Major League Fishing's Owned Livestream Is the AI Overview Hedge That Pays

Major League Fishing's senior ad ops director Jared Collett told AdExchanger that publishers losing traffic to AI Overviews should pivot budget into video, because Google's summarizer can lift article text but cannot replace a watch-the-clip click. Reuters Institute's 2026 publisher survey backs the move: 79% of news publishers plan to produce more video, and a net 74% are increasing effort on YouTube. The threshold most teams underestimate is owning the player, not just the format.

Owning the player is the part that actually pays

The line from Collett's interview that should make every CMO pause is the part about revenue per viewer. MLF generates "significantly more revenue from website viewers than YouTube viewers," he said. That isn't a CPM trick or a creative win. It's a structural one. MLF runs livestream commercials with SSAI-stitched ads on its own infrastructure, then layers homepage takeovers, display banners, email, and social extensions on top. YouTube takes a 45% rev share before any of that math even starts, and it disallows most of the on-page sponsorship inventory MLF actually monetizes.

So when Collett tells publishers to make more video, the implicit recommendation isn't "go viral on YouTube." It's build a destination that AI Overviews can't summarize you out of, then put your video on it.

The contrast with text content is sharper than people give it credit for. Google's summarizer can rewrite your 1,800-word explainer into a paragraph and call it the answer. It cannot show your livestream. It cannot embed your interview with a competing angler. The asset is materially different at the substitution layer, and that's the entire reason video is functioning as Overview insulation right now.

79% of publishers are pivoting. Most won't get the math right.

The Reuters Institute's 2026 trends survey found that 79% of news publishers plan to produce more video this year, 71% want more audio, and a net 74% are increasing YouTube effort. Net +56% are increasing TikTok. X sits at -52, Facebook at -23. The platforms publishers still want to invest in are the ones where you upload a clip and someone actually watches the whole thing through to the ad break.

The same survey says 76% of publishers are getting their journalists "to behave more like creators," with 50% pursuing creator distribution partnerships and 31% planning to hire creators outright. The headline number that should worry CFOs: respondents expect a 43% decline in search engine traffic within three years.

That all sounds like aligned strategy. In practice, most publishers will spend the next 12 months posting clips to YouTube without ever building the owned-distribution layer that gives MLF its per-viewer margin. The vendors and consultants pitching this transition are happy to take video budget and ship it through Google's properties, which is the recursive loop that cornered publishers on text in the first place. Honestly, I'd expect roughly half the teams that say they're pivoting to video to end up subsidizing YouTube's ad business with cleaner production values, and not much else.

The traffic loss is real, and it isn't even

Specific numbers help calibrate the panic. Digital Content Next surveyed 19 of its member publishers (12 news brands, 7 non-news) and found Google Search referral traffic down a median 10% year-over-year across an eight-week window in 2025. Non-news brands took it harder at -14%; news brands sat at -7%. Larger antitrust filings have referenced declines as steep as 58% on AI Overview-eligible queries, and VideoWeek's coverage of the same publisher pivot found the channel mix moving toward direct destinations and creator partnerships rather than search.

The pattern in our coverage of the March 2026 core update looked similar from the visibility side. YouTube lost 567 visibility points and Reddit lost 64 in a single update window, but the recovery curves were not the same across formats. Original video did not get punished the way aggregated text did, which lines up with what Collett is describing from the inventory side. The publishers shipping host-led video on owned domains seem to be holding visibility better than text-first sites running the same topics.

Why "do more video" still leaves most teams stuck

Honest caveat: most marketing teams don't own a streaming sport. MLF has a captive audience watching multi-hour tournament livestreams, which is about the unfairest comparison set possible to a B2B SaaS blog or a niche e-commerce brand. So translating Collett's playbook is less about cloning his ad stack and more about asking which of your content assets are AI-substitutable and which aren't.

From what I've seen, the assets that survive Overview substitution share three traits:

• They show something happening in real time, not summarize it after the fact.
• They have a recognizable host or face, not just a byline.
• The publisher controls the player, the comments, and the email opt-in adjacent to the watch surface.

If you have one of those, you have leverage. If you have none, the pivot to video is mostly going to look like rebranded YouTube SEO with worse CPMs and a dependency on the same recommendation algorithm that already decided your text wasn't worth surfacing.

The reallocation worth testing in the next 30 days

For most marketing teams reading this, the practical move isn't a livestream studio. It's auditing which of your top-10 trafficked pages are Overview targets in the SERP, then reshooting the top three as a 90-second host-led video and embedding it on the same URL. That preserves the SEO real estate you already earned and gives Overviews something they can't summarize cleanly into a paragraph.

Search Engine Journal's coverage of publisher adaptations notes the same pattern: pages with substantive original video tend to retain visibility through Overview rollouts where text-only pages don't. The benchmark worth chasing: if your top-three Overview-vulnerable pages each carry a 90-second video by July 1, you should see a measurable lift in dwell time and a softer click-through hit than the -10% DCN baseline. If you can push that video out as a YouTube clip with the watch-page pointing back to your owned URL, you're moving in the same direction MLF already lives, just at startup scale.

The part that struck me most from Collett's interview wasn't the video advice. It was how matter-of-fact he was about article links being a closed chapter. "Linking to articles died kind of a long time ago," he said, almost as an aside. That's a senior ad ops director at a real publisher describing the open web's primary distribution mechanism in past tense. I'm not sure most marketing teams have caught up to how literally he means it. The blog post as a discoverable, link-driven asset has been quietly downgraded to something that mostly works for direct readers and email lists, and the search layer above it has been rerouted through a summarizer.

If your 2026 content plan still treats blog posts as the primary asset and video as the supporting clip you make later if there's budget, the inversion is the work. Reverse the priority order. Start the planning meeting with the video, then write the page around it. The teams that figure out that inversion this quarter will probably be fine. The ones who don't are going to keep producing service journalism for an audience Google has already decided not to send them.

Notice Me Senpai Editorial