How to test incrementality with MMM

Last-click attribution is still an option in many platforms, but it rarely provides a true picture of ad effectiveness. While it may be comforting, relying on it is unlikely to drive actual growth. Incrementality tests, in particular AB tests, randomised experiments and geo-holdouts, may not show you large "claimed conversion" numbers. The true incremental uplift is often only between 1-10%. However, this is the truth. Think about this from a sensitivity point of view: If the last-click (or last-touch) attribution numbers were accurate (often suggesting 50-80% sales are driven by one campaign), every company running multiple ad campaigns would need to grow as fast as NVIDIA over the past year. The good news: The same platforms that offer last-click or last-touch attribution also tend to provide various incremental lift tests in many regions. For example: Meta - Conversion Lift and AB Testing https://lnkd.in/gr6nyg7m and https://lnkd.in/gTip8uJ9 TikTok - Split Test https://lnkd.in/gaCiZ-G7 Amazon - AB Test and Brand Lift https://lnkd.in/gF2YBmsN and https://lnkd.in/g5KDzgek Google - Conversion Lift https://lnkd.in/gWCYnwsu Trade Desk - Conversion Lift https://lnkd.in/gWBaaGdS Otherwise, there are also great third-party tools that can be used for incrementality tests. Additional comment 1: The comparison of last-click/ touch attribution versus incrementality tests mainly refers to within-channel/within-platform optimisation. Additional comment 2: Not all split or AB test tools will assign users randomly to different ads or test conditions. Sometimes a targeting algorithm decides on the selection of users too. This means we cannot speak of pure causal incremental effects. Unfortunately, often tutorials and explainers are not clear on the settings and manipulations. That's why it's super important to always check this. One great research study outlining the differences in types of experiments is the following study (link below): Braun, Michael, et al. "Leveraging digital advertising platforms for consumer research." Journal of Consumer Research 51.1 (2024): 119-128.

Do you know the top #DigitalAdvertising trends to watch in 2025? ✅The most important trends in advertising is how AI, social media and consumer behaviour are impacting how advertising monies are spent and how creatives are made. 📌 Let's look at the top digital advertising trends in 2025: ➡️1. AI plays a prominent role in digital advertising. 📌 Artificial intelligence and machine learning are creating a revolution in the advertising space. 📌 Advertisers are using AI tools to improve the way they test and optimise impressions, click through rates, bid levels and targeting. 📌 AI can do more than optimise ad campaigns and can also help create them. ➡️2. Video ads get shorter and shorter. 📌 You can expect the majority of the ad spend to go to short form video ads. Some brands are trending towards ultra short, but engaging ads to beat the skip button on YouTube. 📌A recent study shows shows that 100% of the current video advertisers say they will continue to use video ads in 2025. 📌 More than 95% plan to increase their budget for video ads and 70 percent of marketers who don't use video say they plan to start in 2025. ➡️3. Ads lose the sound. 📌 We're on a webpage when video ad we can't see, starts to play at full volume. It turns out that this type of situation annoys a large number of people. 📌Outstream video is one option that may be a popular alternative in the coming year. These types of videos are integrated into the content the user is reading. 📌 They start playing without volume when the user scrolls over them and pauses once they are out of view. Statistics show that Outstream ads have a higher CTR rate. ➡️4. Advertisers capitalise on the popularity of mobile gaming. 📌 Nearly 70% of gamers say they would give up social media or TV before they would give up mobile games. Users have said they are twice as likely to pay attention to in game display ads than other ads. 📌 Along with typical banner ads, there are interstitial ads and reward ads on the mobile gaming apps. 📌Interstitial ads are put at the stopping points in the game, so they don't feel unnatural to users.They also take up 100% of the users screen. 📌 Reward ads offer users game points, currency or another gaming related benefit for watching the entire ad. ➡️5. Consumers demand authenticity in social media ads. 📌 Studies have shown that more than 75% of consumers say they trust content from people more than content from brands. 📌 Marketing professionals are turning to social media to create a balance between brand content and user generated content that conveys the human side of the brand. 📌 Influencer marketing is one way advertisers are working to humanise its brands. 📌 Some companies have invested in non paid advocacy strategies. They have called on dedicated employees to serve in a pseudo influencer role. ✅If you resonate with my views, please like and comment with your thoughts. 📌 Thank you and wish you a great day ahead!

