Weekly Briefing — April 14, 2026: Meta Overtakes Google, PMax Transparency, Creative Is the New Targeting

This week: eMarketer projects Meta will overtake Google as the world's largest ad platform. Google opens the PMax black box with a channel performance timeline. Meta's Andromeda engine makes creative the primary targeting mechanism — broad targeting now beats lookalikes by 49%. Plus: the SEO visibility playbook is shifting from pages to proof, internal wins from a West Coast business school and a B2B SaaS client, and your 6-item cheat sheet for client calls.

Full briefing: Read the HTML version

Transcript

SPEAKER_00 0:00

Hey, it's Will Scott. Welcome to the Search Influence weekly briefing for Tuesday, April 14, 2026. Got a lot to cover this week. Big platform shifts, some internal wins worth talking about, and a few things that should change how we're having conversations with clients. Let's get into it. So the headline of the week, and I think this will be the headline of the year, honestly, is that eMarketer is projecting Meta will overtake Google as the world's largest digital ad platform in 2026. $243 billion for Meta versus $239 billion for Google. First time ever. Now look, the methodology is a little funky. eMarketer strips out Google's traffic acquisition costs, so you can debate the exact numbers. But the directional shift is real and it matters. Meta's AI infrastructure investment, the stuff they've been pouring money into for the last two years, is paying off an actual ad performance. And when the money moves, the leverage moves. For us, the practical takeaway is simple. If you're still positioning Meta as a secondary channel in client conversations, like it's a nice to have alongside Google, that positioning is outdated. Budget conversations should start with what's the best split between Google and Meta? Not, should we add Meta? This is especially true for higher ed and healthcare clients where Meta's targeting used to be tricky. The improvements from Andromeda, and I'll get to that in a minute, make it worth retesting. Sticking with Google for a second. They launched a channel performance timeline view inside Performance Max in early April. This is actually a big deal if you run PMAX campaigns. For the first time, you can see exactly how search, YouTube, display, discover, Gmail, and maps each contribute to results over a specific time period. You can filter by asset type, toggle between cost allocation and performance metrics. This is the biggest PMAX transparency update since the product launched. Before this, understanding where your PMAX budget was actually going required downloading reports, cross-referencing spreadsheets, and making educated guesses. Now you can just look. So here's what I want everyone to do this week. Pull up that channel timeline on every active PMAX campaign. Look for channels that are getting disproportionate budget relative to their conversion contribution. If search is driving conversions at 5x rows and display is burning budget at 0.8x, you want to catch that pattern before it eats another month's spend. Use a 30 to 90 day window. Don't overreact to a single week. And remember, YouTube and display often play upper funnel roles that don't show up in direct conversion metrics. Factor in the attribution model before making cuts. Related signal worth mentioning, Google also quietly killed display and video campaign support and performance planner. You can't build new forecasts around those campaign types anymore. This fits a pattern that's been building for years. Open web display impressions made up only 11% of Google's display advertising as of January 2025, down from over 40% in 2019. Google is following its own data. If standalone display is still a meaningful part of any client's media plan, it's time to start the what comes next conversation. DemandGen is the intended successor. Alright, this is probably the most important strategic shift to understand right now. Meta's Andromeda Retrieval Engine, which rolled out globally by October of last year, fundamentally changed how ads get matched to people. It uses deep neural networks, computer vision, and semantic analysis to read the creative itself and decide who to show it to. Your audience settings are now soft suggestions. The ad itself is doing the targeting. The data from Skeldon's analysis really drives this home. Broad targeting now delivers 49% higher rows compared to look-alike targeting. Let that sink in. A single ad set with 25 diverse creatives produced 17% more conversions at 16% lower cost versus the traditional five ad set structure, and creative fatigue has compressed from 6 to 8 weeks down to 2 to 4 weeks. On top of that, Meta's new adaptive ranking model on Instagram, launched Q4 last year, delivered 3% more ad conversions and 5% higher click-through rate. So what does this mean practically? A few things. One, invest in genuine creative diversity, not creative volume. Five truly different visual concepts beat 50 minor tweaks of the same idea. Andromeda treats 10 variations of the same product shot as a single creative entity. 