These things don't happen by accident. Our people operations and leadership teams have worked tirelessly to keep the best of our in-person culture alive as Level has transitioned to a fully remote world.
Level's Vibe Squad, our employee-led culture committee, recreates the moments that used to happen by the coffee maker through virtual hangouts and small celebrations. Bonusly, a peer recognition platform, makes recognition a daily habit: teammates send real-time shoutouts and reward points the moment someone does great work. And the companies we've acquired have brought more than headcount. Each one carried its own cultural rituals, and we've built the best of them into Level.
And yes, we do still get together in person quite regularly.
It's not complicated, but it's also not automatic. It requires consistent nurturing, every week, for years, before it becomes culture instead of a program.
The same discipline shows up in how we handle continuous professional development and learning, the part of this job that never stops moving.
Platforms like Google and Meta make changes multiple times a year. AI changes something every other day. We treat staying current the same way we treat everything else in our culture: not a special initiative, just something we do.
A team that trusts each other enough to stay close from a thousand miles apart is the same team that shares a hard truth with a client about a campaign before they have to ask.
After all, what really makes a company a great place to work is how people show up for one another.
Let’s get into this week’s updates.
Patrick Patterson
CEO at Level Agency
What's interesting this week
We're now Great Place To Work Certified
This is the full announcement behind our Great Place To Work® Certification™: 84% of our team said yes, this is a great place to work, 27 points above the U.S. average. It includes the full breakdown from our Trust Index™ Survey, plus quotes from the team on what they love about working at Level.
Key takeaways: Congrats to the whole team. Numbers like this don't happen without a lot of people showing up for each other consistently.
Only 40% of searches now send a click to the open web
New Datos x SparkToro data puts the open web's share of U.S. search clicks at a record low of 40%. The EU/UK sits only slightly higher, at 40.7%. Google-owned properties now capture 17.1% of U.S. searches, while paid click share fell hard, from 2.3% to 1.3% between March and June.
Key takeaway: Zero-click behavior keeps climbing, and AI Mode is only part of the story: Google's own properties are absorbing a lot of that share too. Paid click share is shrinking as well, and Wikipedia's climb in the EU search rankings suggests people still verify AI answers against real sources.
Reddit's Q2 revenue hit $805 million, up 61% year over year, with net income up 183% to $253 million. Guidance for next quarter beat expectations too. Shares slid over 10% after hours because CEO Steve Huffman warned that search referral traffic turned choppy, and U.S. daily active users dipped slightly quarter over quarter.
Key takeaway: Investors are pricing in AI's hit to referral traffic even though the underlying business is strong, and Reddit hasn't committed to renewing its data licensing deal with Google past this year. Any brand relying on Reddit for organic reach or community-led SEO should watch that relationship closely.
LinkedIn is testing a feature that lets members privately report posts and comments that read as AI slop, feeding those flags back into its content-quality models. Chief Product Officer Hari Srinivasan also confirmed LinkedIn is retiring its "enhance your post" AI writing tool, replacing it with a proofreading feature that fixes grammar without rewriting voice.
Key takeaway: Posters will get a private analytics alert if their content gets flagged as inauthentic or overly AI-generated, and LinkedIn joins Pinterest, Snapchat, TikTok, Reddit, and X in building this kind of detection into the platform. If your team ghostwrites executive LinkedIn content with AI, treat this as a signal to keep a human editing pass in the loop.
IAB moves to standardize how brands measure AI visibility
The IAB found more than 20 vendors selling AI visibility measurement tools, with almost no agreement between them on methodology or results. Its new "Measuring Visibility in the AI Era" framework tries to fix that with shared vocabulary: four Ps of visibility (presence, prominence, portrayal, and persuasion) and two measurement tiers.
Directional measurement covers early signal detection and internal briefings. Decision-grade measurement is the tier built to guide budget and strategy calls, requiring real sample size, query volume, and reproducibility standards. The IAB is deliberately starting with organic AI visibility only, treating paid and commerce attribution as a later priority.
The stakes explain the urgency. McKinsey estimates brands that don't adapt could see traffic declines as high as 50% from traditional search. Separately, LQ research found more than 40% of brand citations that show up in organic search never appear in AI Overviews for the same query.
Key takeaways: Ask any AI visibility vendor which tier their numbers fall into before you act on the report, and remember that organic citation share and AI Overview citation share are two different metrics worth tracking separately. With only 16% of brands measuring AI search performance systematically today, having a real framework is a competitive edge in itself.
I hope you caught this one last week. I mentioned it during our Virtual Insider Session on AI in creative, but it's worth resharing because the process behind it is the real story.
What separates good AI creative from AI slop
One of our creatives built a mock LEVEL soda can from scratch to test a simple question: does one well-written prompt produce usable creative? The prompt named the mood, the lighting, the wardrobe, and the product in specific detail. The output was still slop, with porcelain skin, a wig-like hairline, and label text that blurred into gibberish.
The version that actually worked took eight steps. The can got its own prompt and a 3D render before any scene existed, locking in the label and proportions first. Midjourney generated dozens of moodboard directions, a person picked the strongest one, and the can got composited into the shot with Krea handling the realism pass. Magnific.ai upscaled the image, and the final pass in Photoshop put the skin's freckles and asymmetry back in, since the model strips them by default.
No single tool ran that whole chain. The judgment calls, like which frame to pick or when to break symmetry, stayed with a person at every step.
Key takeaways: A well-written prompt and a polished process are not the same thing: slop and good work can start from an identical brief, and the real difference is the manual finishing pass that undoes what the model defaults to. If you're evaluating an AI creative tool for your team, ask what it does at step seven, not step one.
Related signal: If you missed last week's live session with Carly Fridhandler and Mike Rubino on where AI actually belongs in the creative process, including the two things Carly won't let AI render and what early performance data shows on AI-made ads versus human-made ones, the full recording plus the 15-minute audience Q&A is available now.
If you were sketching the IAB's new measurement tiers or a client's AI visibility framework this week, Mermaid turns that description into an actual diagram instead of a whiteboard photo. Describe the flow in plain language, paste in a document, or write it in Mermaid's own syntax, and it builds flowcharts, sequence diagrams, and org charts on the spot.
It connects to Claude, VS Code, and Cursor through its own MCP server, so a strategy doc or a messy meeting transcript can become a shareable diagram without leaving the tool your team already works in.
Your AI visibility, scored.
When did you last audit what AI thinks of your brand?
Don't rely on an unreviewed algorithm to manage your brand’s reputation. 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:
Visibility score and share of voice: The percentage of AI prompts your brand appears in, benchmarked against your top competitors across all major models.
Sentiment and positioning: Whether AI describes your brand favorably, neutrally, or negatively, and whether you're showing up as the primary recommendation or an afterthought.
Optimization roadmap: The specific content structure and authority signals you need to improve your score.
