Plan for Google Zero, Not Google #1

On a podcast last week, Conde Nast CEO Roger Lynch said his company is now assuming all of their search traffic will go to zero. He didn't say it might dip. He didn't hedge with "lower than we'd like." He said zero.

That's a publisher CEO saying his bread and butter is finished. Conde Nast owns The New Yorker, Vanity Fair, Wired, GQ, and Bon Appetit, brands that ranked for almost anything you could type into a search bar. Brands that built revenue on the assumption that being #1 for "best Dutch oven" meant traffic, and traffic meant ads. That bet is being called in public.

The web press has been calling this Google Zero for a while. It's the natural endpoint of a few overlapping shifts. AI Overviews summarize the answer on the results page, so the user never clicks. Claude, Gemini, ChatGPT, and Perplexity answer the question without ever rendering Google. The browser bar quietly turned into a chat box for a growing slice of users. Search traffic isn't slowly eroding — it’s being structurally rewritten. The work of adapting already has names: Generative Engine Optimization (GEO) and Answer Engine Optimization (AEO). Both center on the same question. How do you become the source an AI model cites instead of the result a search engine ranks?

We get asked about this in nearly every client kickoff now. Park district leaders ask if their event pages still matter. Nonprofit directors ask whether their blog is still worth maintaining. Small-business owners ask whether the SEO retainer is even worth signing. The honest answers aren't comfortable, but they're not catastrophic either. Here's how we're framing the work.

The cheap traffic is going away. The high-intent traffic isn't.

Not every search is the same. The "explain what a is" query is the one AI eats first. That was always low-intent traffic anyway: readers who bounced after one paragraph and never came back. What survives, for now, is the high-intent search. The parent typing "Hoffman Estates summer day camp registration" still has to land on the actual registration page. The donor researching a specific nonprofit's 990 still ends up on that nonprofit's site. The brand search isn't going anywhere.

So the work shifts. Stop writing thin content that competes for definitional queries you were never going to monetize anyway. Start writing content the model needs to cite: concrete facts, named programs, original numbers, photos of the actual building. Be the primary source, not the summary.

You need distribution you own.

If the algorithm is the landlord, you've been renting. Email lists, community memberships, alumni networks, donor portals, text alert systems — these are owned distribution. They're harder to build than a blog post. They're also harder to take away.

The nonprofits we work with that weather these shifts well all have something in common: a real email list with people who actually open it. The park districts that don't panic about search updates have a printed newsletter, a community Facebook group, or an SMS system that residents have already opted into. Owned channels are slower to grow and impossible to outsource. They're also the only channels that don't go to zero when the algorithm changes its mind on a Tuesday.

Be the answer the model cites, not just the result the user clicks.

When AI Overviews summarize, they cite. When Claude, Gemini, or ChatGPT answers a question, sometimes it links. The citation is now the floor of search visibility, not the ceiling. This is what GEO and AEO come down to in practice: structuring content to be selected as a source, not just surfaced as a link. To get cited, your content has to be the primary source for a specific, concrete claim. Vague roundups don't get cited. Statistics, original research, named case studies, clear product positioning, and unambiguous service definitions do.

Pages that win in this environment are the ones that say something true and specific that nothing else on the web says quite that way. That's harder to write than a top-ten list. It's also what good writing always was.

Treat your site like the canonical record, not a content marketing funnel.

The site stops being a top-of-funnel acquisition machine and becomes the place where the truth about your organization lives. Mission. Programs. Staff. Outcomes. Pricing. Policies. Contact. The model needs that information to cite you accurately. The human who finally clicks needs that information to make a decision. The journalist looking you up needs that information to write the piece.

A lot of nonprofit and city sites still treat their homepage as a campaign landing page for whatever push is happening that week. That worked when traffic came from search. It doesn't work when traffic comes from a referral, a citation, or an AI agent showing up cold. The canonical site is the one that respects the person who showed up not knowing anything and answers their actual questions in two clicks.

What this looks like in practice.

We've reworked a few site strategies around this thesis over the last six months. The pattern that's emerging: fewer pages, written better. Real case studies with numbers. Events pages that get indexed cleanly because they're plain, factual, and well-structured. Leadership pages that answer the "who runs this place" question without making the user dig three layers deep. Less listicle-style blogging, more reference content. Cleaner schema markup. Better internal linking that helps a model understand what your site is actually about. In short, GEO and AEO aren't a new playbook. They're what good SEO becomes when the model, not just the user, is reading your site.

The clients who panic about Google traffic dropping spend the next quarter chasing it. The clients who reframe the question — what do we own, what gets cited, what gets opened in an email — spend the next quarter building something more durable.

Roger Lynch isn't telling investors traffic will be zero because he's giving up. He's telling them so the next phase of the business gets built on something the algorithm can't take away. That's the right read for everyone else too. The traffic spigot was never going to last forever. The work now is to make sure your organization is still findable, still useful, and still trusted when the spigot is closed for good.