Attribution has become the most expensive fiction in the building.
I see teams tighten knobs because it feels like control, then act surprised when the story changes overnight. Shorten the attribution window and brand looks useless. Expand it and suddenly everything “works.”
In paid social, we’re hearing the same song remixed: less obsession with micro-precision, more acceptance that the cleanest numbers are often the least true.
I get why this happens. Leadership wants certainty. Teams want defensible decisions. Platforms happily offer both, as long as you accept their definitions of reality. The problem is that buyers stopped cooperating with those definitions a long time ago. People triangulate now.
They bounce from search to feeds to Reddit threads to AI summaries, then back to a friend’s screenshot. The journey isn’t linear, and pretending it is creates bad incentives. You end up optimizing for whatever gets credit instead of whatever builds conviction.
So how do we fix it? Choose your rules on purpose. Pick attribution windows that match the buying cycle, and decide what counts as proof before you look at the dashboard. Build timing and creative systems that can hold up when measurement gets noisy.
If you don’t define your measurement story, the platform will. And it will always pick the version that makes spending feel inevitable.
Patrick Patterson
CEO at Level Agency
Podcast spotlight
How to Scale AI with Empathy and Rebuild Trust in a Digital World
I sat down with Myles Biggs and our new SVP of AI, Krish Sailam, and it was one of those conversations that cuts through the “AI = efficiency” trope and gets into what actually breaks in the real world: trust, hiring, and customer experience.
Krish’s background spans Amazon, Cisco, and BMW, and you can feel the difference in how he thinks. He’s not chasing tools. He’s translating disruption into an operating plan.
We talk about what happens when deepfakes make “proof” slippery, why the resume is basically dead in a market flooded with AI-generated applications, and how leaders can keep teams moving forward without pretending the learning curve isn’t real.
If you’re trying to scale AI inside a business without turning your culture into collateral damage, this one’s worth your time.
Paid social is trading precision for defensibility Some brands are abandoning overly granular metrics because “precise” doesn’t always mean “true.” The shift is toward broader, more reliable gauges that hold up in budget conversations. If your reporting wins dashboards but loses CFO trust, this is your warning. Read more
Attribution windows decide your narrative Attribution windows aren’t neutral settings, they’re a decision about what gets credit. Pick the wrong one and you’ll starve the work that creates demand, then celebrate the work that captures it. This is one of the easiest ways to accidentally train the business on the wrong truth. Read more
Social discovery is rewriting distribution rules Discovery is happening in feeds, comment sections, and creator ecosystems, not just “your content.” This frames distribution as the constraint: how you earn reach, not how often you post. If your social strategy is still built like a publishing calendar, you’re playing the wrong game. Read more
“Gen Z proof” Google Ads are really a trust test 18–24 buyers cross-check fast and punish vague claims. This pushes search teams toward tighter intent mapping, clearer language, and landing pages that validate credibility immediately. If your ads assume a linear funnel, you’ll misread what’s actually driving response. Read more
The Level way: Culture as a client-facing operating model Culture is usually treated like internal fluff. This reframes it as a client-facing operating model: how decisions get made, how quality gets protected, and how consistency shows up in delivery. If you’ve ever felt an agency “drift” over time, this is the antidote. Read more
Deep dive: AI
Why your website needs to be legible to machines
AI systems are increasingly acting as gatekeepers between brands and buyers. Whether it’s ChatGPT summarizing options or Google’s AI Overviews pulling key information, models decide what to surface based on clarity, structure, and perceived authority.
Marketing Dive’s piece on optimizing mobile apps and websites for AI makes a strong case that technical hygiene is now strategic. Clean schema, structured data, fast load times, and clear entity signals are not just SEO best practices. They determine whether AI systems understand your brand well enough to cite it.
This changes the optimization mindset. You’re no longer just writing for users or crawlers. You’re writing for summarizers. If your product pages bury value propositions in vague language, or your site architecture obscures relationships between services, AI tools will struggle to represent you accurately.
The brands that win will treat machine legibility as part of brand strategy.
WordStream’s roundup of 2026 marketing predictions is useful because it moves past shiny-object forecasting and into structural change.
The consistent theme: automation will increase, but differentiation will come from human inputs. Creative quality, offer strength, and first-party data discipline matter more as platforms abstract execution. AI will handle bidding, targeting, and testing at scale. What it cannot generate on its own is positioning clarity or strategic restraint.
Another throughline is measurement realism. As journeys fragment and privacy constraints persist, leaders will need to get comfortable making decisions with directional, not deterministic, data.
The prediction I’m betting on: teams that tighten their systems and simplify their signals will outperform those chasing every new feature release.
Claude Cowork is Anthropic’s workspace-oriented interface designed to function more like a persistent collaborator than a one-off chatbot.
What I like about it is the emphasis on shared context. Instead of starting fresh each session, it allows teams to maintain continuity across projects, documents, and conversations. For growth leaders juggling strategy docs, campaign planning, and internal briefs, that continuity matters.
It’s not magic. You still need clean inputs and clear prompts. But as AI shifts from “assistant” to “teammate,” tools that preserve memory and project-level context will create real leverage.
If you’re experimenting with AI inside your marketing workflow, this is worth a serious look.
Get your AI Visibility Scorecard
AI platforms are shaping how buyers discover and vet brands. But most marketers don’t know if their brand even shows up.
Level’s AI Visibility Scorecard shows how often your brand appears in ChatGPT, Gemini, and Perplexity, how you’re described, and how that compares to your competitors.
You'll get:
A visibility score and share of voice vs top competitors
Sentiment analysis of how your brand is framed in AI answers
A tactical roadmap to improve your citations, content structure, and authority signals
Used by leading brands and fine-tuned across real buyer prompts, this scorecard is your opportunity to drive real visibility in the zero-click search environment.