The Attribution Models: When to Use 3rd Party Tools? As a media buyer for DTC Brands, I've had my fair share of conversations about attribution tools. Recently, One founder asked about when to use third-party attribution tools for tracking data. I responded on how many platforms he was advertising, He said he is mainly running ads on a single platform i.e. Meta. But he wanted to use the tools like Triple Whale or Northbeam. It got me thinking - Does he have clarity on when to use these tools? Well, I guided him, the third-party attribution tools are fantastic, but you need to know exactly when to use them. If you're advertising on multiple channels like Meta, TikTok, and Google etc it's worth investing in these tools. But if you're only advertising on one platform, you might not need them. Here's the thing: if you rely solely on in-platform conversion reporting, you'll see a lot of double or triple reporting. Each ad channel will claim the same conversion, making it difficult to understand the true impact of your ads. 𝘓𝘦𝘵 𝘮𝘦 𝘤𝘭𝘢𝘳𝘪𝘧𝘺 𝘪𝘵 𝘸𝘪𝘵𝘩 𝘢𝘯 𝘦𝘹𝘢𝘮𝘱𝘭𝘦. 𝘐𝘮𝘢𝘨𝘪𝘯𝘦 𝘢 𝘤𝘶𝘴𝘵𝘰𝘮𝘦𝘳 𝘸𝘩𝘰 𝘤𝘭𝘪𝘤𝘬𝘴 𝘰𝘯 𝘢 𝘍𝘢𝘤𝘦𝘣𝘰𝘰𝘬 𝘢𝘥, 𝘵𝘩𝘦𝘯 𝘰𝘯 𝘢 𝘛𝘪𝘬𝘛𝘰𝘬 𝘢𝘥 𝘰𝘧 𝘵𝘩𝘦 𝘴𝘢𝘮𝘦 𝘣𝘳𝘢𝘯𝘥, 𝘢𝘯𝘥 𝘧𝘪𝘯𝘢𝘭𝘭𝘺 𝘮𝘢𝘬𝘦𝘴 𝘢 𝘱𝘶𝘳𝘤𝘩𝘢𝘴𝘦. 𝘐𝘯 𝘢 3𝘳𝘥 𝘱𝘢𝘳𝘵𝘺 𝘢𝘵𝘵𝘳𝘪𝘣𝘶𝘵𝘪𝘰𝘯 𝘵𝘰𝘰𝘭, 𝘵𝘩𝘦𝘴𝘦 𝘢𝘥𝘴 𝘸𝘰𝘶𝘭𝘥 𝘳𝘦𝘤𝘦𝘪𝘷𝘦 𝘱𝘢𝘳𝘵𝘪𝘢𝘭 𝘤𝘳𝘦𝘥𝘪𝘵, 𝘮𝘦𝘢𝘯𝘪𝘯𝘨 𝘵𝘩𝘦 𝘤𝘳𝘦𝘥𝘪𝘵 𝘪𝘴 𝘥𝘪𝘴𝘵𝘳𝘪𝘣𝘶𝘵𝘦𝘥 50% 𝘵𝘰 𝘣𝘰𝘵𝘩 𝘢𝘥𝘴. 𝘞𝘩𝘦𝘳𝘦𝘢𝘴 𝘪𝘧 𝘺𝘰𝘶 𝘶𝘴𝘦 𝘪𝘯-𝘱𝘭𝘢𝘵𝘧𝘰𝘳𝘮 𝘳𝘦𝘱𝘰𝘳𝘵𝘪𝘯𝘨 𝘣𝘰𝘵𝘩 𝘱𝘭𝘢𝘵𝘧𝘰𝘳𝘮𝘴 𝘸𝘪𝘭𝘭 𝘵𝘢𝘬𝘦 𝘵𝘩𝘦 𝘤𝘳𝘦𝘥𝘪𝘵 𝘰𝘧 𝘵𝘩𝘦 𝘴𝘢𝘮𝘦 𝘤𝘰𝘯𝘷𝘦𝘳𝘴𝘪𝘰𝘯 𝘴𝘰 𝘪𝘵 𝘸𝘪𝘭𝘭 𝘣𝘦 2 𝘤𝘰𝘯𝘷𝘦𝘳𝘴𝘪𝘰𝘯𝘴 𝘣𝘶𝘵 𝘪𝘯 𝘢𝘤𝘵𝘶𝘢𝘭𝘪𝘵𝘺, 𝘵𝘩𝘦𝘳𝘦 𝘪𝘴 𝘰𝘯𝘭𝘺 𝘢 𝘴𝘪𝘯𝘨𝘭𝘦 𝘤𝘰𝘯𝘷𝘦𝘳𝘴𝘪𝘰𝘯. So, when do you use 3rd party attribution tools like Northbeam & TrippleWhale? In my POV -> when you're advertising on multiple channels. These tools help you understand the true impact of your ads and allocate your budget more effectively. Are you using 3rd party attribution tools? would like to know your thoughts #adtracking #attributiontool

