings all month long.
    As a small gesture of my appreciation, I want to take this a step further and publicly recognize a colleague I deeply respect & and has had a profound impact on my life. While I’d love to see more of this across our industry, I understand that publicly vouching for someone comes with the risk of them getting “poached”. So it hear it goes…
    I have had the honor and privilege to work alongside Amanda Bookout (not on LinkedIn ATM, otherwise would have linked) for the past 15 years. Amanda is Viant’s Director of Business Intelligence. There is simply no one better in our industry that understands both the data complexities that AdTech brings and the business application of that data to bring it to life for our internal stakeholders and clients. She handles the most challenging of issues with calmness and intellect through experience. Her positive, can do attitude is infectious and I’m thankful to have been able to witness it as long as I have. Amanda‘s ability to find, hire, onboard, and develop talent is nothing sort of amazing.
    For those who have ever worked with Amanda, I encourage you to show your appreciation and continue this shout-out, whether it’s through internal Slack, an email, or a comment below. I’ll compile all the comments and include them in an Amanda dedicated episode of the AdTech AI Files Podcast to round out this shout out. Let’s celebrate the people who make a difference.
    Throw the image in a PPT and introduce it to your next team meeting and hashtag#measure what happens!hashtag#ShoutOut hashtag#Recognition hashtag#Teamwork hashtag#NoAI hashtag#IWD25
    6 min
  • #20 | MMP: Use it or Lose it
    Use it or Lose it: a Mobile Measurement Partner ofcourse
    Mobile in-app environments convert 3x more often than mobile web. Think about how often you buy Starbucks coffee through the app vs. opening Safari or Chrome to place an order. For most of us, it’s a night-and-day difference, in-app is faster, more convenient, and personalized. In cases like this, app conversion rates aren’t just higher, they’re the dominant way customers engage and buy.
    Yet, there are no cookies or traditional pixels in an in-app environment, meaning the standard ways most marketers measure performance don’t include mobile in-app conversions. This leaves a huge gap in understanding your true media impact.
    The in-app environment is a different world, frictionless UX, higher engagement, and better retention. To truly measure the impact of your media spend, you need a Mobile Measurement Partner (MMP) integrated into your DSP.
    The good news? The best DSPs can do two key things:1️⃣ Pre-built integrations with your MMP – With simple provisioning, you can securely share data with your DSP for retargeting and measurement.2️⃣ Off-the-shelf incremental lift measurement – The best DSPs work seamlessly with MMP data to provide clear insights into the true impact of your campaigns.
    At the end of the day, it’s simple: use your mobile data or lose viability. If your measurement strategy ignores in-app conversions, you’re making decisions with incomplete data—and in a world where mobile dominates, that’s a risk no marketer can afford
    #MobileMarketing #MMP #DigitalAdvertising #AppEngagement #MarketingROI #AdTech
    If looking for an MMP, check out: AppsFlyer Branch Kochava Singular Adjust
    5 min
  • #19 | Privacy vs. Blocking: The Real Story Behind Consumer Choice
    Privacy vs. Blocking: The Real Story Behind Consumer Choice
    There’s a big difference between consumers opting out of tracking (which would be every consumers right) and Big Tech outright blocking measurement for their own gain. Yet, the lines keep getting blurred.
    Take Apple’s Private Relay, a feature that allows users to effectively be routed through a VPN while browsing in Safari, making you invisible across the web. It sounds like a major privacy shift, but in reality? After nearly 4 years of Apple pushing it, we see only ~16% of Apple related bid requests routed through Private Relay, meaning most users aren't opting-in. Perhaps due to widespread complaints of Private Relay interfering with the user's experience. I personally experienced an issue with getting onto Netflix while testing Private Relay, as Netflix couldn't verify my location properly.
    Now, compare that to a marketer simply adding a pixel on their own website to share their first-party data with their preferred measurement or DSP providers. Apple’s broader privacy policy automatically block nearly 2/3rds of pixel-based measurement hits (and upwards of 85% for some marketers), regardless of whether the consumer has explicitly opted out. That’s not consumer choice, that’s a platform-driven decision that limits visibility into what’s actually driving business outcomes for all marketers alike. Benefiting Big Tech in the name of “privacy”, forcing marketers to integrate and further leverage Big Tech's products.
    So what can you do about it?✅ Leverage 1st-party data: Build direct customer relationships and own your data.✅ Adopt clean rooms: Securely match and analyze data without relying on big tech.✅ Use server-side pixels: Bypass browser-based restrictions to improve measurement accuracy.
    The future of measurement isn’t about waiting for Big Tech’s next move, it’s about taking control of your own data strategy.
    #Marketing #AdTech #Privacy #FirstPartyData #PrivateRelay #Measurement #AdTechAIFiles
    7 min
  • #18 | What the Freq? Linear vs. Connected TV
    Linear TV vs. Connected TV: The Frequency Problem Marketers Can’t Ignore
    After running Linear TV vs. Digital measurement studies for the past decade, one pattern remains consistent:
    Dolllar for dollar, more times than not, Linear TV campaigns have 3x higher frequency than CTV. And here’s the issue, without frequency capping, that’s a lot of wasted impressions.
    Why This Happens:
    ✅ Linear can’t control frequency. If someone watches the same channel daily, they’ll likely see the same ad over and over due to the upfront nature of Linear TV.✅ CTV is transacted programmatically, and therefore lets you set caps, allowing the marketer to control how often someone sees your ad✅ Heavy TV viewers skew the data. A portion of the audience gets bombarded, while others barely see the ad at all.
    What This Means for Marketers:
    🔹 Wasted Spend: Those extra impressions aren’t incremental reach, they’re just oversaturation for the same people.🔹 Diminishing Returns: Seeing an ad 5+ times in a week doesn’t necessarily make someone more likely to convert.🔹 Smarter Budgeting: CTV gives brands more control, meaning less waste and more efficient reach.
    How to Fix It?
    📌 Use digital measurement to quantify true reach & frequency.📌 Balance TV & digital to optimize reach without overexposure.📌 Consider CTV as a bridge, more control than linear, but the big-screen impact remains.
    Linear TV still has its place, but without frequency controls, a huge chunk of ad dollars get wasted on the same eyeballs. Smart marketers are shifting spend to platforms where they can actually control the user experience. 🚀
    #AdTech #MarketingMeasurement #TVAdvertising #DigitalMarketing #CTV #Programmatic #MediaBuying #GrowthMarketing #Advertising
    7 min
  • #17 | Death of the 3rd Party Pixel
    The Death of Third-Party Pixels: Why First-Party Data is the Future
    If you launched your own website tomorrow, you’d own everything: your customer relationships, purchase history, and insights. That’s first-party data, it belongs to you, and no one can take it away.
    Your same business relies on pixels to understand performance, retarget your customers that have added your product to cart but yet to purchase, to optimize and measuring success. You rely on your partners, whether measurement providers or DSPs to help in this endeavor, which makes these Pixels “Third-Party Pixels”. Easy enough…
    Only 1 problem, let’s take a look at who owns the browsers and operating systems that enables your business to pass your data to your partners:
    📌 Google + Apple control 97% of the mobile operating system market (Android & IOS)📌 Google Chrome + Apple Safari account for over 85% of browser usage.
    And when it doesn’t benefit BIG tech, they squash tracking in the naming of privacy (hello Private Relay and Cookie deletion), all of which is increasingly making third-party trackers obsolete.
    What happens when your measurement strategy relies on pixels that no longer work?🚫 You lose visibility.🚫 You increasingly make decisions on a small part of the pie.🚫 Marketers without a first-party measurement strategy fall behind.
    What can you do? The best marketers don’t rent data, they own it. They need to take matters into their own hands in a way that big tech can’t touch.
    Winning in a cookieless world means:
    ✅ Implementing server-side tracking & clean rooms.✅ Building privacy-compliant, deterministic attribution models.
    The brands that control their data will control their future.
    AdTech AI Files latest Pod drop discussing this very topic. On Spotify now: https://lnkd.in/gwTMafsm
    #FirstPartyData #MarketingMeasurement #AdTech #PrivacyFirst #CookielessFuture #DigitalAdvertising #DataOwnership #MarketingStrategy
    5 min
  • #16 | Goodbye DCM
    Why Cookie-Based Measurement Sucks in 2025
    Remember when cookies were amazing? Back in the ‘90s and early 2000s, they revolutionized online tracking. But today, they only capture 9% of online conversions. Why? Because cookies simply don’t exist in most of the places where modern digital engagement happens:
    • ​ Connected TV – No cookies in your streaming ads.
    • ​ Most computer browsers – Safari, Firefox, and even Chrome (if users opt out).
    • ​ Mobile in-app – Cookies only work in mobile web (mostly Chrome).
    • ​ Audio ads on smart devices – Your podcast ads aren’t tracking conversions.
    • ​ Digital out-of-home (DOOH) billboards – Good luck using cookies on a Times Square billboard.
    • ​ Consumer opt-outs – Chrome lets users block third-party cookies.

