Search Queries Are Getting Longer, Four Quarters Running

Six-word searches reached 6.57% of desktop Google queries by June 2026. Seven-word searches climbed too. Both have now extended their upward trend for four consecutive quarters, while five-word queries sat almost perfectly flat around 9.3% for the entire period.

Short queries are not disappearing. Something is being added on top of them, and it looks a lot like people talking to a search box the way they talk to an assistant.

Key takeaways

• Mid-length queries of six and seven words have grown for four straight quarters in the US, reaching 6.57% for six-word searches by June 2026.
• Short queries held steady rather than declining, so this is additive behavior rather than a wholesale replacement of how people search.
• In Europe the effect runs further, with six, seven, and eight word searches all rising and very long queries of fifteen words or more climbing from 0.53% to 0.69%.
• Informational and commercial research intent rose across every major engine over the same period, which is the same shift showing up in a second measurement.

What the numbers show

In the US, mid-length queries continued their climb through Q2 2026. Six and seven word searches both extended a trend that has now run four quarters without interruption. Very long queries of fifteen words or more stabilized after rebounding in Q1.

Europe shows the same pattern with more force. Six, seven, and eight word searches all increased during Q2, and the longest queries kept building rather than levelling off, rising from 0.53% in March to 0.69% by June. Five-word searches stayed pinned between 7.75% and 7.80% across the whole quarter.

The flat short-query line is the part I find most useful. If long queries were replacing short ones you would see the short band shrink, and it has not. People still type two words when two words will do. They are adding a new mode alongside the old one.

What six words actually looks like

The percentages are abstract until you say the queries out loud. A two-word search is plumber detroit. A six-word search is why is my water heater leaking. An eight-word one is how much does it cost to replace a water heater.

Those are not the same request wearing different clothes. The short one is a navigation instruction and the long ones are questions with an implied situation behind them. The person asking the second one has a problem, a timeline, and a budget they have not stated.

A page built to rank for the two-word version is a landing page. A page that answers the six-word version has to actually contain the answer, near the top, in words a machine can lift.

Where the longer queries are coming from

The report does not isolate a cause, so this part is my read rather than a finding. Two things line up with it.

The first is habit transfer. People who have spent a year typing full sentences into an assistant stop shortening them for a search box, because the search box stopped punishing them for it. Google handles a sentence now.

The second is that the intent data moves in the same direction. Across all five major engines, informational and commercial research queries increased over the year while navigational searches fell on every single one. Longer queries and research-shaped intent are the same behavior measured two ways, and I dug into the intent side of it in the piece on the searcher becoming software.

Europe running ahead of the US on both query length and AI adoption fits the same explanation, though a correlation across two regions is a long way from proof.

What this changes about content

The practical shift is away from writing pages about topics and toward writing pages that answer questions. A six-word query is usually a question, and a question has a correct answer that can be stated in a sentence or two.

That means a heading phrased the way a person would ask it, followed immediately by the answer rather than three paragraphs of throat-clearing before it. It is the format that gets pulled into an AI summary, and it is also just easier to read, which is the whole argument in AEO being mostly SEO with new acronyms.

It also means your keyword research is probably looking at the wrong shape of term. Six-word questions carry little or no reported search volume individually, so a volume-first process filters them out before you ever see them. The volume exists, distributed across thousands of phrasings of the same underlying question.

How I would find the questions

Search Console is the least glamorous and most reliable source. Filter your query report to phrases containing how, why, what, or can, and sort by impressions rather than clicks. Those are questions your site already appears for and probably does not answer directly.

Your inbox and your call log are better still. The questions customers actually ask, in the words they actually use, are the six-word queries before anyone types them, and they cost nothing to collect. I made the fuller case for reading real customer language rather than guessing at it in stop guessing how customers search.

Then write the answer plainly and put it where it can be found. That is the entire play, and it works for the human reading the page and the model summarizing it, which is what I was getting at in getting cited by AI as a service business.

What this data does not cover

These are desktop figures from a clickstream panel across the US, EU, and UK. Voice and mobile search skew longer and more conversational than desktop typing does, so the true drift toward sentences is probably understated here rather than overstated.

The shifts are also small in absolute terms. Six-word queries moving to 6.57% is a real trend across four quarters, not a sudden reordering of how everyone searches. Short queries remain the bulk of the volume and will for a long time.

FAQ

Four quarters in the same direction is long enough to stop calling it noise. People are asking search engines questions in full sentences, and the pages that win are the ones that answer the question instead of covering the topic.

Sources

1. Datos (a Semrush company), “State of Search Q2 2026: Behaviors, Trends, and Clicks Across the US & Europe.” Desktop clickstream panel, April 2025 to June 2026. datos.live
2. Same report, query length distribution section, on six, seven, and eight word searches across the US and Europe.
3. Same report, search intent section, on rising informational and commercial research queries and falling navigational intent across all five major engines.