Embracing the digital age, I often find myself thrilled by the potentials unlocked by Online Advertising. 🌐💼 • It's an incredible era where our messages can cross borders and time-zones, reaching individuals who were impossible to connect with before. And it's not just about the reach, it's about the refined, intricately detailed targeting like never before. • Think geo-targeting, retargeting, behavior targeting... The list goes on. These tools empower us to send precise, personalized messages that drive grand results! Through online advertising, power shifts to the consumer. They choose when to engage, and how. ✌️ • This shift triggers creativity and innovation for us marketers – we need to create content that attracts, engages, and resonates with our audience, shifting from invasive push tactics to more persuasive attract methods. Yet, we must never forget: data privacy is critical. It's a fine balance. 🎯 • A debate that often comes up is the trade-off between personalization and privacy. How much is too much? That's a question we must continuously ask ourselves as avid online advertisers. Online advertising is an art. An art where the canvas is your digital screen and the brushes are your marketing tools! 🎨 • And each campaign is a new masterpiece. Requires the precision of a surgeon, the imagination of a painter, and the pragmatism of an engineer. It's a fascinating blend of creativity and analytics. But, like all art, it is not immune to criticism. • Yes, online advertising is under scrutiny – ad blockers, fake news, and privacy intrusions are real challenges we face. How we adapt, address these issues, and evolve is what will set legendary brands apart. Online advertising, fun as it may be, holds a steep learning curve.🎢 • The landscape is ever-changing, emerging technologies such as AI, big data, VR, AR, blockchain are revolutionizing online advertising. Staying atop these trends is no walk in the park. As we sail this digital sea, let's embrace the change, adapt quickly, but never compromise on our brand value and audience trust.🚀 • Remember, at the heart of every successful Online Advertising campaign is a customer who feels valued and understood. Let's continue the conversation. What are your thoughts on the evolution and challenges of Online Advertising? Drop your thoughts in the comments. Let's learn together.🙌 #OnlineAdvertising #DigitalMarketing #ChangingLandscapes

Your Creative Testing is Wasteful! The idea that you can throw as many creatives as you can churn out into one ASC campaign and expect success is a myth. Even if you’ve seen some improvement, remember that better metrics are relative. If you started with a poor account structure/optimization, any improvement doesn’t necessarily mean your approach is optimal. At Social Bulldog, we have a budget strategy that ensures sustainable growth while allowing room for testing. Here’s a guide we tweak based on each client's needs: -70% of your budget to acquisition: Excluding lifetime customers and all website visitors, ensuring that most of your budget reaches new users and grows the business. -10-15% for retargeting: This depends on business style and channel mix. I prefer it close to 10% or even lower, letting other channels handle retargeting. -10-15% of the budget for creative testing: The number of creatives we can test depends on your budget and CPA. We only test creatives if we’re confident they’ll generate enough conversions to provide meaningful data. For example: With a $400,000 monthly budget and an average CPA of $100, allocating $50,000 for testing could drive around 500 conversions. This allows us to test 10-20 new creatives monthly, each achieving 25-50 conversions. For a $100,000 monthly budget with the same CPA, a $15,000 test budget would drive about 150 conversions. Here, we’d test 3-6 new creatives, focusing on BIG ideas rather than smaller iterations due to limited testing opportunities. Conclusion: Plan your budgets based on your business KPIs and larger goals. This dictates how much you can test. Stop throwing everything at the wall to see what sticks—hope is not a strategy. #FacebookAds #MarketingStrategy #DigitalMarketing #AdOptimization #CreativeTesting #SocialMediaMarketing #socialbulldog #digitalmarketing

Meta Advertisers: If your conversion tracking & ad structure are wrecked, creative can’t save you.   Just talked to a $25M brand running a strong creative motion.   But it looked like a 5th grader set up conversion tracking and ad accounts.   They’re putting out great statics & videos every week.   Starting with great creative frameworks & high-converting ad templates.   Even running a proper whitelisting play.   But it wasn’t enough to scale ads aggressively. B/c they had:   😡 Limited server-side tracking   😡 Only CAPI v. better options - Blotout, Elevar   😡 No micro-conversions - only orders   😡 Broken attribution - inaccurate GA4, no Northbeam   😡 Too many campaigns - crushing Meta’s ML’s potential   😡 Bad budget allocation - to Tests vs. Winners   😡 Too much ABO, too little CBO   😡 Too much retargeting - faking growth w/o incremental profit   --   Bums me out when good brands get in their own way (often bc of bad agencies). Yes, creative is a massive lever. But you’ll never be A+ if sleeping on the above.

