There's a pattern I keep seeing. An agency wins an award, puts it in the deck, and starts leading with it in every pitch. The badge becomes the story. Meanwhile, the work that earned the badge stops being the priority.
I get why it happens. Recognition feels like validation, and validation is addictive. But there's a version of this that becomes a trap, where you spend more energy on the signals of good work than on the work itself.
The reason I'm thinking about this is that we had a good week by the external scoreboard. Level was named a 2026 Google Premier Partner, placing us in the top 3% of Google Partners in the country. Patrick Van Gorder, our Chief Partnership Officer, said it better than I would: "Our clients don't hire us for badges — they hire us to grow."
And then Google said something back to us in a different way.
Think with Google featured Level.Signal for how it uses Google Cloud infrastructure and predictive search signals to help brands connect fragmented marketing data into sharper decisions. For one higher education client, that work delivered a 48% increase in applications and a 30% reduction in cost per application.
Think with Google doesn't write about agencies. It writes about ideas worth scaling. That's the distinction that matters.
Neither of these things was built to be recognized. They were built because the alternative, managing campaigns without accountable systems, produces results nobody can defend when leadership asks hard questions.
The recognition is a byproduct. The moment it becomes the goal, you've already started losing the thing that earned it.
Patrick Patterson
CEO at Level Agency
Tomorrow: The Virtual Insider Session with Reddit you should attend
We're hosting a session tomorrow at 1 PM ET with the team at Reddit and Level, and I think it's going to be one of the more useful hours you spend this month.
The topic is Reddit as a serious marketing channel, not as a social media checkbox, but as a place where real buyer conversations happen, AI systems listen, and brand reputations get built or eroded without most marketers ever noticing.
We'll get into the AI visibility angle, how to approach the platform without getting it wrong, and what the data actually shows about how Reddit influences pipeline.
Poetry, private equity, and the marketing principles that outlast everything
With Cody Lee, Principal, Peak Performance Group at Summit Partners
Cody Lee has worked in pharmaceutical advertising, poetry, copywriting, growth marketing, and private equity. That's not a resume that follows a straight line, and that's exactly what makes this episode worth your time.
The mental model he introduces early in the conversation is one of the most useful things I've heard in a while. Separating what is valid from what is relevant. It applies everywhere, but it's especially sharp right now when AI can produce an enormous volume of valid content that has nothing to do with what your specific audience actually needs.
We also get into what it looks like to connect marketing to a financial investment thesis, which is a conversation most marketing leaders avoid having and most PE-backed companies desperately need.
And yes... there's a Fyre Festival reference in defense of the four Ps. It holds up.
P.S. The best episodes come from people who didn't follow the obvious path. If you've got a "good enough isn't" story, broken a rule that needed breaking, or figured something out about AI and growth the hard way, I'd like to hear it.
What's interesting this week
Shopify says purchases are coming inside ChatGPT Agentic storefronts are moving from concept to reality faster than most teams are ready for. OpenAI stepping back from Instant Checkout doesn't slow the trajectory, it just changes the infrastructure. The brands that have clean product data, strong signals, and clear conversion paths will be the ones that show up when agents start making buying decisions on behalf of users. Read more
AI search barely cites syndicated news or press releases This one is worth sitting with. If your content strategy relies heavily on syndication or PR distribution to build authority, AI search is largely ignoring it. What gets cited is original thinking, specific expertise, and content that earns attention on its own terms. The implications for how brands invest in content are significant. Read more
LinkedIn might have two algorithms, and marketers should care SparkToro's breakdown suggests LinkedIn surfaces content differently depending on whether it's being distributed through search or through the feed. That means a post optimized for one may underperform in the other. If LinkedIn is a serious channel for you, understanding which algorithm you're actually trying to reach changes how you approach content architecture. Read more
Google is reimagining Maps with Gemini Navigation is becoming conversational. Google's new immersive Maps experience lets users ask questions, explore areas, and get contextual answers the way they would from a person who knows a city well. For marketers, local presence and structured data matter more as discovery shifts from search bars to natural language. Read more
What UGC actually does for brands User-generated content keeps outperforming brand-made creative, and this piece is a useful reminder of why. It's not about production value or reach. It's about trust transfer. When a real person says something works, it lands differently than when a brand does. The practical question isn't whether to use UGC, it's how to build systems that generate it consistently without manufacturing it into something that feels fake. Read more
Claude now builds interactive visuals inline Anthropic rolled out the ability for Claude to create charts, diagrams, and interactive visualizations directly in conversation. For growth teams using AI in reporting, analysis, or client communication, this closes a meaningful gap between generating an insight and being able to show it clearly. Read more
Deep dive: AI
Automation layering and the new job of the PPC manager
Paid media automation has been reshaping campaign management for years, but the conversation has shifted. It's no longer about whether to use Smart Bidding or automated assets. It's about how to layer multiple automation systems together so they produce better outputs than any single tool would on its own.
The concept this piece introduces is useful: automation layering, combining Smart Bidding, scripts, rules, third-party tools, and LLM-assisted analysis into a system where each layer adds signal or guardrail to the one beneath it. The platforms handle mechanics. The marketer's job is increasingly about the quality of inputs going in and the quality of interpretation coming out.
The piece is practical rather than theoretical, with specific use cases for Smart Bidding safeguards, how to actually use the Google Ads Recommendations tab rather than ignoring it, and where LLMs fit into analysis workflows without replacing human judgment.
The underlying argument is one worth sitting with. Automation doesn't replace the need for strategic thinking in paid media. It raises the floor on execution while making the ceiling entirely dependent on how well the human operating the system understands what good looks like.
Most organizations treating connected customer experience as a technology problem are solving the wrong thing. The MarTech panel at the center of this piece makes a point that's uncomfortable but hard to argue with: the data exists, the tools work, and customers are still repeating themselves every time they reach a new department.
The culprit isn't the stack. It's the operating model. Marketing is measured on lead volume, sales on revenue, support on resolution time. None of those metrics require teams to work together, and no amount of CDP investment fixes that when the incentives point in opposite directions.
The practical framing here is worth keeping: you don't need a 360-degree view of every customer. You need 80% of the right data, clean enough to act on, in a single authoritative system. The brands that get this right aren't the ones with the most sophisticated infrastructure. They're the ones that decided who owns the customer and built compensation around outcomes that require actual coordination.
If your teams can't answer who owns data quality, that's where to start. Not with the next platform purchase.
Most AI image tools get you 80% of the way there and leave you with something that looks like AI made it. Krea is useful for closing that gap.
Where it's earned a place in our workflow is refining AI-generated imagery into something that actually looks real enough to run in paid placements.
The practical value is in the iteration speed: you can take a generated asset that's directionally right and push it toward something that actually holds up in a paid placement without handing it off to a full production workflow.
What does ChatGPT actually say about you?
When a potential buyer asks an AI agent for a Top 5 list in your category, do you make the cut? More importantly: Why or why not?
Don't leave your brand’s reputation to an algorithm you haven't audited. Our scorecard pulls back the curtain on how AI platforms frame your brand sentiment and which sources they use to vet your authority.
Inside your custom report:
Sentiment analysis: Is the AI's summary of your brand accurate or outdated?
Citation audit: Discover which third-party sites are feeding the AI your info.
AEO roadmap: Immediate steps to improve your "AI discoverability."
