Single-purpose study, writing and translation tools lost web visits over the twelve months to April 2026 even as chatbots grew 90.82% year over year, according to OneLittleWeb, whose analysis of 9,531 AI tools found Chegg down 61.45%, Quizlet down 29.09% and Grammarly down 15.45%.

The study, published July 13, 2026 and last updated July 28, 2026, tracked estimated web visits across more than 170 categories of AI software, drawing on monthly data from Semrush and Ahrefs covering May 2024 through April 2026. It is authored by Sujan Sarkar, who runs the search engine optimization agency OneLittleWeb.

Chatbots and single-task tools moved in opposite directions inside the same market. According to OneLittleWeb, the chatbot category grew from 45.1 billion to 86.0 billion web visits over the latest twelve-month period, a rise of 90.82%, giving it a 59.49% share of all AI tool traffic tracked in the study. Translators, the category that includes Google Translate and DeepL, fell 22.52% over the same period, losing 1.6 billion visits.

What the numbers show

The study drew on OneLittleWeb's AI Tools Dashboard, a database the company built by combining estimated monthly web visit figures from Semrush and Ahrefs for 9,531 tools across more than 170 categories. Together, those tools generated 144.5 billion web visits over the twelve months to April 2026, up 40.12% from the prior year. The Top 100 tools alone accounted for 129.6 billion of those visits, or 89.69% of everything tracked, meaning nearly nine out of every ten visits to an AI tool went to just 100 platforms.

Set against the wider internet, the figure looks smaller. OneLittleWeb's separate Most Popular Websites dashboard, which tracks 10,173 major domains including the AI tools in this study, recorded 3.44 trillion visits over the same period. AI tools accounted for 144.5 billion of that total, or 4.20%.

Inside the AI category, the concentration sharpens further. ChatGPT generated 64.7 billion web visits over the twelve months to April 2026, up from 38.1 billion the year before, a gain of 69.80%. That single tool captured 44.76% of all AI tool web visits tracked in the study, and generated nearly as many visits on its own as the other 99 tools in the Top 100 combined. Canva ranked second with 10.5 billion visits, followed by Gemini at 6.9 billion, a fivefold increase from 1.3 billion the year before.

Growth was not confined to ChatGPT. Gemini added 5.6 billion visits, the second-largest absolute gain of any tool in the study, rising 450.33% year over year. Claude added 2.4 billion visits, up 250.07%, moving from 958.0 million to 3.4 billion. Grok posted the fastest percentage growth among the ten largest chatbots, up 731.26% to 2.5 billion visits, while Perplexity nearly doubled to 2.3 billion, a gain of 98.38%.

Where the declines concentrated

Six named tools illustrate a pattern OneLittleWeb frames as a shift away from single-purpose software. Chegg, a homework-help and study platform, fell from 281.2 million to 108.4 million visits, a drop of 61.45%, the steepest percentage decline recorded among the study's Top 100 tools. Quizlet, a flashcard and study application, fell 29.09%, from 1.1 billion to 782.7 million visits, a loss of 321.1 million, the second-largest absolute decline in the Top 20 behind Google Translate. QuillBot, a paraphrasing and academic writing tool, dropped 25.24%, from 862.6 million to 644.9 million visits.

Grammarly fell 15.45%, from 823.2 million to 696.0 million visits, a loss of 127.2 million. According to OneLittleWeb, every month in the latest twelve-month period finished below the corresponding month a year earlier for Grammarly, even though the tool still held 69.60% of all web visits inside the Grammar Checkers category. A separate Grammarly product, its AI plagiarism checker, fell further still, down 38.10% to 108.4 million visits.

Google Translate recorded the largest absolute decline of any tool tracked in the study, losing 1.5 billion visits, down 23.46% to 4.8 billion. OneLittleWeb's monthly breakdown shows the tool falling below 400 million visits in November 2025 and reaching a 24-month low of 314.7 million in February 2026, with a brief recovery to 419.2 million in March before dropping again to 349.6 million in April. Despite the decline, Google Translate still controlled 83.66% to 84.27% of all web visits inside the Translators category, depending on which of the study's two tables is used, indicating the category itself lost ground rather than losing share to a rival. DeepL Translate fell 12.69%, from 758.4 million to 662.2 million visits, with its final six months down 24.1% against its first six.

The pattern extended across the translation category as a whole. Eight of the ten largest AI translators tracked in the study lost web visits. Papago fell 28.79%, Reverso Text Translation dropped 41.08%, and Rask AI declined 37.94%. Only two of the ten grew: OpenL and MachineTranslation.com, which together added roughly 2.1 million visits, a fraction of what Google Translate alone lost.