Meta’s ad attribution update gives clearer view of conversions Meta has introduced new attribution settings in its Ads platform, allowing advertisers to distinguish between all conversions and first-time conversions. How it works. Advertisers can now choose between two attribution options: All Conversions: Shows every conversion after an ad view or click First Conversion: Displays only the first conversion after an ad view or click Why we care. Advertisers have long struggled with discrepancies between Meta’s ad data and their own backend numbers. This update aims to bridge that gap, providing more accurate insights into ad performance and customer acquisition costs. By the numbers. One advertiser found that Meta was counting about 25% of non-first purchases towards new customers, significantly inflating reported performance metrics. Between the lines. Previously, Meta counted all conversions a user made after interacting with an ad, potentially including multiple purchases by the same customer. This led to inflated performance metrics and skewed customer acquisition costs. How to access. Advertisers can find the new feature in Ads Manager under Attribution settings, where they can select their preferred attribution window and conversion view. What’s next. As advertisers begin to use these new attribution settings, we may see a shift in how businesses evaluate the effectiveness of their Meta ad campaigns and allocate their marketing budgets. First seen. We discovered this update via Bram Van der Hallen on LinkedIn: July 04, 2024 at 01:29AM

Day 14 Still on Performance Ads, I learnt about Performance Ads type/objective on X and Ad format. Performance Ads is a type of ads that seeks users to complete an action. In this type of ads, advertisers will only pay when a user complete a specific predefined action. Performance Ads is a conversion ads which is at the bottom of the funnel. The aim of advertisers is to get large audience to perform a specific task. The major KPI for performance Ads is Return on Ads Spend (ROAS) and in this case, the ads budget is an investment. The ROI of the ads is tied to the number of people that performed the action. Other Performance metrics used for Performance Ads as it pertains to X ads include; - Click-through rate (CTR) - Cost per clicks (CPC) - Conversion rate (CVR) There are 3 Performance Ads type and 3 ads Format you launch on X. The Ads type are; - App download - App user re-engagement or retargeting and - Website visit. The Ads formats are; - Carousel - Still image - Video Performance ads on X are also placed on X Audience Platform (XAP) although X advices that only App download ads should be place on there Audience Network.

Every brand who buys ads on Meta, Google, or TikTok is looking for the same thing: a savior product that restores their ad performance to "the good old days". They see the promise of Popsixle or any other data connector as the potential silver bullet that gets them back in control and back to growing. Here's the sad truth: a quick fix that magically turns around your business simply is not possible. Anyone who tells you there's an instant fix to restoring growth is either lying or doesn't know what they're talking about. "Restore performance to pre-IOS14" isn't possible. "Protect yourself from the cookie-less future" isn't possible. "Double your ROAS overnight" isn't possible. Here's what is possible, and what we've seen over and over again with all of our customers. Great ad performance is built on a foundation of four elements: ▶ Great data ▶ Great creative ▶ Great media buying ▶ Great landing page/site/offer These set a positive feedback loop in motion that unlocks the super-powers of Meta/Google/TikTok machine learning, as follows: ↳ high click-through ads ↳ high conversion-rate websites ↳ high match-rate conversion data ↳ better ad quality scores ↳ better auction performance ↳ better ad efficiency, scale, and stability While there's no silver bullet that instantly transforms your business, fixing your data can help 10-20%. Setting up optimal ad campaign structure can help another 10-20%. Finding optimal creative with message-market fit can help another 10-20%. Optimizing your landing page and site speed can help another 10-20%. All of these incremental gains close the gap from a pretty good business to a well oiled conversion machine that is scalable, predictable, and stable. We're proud to offer a data service that is a part of this success equation. But there are no magic wand solutions. If you know someone who needs to hear this and wants to rebuild their business piece by piece, share this post and tell them about Popsixle.

7 things that destroy a Meta ad account: 1). Not understanding your customer → If you can’t think like your customer, your ads won’t work. Spend time researching, reading reviews, and talking to them. 2). Being too conservative with creative → Don’t be afraid to try different creative styles. Your ads should stand out, not fit in. 3). Spending too much on retargeting → Sure, you can spend 80% of your budget on retargeting and get a low CPA, but you’re not filling the funnel with new prospects. 4). Not testing consistently → Make sure you’re regularly testing, learning, and refining your creative. 5). Ignoring your landing page → If your landing page doesn’t convert, even the best ads won’t save you. 6). Trusting your taste instead of your data → Don’t fall in love with your ideas. If the data says it doesn’t work, kill it. 7). Sweating the small stuff → Take big swings. Try very different ad creatives. Try different landing page layouts. Worry less about micro-optimizations.