AI Overviews Cut Google Clicks to 8%, and 'Safe' Marketing Pays the Price

Users click a traditional Google link in only 8% of visits when an AI summary sits atop the results page, down from 15% without a summary. That figure comes from a Pew Research Center study of nearly 69,000 searches, and it anchors a new Entrepreneur contributor column arguing that "safe" marketing — the copied channels, messaging and metrics of the category leader — is now the riskiest plan on the table.

"The safest-looking marketing plan in the room is usually the one most likely to keep you invisible," the author writes. Copying the leader means competing on terms that favor whoever has the bigger budget and the longer head start. "That feels like prudence. It is actually the slowest and most expensive way to stay a follower."

Why does safety fail?

Best practices earn the label because everyone already uses them, the column argues. By the time a tactic is safe enough to be named one, it is crowded and drained of whatever edge it once held. A brand can execute it flawlessly and still finish exactly where it started — one more voice repeating the same message in the same place as its competitors.

The same trap distorts measurement. Chasing the impressions, followers and engagement numbers every rival reports is "the most crowded play in marketing, because those figures are easy to hit and easy to ignore," the author writes. They make a quarter look healthy while the business underneath sits still.

Safety also carries a psychological pull. Failure is easy to defend when you followed the same plan as everyone else, and nobody gets second-guessed for matching the industry norm. The column's verdict: "a plan built to avoid blame is not built to win," and markets reward brands willing to be wrong in public before they are proven right.

What does the author's own track record show?

The column cites a 2013 call. The author argued then that mobile apps needed a strategy of their own rather than a shrunken version of web marketing, at a time when that position "sounded like a distraction." The people who moved before consensus formed didn't have better tools. They simply refused to wait for permission, and had already taken ownership of the space by the time the cautious crowd arrived.

Bold, the author stresses, does not mean reckless. The recommended structure is an asymmetric wager: a small, survivable downside paired with the upside of owning territory no rival has claimed. The highest-return version refuses to fight inside a crowded category at all — instead of scrapping for a sliver of space the leader defined, a brand reframes the problem so it becomes the obvious answer to a question nobody else is asking.

The column offers a one-meeting pressure test. Ask what happens if the move fails — whether the honest answer is a bruise or a catastrophe. Then ask what happens if it works — whether the payoff is a marginal gain or a foothold nobody can take from you. "When failure is survivable and success is defining, hesitation becomes the expensive choice."

What does AI search change for marketers?

The Pew data gives the argument its sharpest edge. Ranking a page and waiting for the click is losing its footing, yet most brands are still optimizing for rules that have already changed, the author writes.

AI systems are becoming the layer between the customer and the answer, and "the brands that appear inside those answers are the ones deciding today to be there." Waiting until the behavior is undeniable means arriving after the winners are chosen.

Winning that visibility depends less on budget than on being the clearest and most credible source on the questions a brand wants to own. When an AI system builds an answer, it pulls from information it can trust and quote, so the work is making expertise easy to find and hard to leave out.

In practice, the column prescribes unglamorous fundamentals:

• Publish direct answers to the questions customers actually ask, in language a machine can lift without guessing at meaning.
• Keep pages clean enough that facts are easy to extract and quote.
• Earn mentions on the outside sources these systems already trust, so your material is the evidence within reach when an answer gets built.

"That is unglamorous work, and it is exactly why most of your competitors will skip it until the advantage is gone," the author writes.

The closing warning lands on the cost of inaction: playing it safe shields a brand from the danger it can see and hands it the one it cannot — the slow erosion of being easy to overlook. As AI answers keep absorbing the clicks that traditional rankings once delivered, the brands that publish machine-quotable expertise now will be the ones those systems cite when the rest of the market finally moves.