    Here’s a real-world example: You see a Starbucks ad while searching the web on your phone. Later, you open the Starbucks app and place an order. Does a cookie follow you from web to app? Nope.
    This is why cookie-based measurement is great for companies like Google (who control the walled gardens) but terrible for marketers at large. If you actually want to measure conversions, it has to happen at the household and person level—which, by the way, Viant may or may not have patented back in 2012.
    Now, Viant is doubling down on this future, partnering with TransUnion to bring even stronger, cookieless identity resolution to marketers. With first-party data and identity-based measurement, brands can finally connect the dots across CTV, mobile, web, and beyond—no cookies required.
    The digital world has moved beyond cookies. Has your measurement strategy?
    #AdTech #MarketingMeasurement #CookiesAreDead #DigitalAdvertising #MarketingAnalytics #ConnectedTV #PrivacyFirst #FutureOfMarketing #DataDriven #MarketingStrategy #ConversionTracking #AdMeasurement #MarTech #CustomerJourney #OmnichannelMarketing #IdentityResolution #CookielessFuture
    5 min

About AdTech AI Files

From the publisher's feed

Byte Size AdTech Insights for Those of Us Who Live in an AI World For AdTech marketers navigating the intersection of AI, Measurement, and CTV, this podcast delivers byte-sized insights to keep you…