The interpretation OneLittleWeb offers

According to OneLittleWeb, the pattern suggests some study, writing, translation and research tasks are shifting from single-purpose tools into general-purpose AI assistants capable of handling several tasks inside one conversation. The company's published analysis states that Grammarly's decline "may reflect a broader change in how people access writing assistance, with grammar correction and rewriting now available inside chatbots, document editors, and other general-purpose AI tools." For Quizlet, OneLittleWeb notes that the tool still accounted for 95.55% of visits inside the Students Tools category, meaning no rival study app absorbed the lost traffic. Instead, the company writes, "the rise of broader AI tools that can explain lessons, create quizzes, summarize notes, and answer follow-up questions may be changing how students choose to study online."

The study is explicit about what its methodology can and cannot show. It measures estimated web visits only, not unique users, revenue, subscriptions, app installations or API traffic. A decline in web visits to Grammarly's website, for instance, does not necessarily mean fewer people are using Grammarly's browser extension or desktop application, both of which operate without generating a tracked page visit. OneLittleWeb states plainly: "This study measures estimated web visits, not unique users, revenue, subscriptions, app usage, or API usage. Web visits are useful for comparing public demand and discovery patterns across tools, but they do not capture every way people use AI products."

Category-level detail

Ranked by total category traffic, chatbots led all AI tool categories at 86.0 billion visits and 59.49% market share, more than 2.6 times the combined total of the next nine categories in the Top 10. Design Generators ranked second at 10.9 billion visits, growing 12.87%, a category almost entirely dominated by Canva, which held a 96.28% share within it. Translators ranked third at 5.7 billion visits but moved in the opposite direction from every category above it, falling 22.52%.

Not every category followed the pattern of chatbot gains and single-purpose losses. Image Generators remained nearly flat, up just 0.19% to 3.5 billion visits, while individual tools within the category moved sharply in different directions: SeaArt.ai grew 78.81% and Photoroom grew 41.69%, even as the category leader, Magnific, formerly branded as Freepik, lost 7.77%. Video Generators grew 31.24% to 2.6 billion visits, the strongest growth rate outside chatbots among the ten largest categories. Background Removers and Sales Tools both declined, down 5.13% and 9.15% respectively, in each case because a single dominant tool, Remove.bg and Salesforce, pulled its category backward even while retaining more than 80% of category traffic.

Context for the marketing and publishing industry

The OneLittleWeb figures land inside a debate that PPC Land has tracked across several measurement releases through 2026. Similarweb data covered by PPC Land in June 2026 showed ChatGPT's worldwide traffic share falling from 76.4% to 52.7% over twelve months, while Claude's share nearly tripled to 8.9% and Gemini reached 27.3%. That report measured traffic share within the chatbot category rather than absolute visit volume, which is why its findings differ in shape from OneLittleWeb's figures without necessarily contradicting them: a shrinking share of a fast-growing category can still coincide with rising visit counts in absolute terms, which is exactly what OneLittleWeb records for ChatGPT.

A January 2026 Similarweb dataset reported by PPC Land found consumer EdTech platforms down 5% year over year as a category, with Chegg falling 68%, Mathway 63%, and Coursehero 48%, while Grammarly's decline in that dataset measured a comparatively modest 7%. The percentages differ from OneLittleWeb's figures because the two studies use different measurement windows, different underlying data providers and different category definitions, but the direction is the same: Chegg's decline appears as the steepest among named study tools in both datasets, and Grammarly's decline, while present in both, is consistently smaller than the declines recorded for Chegg, Quizlet or QuillBot.

Comscore data covered by PPC Land in May 2026 measured Claude's U.S. desktop growth at 130.1% month over month in March 2026, reaching 2.66 million unique visitors, a different metric again from OneLittleWeb's 250.07% year-over-year web visit growth figure for the same product. None of these figures are directly comparable on a like-for-like basis. Comscore counts unique visitors on desktop devices in the United States; Similarweb measures traffic share within a defined chatbot category worldwide; OneLittleWeb measures estimated total web visits, combining Semrush and Ahrefs data, across all device types worldwide. A marketer citing any single figure as the definitive measure of AI adoption risks conflating incompatible units.

This measurement fragmentation itself has become a recurring theme in PPC Land's coverage. Microsoft Clarity data reported by PPC Land in December 2025 found AI referral traffic to publisher sites growing 155% over eight months while converting to sign-ups at 1.66%, compared with 0.15% from organic search, an eleven-fold conversion advantage. That study, like several others PPC Land has covered, measured AI referral traffic arriving at third-party sites, a different phenomenon from OneLittleWeb's measurement of direct visits to the AI tools themselves. Google added a dedicated AI Assistant channel to Google Analytics in May 2026, as PPC Land reported at the time, automatically classifying referral traffic from ChatGPT, Gemini and Claude, a response to the same measurement gap that studies like OneLittleWeb's are attempting to fill from a different angle.

