Reddit’s CEO Steve Huffman said something during his most recent earnings call that every CMO should sit with for a moment. Talking about Google’s AI Overviews, he said: “10 blue links has driven tremendous value and growth to the broader ecosystem… from where we sit, AI Overviews has yet to make a similar level of positive impact, and I think that’s consistent across the broader landscape. As businesses, publishers, retailers, we’re still looking for that win-win.”
That is a public company CEO, in an investor context where measured language matters, saying out loud that the deal has changed and nobody has figured out the new terms yet.
If you are a CMO making budget and channel decisions right now, that should inform how you think about organic search as a reliable traffic source.
The number that actually matters
A Pew Research study looking at data from 900 US adults found that Google’s AI Overviews cut referrals to publisher sites by almost half compared to the traditional search results approach. Not a marginal dip. Not a rounding error. Close to half.
I want to be careful here: correlation and context matter. Google had already moved away from pure “10 blue links” results before AI Overviews arrived. The erosion of referral traffic is not a single-cause story. But when you pair that Pew finding with Huffman’s public comments, and with a Wall Street Journal report that Reddit, The Economist, Reuters, Politico, and USA Today are all reconsidering their data licensing relationships with Google, a pattern emerges that is hard to dismiss.
The implicit agreement between search engines and content publishers for decades was: you provide the content, we send you the audience. AI Overviews compress that content into a summary and answer the query without the click. The audience stays on Google. The referral never happens.
What this means for how you allocate
Most CMO traffic models still treat organic search as a relatively stable, cost-efficient acquisition channel. It earns that reputation over years of consistent performance. But the underlying mechanism driving that performance has shifted, and your model probably has not caught up yet.
Here is what I think the right posture looks like right now:
Audit your organic dependency. If a significant share of your site traffic, lead flow, or revenue pipeline runs through organic search, you need to understand what percentage of that comes from informational queries versus navigational or transactional ones. AI Overviews are most likely to intercept informational queries. If that is where your funnel starts, the top is getting narrower.
Do not assume last year’s benchmark is this year’s baseline. Traffic benchmarks from 2023 or even early 2024 were set under different search behavior conditions. Holding this year’s performance to those benchmarks without accounting for structural channel change will produce misleading conclusions about what is working.
Get curious about where the engaged audience actually lives. Huffman made an interesting argument in his investor letter: “As AI makes information more abundant, the challenge is no longer finding content. It’s finding context, personal opinion, and first-hand accounts.” He is talking about Reddit’s value proposition, but the underlying point applies to your brand too. If your content is the kind that gets compressed into a summary and consumed without a click, you need to ask what you are building that requires someone to actually show up. Start by mapping the topics your audience actually asks about, and where your brand genuinely has something first-hand to say.
Diversify the measurement logic, not just the channels. The instinct when a channel weakens is to shift budget somewhere else. That is often the right call. But the more important discipline is building measurement infrastructure that can detect channel-level shifts early, before they become budget crises. That is marketing run as a loop with a memory: read this month against last month and last year before you touch anything else, and a structural shift shows up as a trend you caught rather than a crisis you inherited. If you are still primarily measuring marketing success through last-touch attribution models anchored to organic search clicks, you are flying with instruments that do not account for the new terrain.
The honest difficulty here
I am not going to pretend there is a clean answer to this. Organic search has been a foundational channel for so long that the playbooks, the team structures, and the measurement frameworks of most marketing organizations are built around it. Rebuilding that infrastructure while it is still partially functional is genuinely hard. You cannot just declare organic search dead and move on, because it is not dead. But you also cannot operate as if the structural shift Huffman described is not real.
The job right now is to hold both of those things at once: continue working the channel while honestly reassessing its long-term role in your mix. That requires measurement clarity, not optimism.
Find out how to grow your marketing with the professional marketing minds at Ambient Array.
Original source: As Reddit stock falls, CEO questions value of Google’s AI Overviews