Personalized Ad Experiences

Uber is a master in ad personalization at scale. The behind-the-scenes of how we pulled it off when I led growth efforts at Uber: As a global company, Uber’s complexity is no joke. We’re talking: • Global markets • Multiple products • Diverse languages Without a streamlined system, personalizing ads would be chaotic. That’s why we built a well-oiled machine for ad creative and messaging across regions. Which resulted in ads so personalized... They felt tailor-made for each region. And believe it or not, our backbone was... Google sheet. But this sheet didn’t just house creative assets and messaging. It was our command center. We hooked our sheet into what’s called an FMP (Facebook Marketing Partner.) We could update and push hundreds of campaigns across markets without opening Ads Manager. Need to tweak copy in a certain region? Adjust in the sheet, hit refresh, and it's live. That's how we personalize our ads at scale. If you don’t have the budget for an FMP, you can still achieve personalization at scale. Instead of overwhelming yourself by testing in 20 markets at once. The key is to start small: ↳ Test messaging within a single category or region ↳ Identify winning variations ↳ Scale them across other markets A concrete example: Uber’s rider campaigns weren’t just about translating content. But also about cultural relevance and buying power. An ad in Mexico City is entirely different from one in Los Angeles: • Language: English vs. Spanish • Offers: Discounts adjusted for buying power • Location: Specific callouts like "Hey Los Angeles" vs. "Hey Mexico City" We wouldn’t give a $10 discount in Mexico as $10 is worth way more there. Beyond ad creative, we ran advanced growth tests. For instance, we measured user propensity: The likelihood of someone re-engaging if they saw an ad and received an email versus just seeing an ad. We dug into timing too - Were 30-day inactive users more likely to convert than those inactive for 90 days? These granular insights fine-tuned every touchpoint of our campaigns. For less mature growth teams: The goal is to optimize the first 95% of performance with best practices. But for advanced teams like Uber’s... The challenge is squeezing out that last 5%. And that requires hyper-specific: ↳ Testing  ↳ Segmentation  ↳ Understanding how every variable affects your outcome That’s why we ran advanced growth tests – to nail the final 5%. While Uber was already a highly mature company when I joined... This level of personalization applies whether you’re an enterprise or startup. It’s a scalable approach regardless of company size. The best part? Personalized messaging yields incredible results. As competition rises, this precision is what separates you from the rest.

When creating tailored marketing campaigns, handling data privacy and personalization can be tricky. Why? Because we have to respect people's data. It's easy to think, "The more data, the better!" But that's dangerous because it can lead to personalization, where people feel they're being watched. Think about those ads that follow you around the internet after you've browsed a product. Annoying, right? Well, that's what happens when personalization goes wrong. We need to shift our thinking. It's not just about what we can do with data but what we should do. It's about building trust, not destroying it. In my experience, ethical personalization starts with transparency. 👉🏾Be clear about what data you're collecting and why. 👉🏾Give people control over their information. 👉🏾Make it easy for them to opt out. It's also about providing real value. Do not try to trick people into buying something. It should enhance their experience and make their lives easier. For example, think about a streaming service that suggests movies you might like based on your viewing history. That's helpful. That's valuable. Or consider an e-commerce site that remembers your shipping address and payment information. That saves you time. That's a good use of personalization. At the end of the day, we need to create personalized experiences that are both effective and respectful. #marketingstrategy #b2bmarketing #demandgeneration

For years, I thought personalization was this: - Using a landing page template - 98% of the content is the same - Swap a company logo and name  - Maybe if we're getting crazy, add a title Wow, so ABM of me.  Not to mention the risk I was taking on: What if I used the wrong logo?  Am I even allowed to use the logo?  What if I misspell the company name?  What if I show the page to the wrong company? Now, add the volume I had to manage. Having approx. 0 time for analysis and testing. Then it hit me — Personalization ≠ being *personal*. 70% of users want custom experiences. But that doesn't mean having their PII and pictures of grandma splattered across pages. I simply need to meet them where they are. Key = relevant experiences Amplitude released a report on personalization at scale and had cool examples that were neither too generic nor too personal. Like Goldilocks' porridge. 1) Segment users by behavior, not PII  For example, to move users from free -> paid. - Look at who upgraded in the last 60 days.  - Create a cohort who show similar behaviors. - Create landing pages relevant to that cohort. 2) Message timing - Pop-ups directing users to a product demo  - Use on BOFU high-intent pages, not TOFU - Look at intent to learn vs purchase 3) Recency effect How often have you gotten an ad for a product you purchased? Or a reminder for a bill you already paid? WORST. Exclusions are as important as inclusions. Personalization ultimately comes down to good, targeted, and relevant marketing. (Link to report in comments)