10 ads, one retrieval ticket, 2. Design for multiple placements, 1x1 for feed, 4x5 for mobile, 9x16 for stories and reels. Each format registers as a separate entity in retrieval. 3. Shorten your creative refresh cycle. If you're still running the same ads for 6 plus weeks, you're past the fatigue cliff. And 4. Implement conversions API with proper deduplication, because Andromeda uses conversion signal quality as a retrieval factor. This is probably the biggest shift in paid social strategy since iOS 14.5, and I don't say that lightly. Switching gears to the organic side, the industry voices we track every week, Near Media, Sterling Sky, Whitespark are all pointing in the same direction. The old playbook of generic pages and vague claims is losing ground. Search engines and AI tools are rewarding clearer intent, better business data, and real third-party mentions. Joy Hawkins' latest Sterling Sky post explains how GBP relationships and chain IDs affect whether Google reads a business as a department, a co-located entity, or a standalone location. If you have multi-location clients, this matters. Whitespark's New Guidance says unstructured citations now matter for both local search and AI visibility. For client conversations, I want us leading with three ideas better service pages, plain English, specific to one buyer type, cleaner listing relationships, make sure Google understands the parent, department, and location structure, and stronger third-party mentions, quotes, coverage, citations outside their own website. That's the practical version of visibility in 2026. A couple of things worth celebrating this week. First, a prospect at a major West Coast business school searched a graduate school marketing GEO audit and converted on one of our AI search optimization content pages. Right content, right searcher, right moment. That's product market fit in action. She came in with a budget in hand, four other vendor conversations lined up, and enough technical fluency to ask about FAQ schema, semantic triples, and Reddit. The bar for sounding credible in these conversations has gone way up, and prospects are doing serious pre-work before the first call. The other thing worth noting from that conversation, AI referred traffic converts at a measurably higher rate than organic search, even though there are fewer clicks. Fewer clicks that convert better is exactly the argument for investing in visibility before the click, not just after it. That applies to both organic and paid. Second, on a B2B SaaS client we're working with, title tag fixes on high impression and low-click pages, move the dashboard to about 40% prompt visibility with an average rank of 9.6 across 33 keywords, but the boring fixes still work. When a page gets impressions but not clicks, fix the title first. It's the fastest win and the easiest proof point to show a client. A few more things to be aware of. Tracking is getting simpler on both platforms. Google consolidated enhanced conversions into one toggle. Meta released an official GTM pixel template and changed click-through attribution to count only real link clicks. If a client's ROS looks different this week, check whether it's the metric change and not an actual performance shift. Microsoft Ads has a real deadline. New API features a REST only starting October 1st. Legacy SOAP API fully deprecates January 31st, 2027. If anything in our workflow touches Microsoft Ads, that needs a migration plan this quarter. And the bigger picture. AI is automating ad tech faster than anyone expected. Ad Exchanger reports that agentic creative testing and automated media buying are the most common new product category. Companies are openly saying they can do the same work with a third of the people. That used to be unsayable. It's not anymore. For us, that means our value is in strategy, creative quality, and clean data. The parts AI still can't automate. Alright, your cheat sheet for client calls this week. Six things. One, fix the signal first. If the account doesn't have one clear conversion event, better bids and better creative will both get confused. Clean data before you scale. Spend. Two, creative is targeting now on Meta. Five genuinely different ad concepts beat 50 variations of the same idea. Three, P Max is less opaque. Pull up the channel timeline and show clients where their budget is actually going. Four, stop planning standalone display. Google's own tool doesn't support it anymore. Move upper funnel planning to demand genders. Five, lead with three visibility ideas better service pages, cleaner listing relationships, stronger third party mentions. Six, don't sit on API debt. Microsoft's October deadline is real. Get a migration plan on the calendar. That's it for this week. Go win something.