SOFIA: The spice-and-recipe site reformatted its content to chase featured snippets, those answer boxes at the top of search, restructuring articles into snippet-friendly chunks, and traffic dropped. The assumption was that winning the box means winning the click. Sometimes it’s the opposite. A snippet that fully answers the question can mean the searcher never needs to click through at all.
ELENA: So the format change optimized for a position that doesn’t always pay, and may have cost something to get.
SOFIA: Right. They reshaped good articles into terse question-answer blocks to fit the box, and in doing so may have stripped the depth and flow that kept readers engaged once they did arrive. So they optimized for a feature that can satisfy the searcher in the result itself, while degrading the page for everyone who clicks.
MARCUS: Hold on, I want to challenge the blanket retreat before they rip all the formatting back out, because some snippets are absolutely worth winning.
SOFIA: Meaning it’s not that snippets are bad, it’s that they’re not uniformly good.
ELENA: And the no-click side of this got bigger, not smaller, which the team should factor in. With AI-generated overviews now sitting above many results and synthesizing the answer directly, the pull toward answering in the search page and never sending a click has intensified. So chasing a box on a query that an overview already resolves is chasing visibility that converts to nothing. The queries still worth winning are the ones where the searcher needs more than a synthesized summary can give.
MARCUS: Exactly. A snippet for “how long to toast cumin seeds” might fully answer it and cost the click, true. But a snippet for “best spices for a chili blend” puts you at the visible top, builds brand recognition, and the searcher often still clicks for the full list and the buying options. So the question per query is whether winning the box satisfies the searcher completely or just teases them into clicking. Chase the snippets that lead to clicks, not the ones that replace them.
SOFIA: That’s the correction, the value of a snippet depends entirely on whether the query is fully answerable in a box or needs the page.
HANNAH: And there’s a way to tell which is which, rather than guessing per query. Look at the query type. A simple factual question, a temperature, a substitution ratio, a yes-or-no, is fully answerable in the box, so winning it often means losing the click. A query that needs depth, a full recipe, a comparison, a technique with steps, can’t be satisfied in a snippet, so the box drives clicks instead of replacing them. It also helps to know the snippet format the query pulls, a paragraph box for a definition, a list for steps or rankings, a table for comparisons, because that tells you how to structure the answer to win the spot people still call position zero. The split is answerable-in-a-sentence versus needs-the-page.
NOAH: The pattern is optimizing for a visible win that isn’t the actual goal, the snippet is a trophy you can see and chase, but the goal was traffic and engagement, and the trophy sometimes works against both. The tell is restructuring content around a SERP feature rather than around the reader, formatting for the box first and the visitor second.
THEO: So the rule is to pursue snippets selectively, by whether they drive clicks, and never at the cost of the page. For queries that need the page, a recipe, a comparison, a guide, structure a clear answer that earns the snippet and pulls the click into the full content. Winning the box is mechanical once you decide to, phrase the question as a heading the way searchers ask it, put a concise direct answer right under it in the format the box wants, a short paragraph for a definition, an ordered list for steps, a table for a comparison, then let the full depth follow for the reader who clicks. For purely factual queries where the box ends the search, decide whether the brand visibility is worth a likely no-click, and don’t degrade a good article to chase one. Win the snippets that lead somewhere, and keep the page strong for the reader who arrives.
HANNAH: And you can confirm whether a won snippet is actually helping by watching the click-through, if you take the box and clicks fall, that’s a no-click snippet you may not want.
THEO: So the click-through data tells you whether a snippet you won is paying or quietly costing.
AIKO: Operationally this means snippet targeting is a per-query decision in the content plan, not a sitewide reformat. Tag the target queries by whether they’re snippet-and-click or snippet-and-done, format for the box only where it pulls the reader in, and monitor click-through on the ones you win so a no-click snippet gets caught rather than celebrated. And the page structure serves the reader first, with snippet-friendly clarity as a bonus where it fits, never a reformat that hollows out the content. The feature serves the strategy, the strategy doesn’t bend to the feature.
DANA: The move is selective snippet pursuit, never at the page’s expense. Traffic dropped because winning the box can mean the searcher never clicks, and reformatting good articles into terse blocks degraded the pages for everyone who did. Per Marcus, snippets aren’t uniformly bad, the ones for queries that need the full page, recipes, comparisons, guides, drive clicks and brand visibility, while the ones for simple factual queries can end the search in the box. So per Hannah’s split, we chase the snippets where the query needs the page and the box pulls the click, decide case by case on the purely factual ones, and watch click-through to catch no-click snippets we won. And we never hollow out a good article to fit a box. The instinct that top visibility matters was right. Chasing the feature regardless of whether it leads to a click, and degrading the page to do it, is what cost the traffic.
SOFIA: A snippet is worth winning only if the reader still has a reason to click. Win the ones that open the door, not the ones that close it.
DANA: The answer box isn’t the goal, the visit is. Take the snippets that lead readers in, skip the ones that end the search, and never gut the page to fit the box.