For publishers and content platforms whose business models depend on direct site visits rather than API integrations, the decline recorded for Chegg, Quizlet, QuillBot and Grammarly carries a specific implication distinct from the broader AI Overviews traffic debate that has dominated search industry discussion through 2026. AI Overviews concerns whether Google's search results page still sends clicks to publisher websites. OneLittleWeb's study concerns a related but separate question: whether users who once visited a specialized tool directly, without an intervening search, are now routing that same task through a general-purpose assistant instead. The mechanism differs. AI Overviews intercepts a search query before it reaches a publisher's page. The pattern OneLittleWeb describes intercepts a task before it reaches a specialized tool's homepage. Both reduce the same metric, web visits, but the competitive response required is not identical: a study platform losing visits to a chatbot competes on task completion and depth of feature set, not on search ranking.

Methodology details

OneLittleWeb's dashboard combines two commercial estimation sources, Semrush and Ahrefs, rather than relying on either alone. Both platforms estimate web traffic using clickstream panels, DNS data and other indirect signals rather than server logs or first-party analytics, a limitation common to nearly all third-party web traffic estimation tools and one the study does not claim to overcome. The Top 100 ranking compares total web visits from May 2025 to April 2026 against the preceding twelve months to calculate year-over-year change. Market share figures are calculated against the full 144.5 billion visits generated by all 9,531 tools in the dataset.

The study explicitly includes several tools that are not AI-native products in the conventional sense. Salesforce, Zoom, HubSpot and Qualtrics all appear in the Top 100 ranking despite being established software platforms that have added artificial intelligence features rather than AI-first startups. OneLittleWeb states these tools are included "because AI functionality is a core component of the overall product experience or classification," a methodological choice that affects the interpretation of category totals. Sales Tools, for example, recorded a 9.15% category decline driven almost entirely by Salesforce's own 10.17% drop, a figure that says more about enterprise software purchasing cycles than about generative AI adoption specifically.

Timeline

  • May 2024: OneLittleWeb's 24-month measurement window begins, tracking monthly web visits across 9,531 AI tools.
  • December 2025: Microsoft Clarity reports AI referral traffic to publisher sites grew 155% over the prior eight months.
  • January 2026: Similarweb data covered by PPC Land shows consumer EdTech platforms down 5% year over year, with Chegg falling 68% in that separate dataset.
  • February 2026: Google Translate reaches a 24-month low of 314.7 million monthly visits, according to OneLittleWeb's tracked data.
  • May 2026: Google adds a dedicated AI Assistant channel to Google Analytics, automatically classifying referral traffic from ChatGPT, Gemini and Claude.
  • May 2026: Comscore reports Claude's U.S. desktop unique visitors grew 130.1% month over month in March 2026.
  • June 2026: Similarweb data covered by PPC Land shows ChatGPT's worldwide chatbot traffic share falling to 52.7% while Claude's share nearly triples to 8.9%.
  • April 2026: The study's twelve-month measurement window closes; Google Translate visits fall to 349.6 million for the month.
  • July 13, 2026: OneLittleWeb publishes "Inside the AI Tools Market 2026," its 24-month study of 9,531 AI tools.
  • July 28, 2026: OneLittleWeb's study receives its most recent recorded update.
  • August 20, 2026: OneLittleWeb representative Alex Larson first sends the study's findings to PPC Land by email.
  • August 24 and August 26, 2026: Follow-up emails from OneLittleWeb reiterate the study's core findings on declining study, writing and translation tools.

Summary

Who: OneLittleWeb, a search engine optimization agency, authored by Sujan Sarkar, published the study. Chegg, Quizlet, QuillBot, Google Translate, Grammarly and DeepL Translate are the named tools that lost web visits; ChatGPT, Gemini, Claude, Grok and Perplexity are the named chatbots that gained.

What: A 24-month analysis of estimated web visits across 9,531 AI tools found that chatbots grew 90.82% year over year to 86.0 billion visits while Translators fell 22.52%, Chegg fell 61.45%, Quizlet fell 29.09%, QuillBot fell 25.24% and Grammarly fell 15.45%.

When: The study covers web visit data from May 2024 through April 2026. It was published July 13, 2026, and last updated July 28, 2026. OneLittleWeb representatives sent the findings to PPC Land by email on August 20, August 24 and August 26, 2026.

Where: The data covers global web visits, with country-level detail showing Singapore, the Netherlands and the United States as the highest-adoption markets by visits per capita.

Why: The figures matter to marketers, publishers and product teams building single-purpose AI tools because they indicate that visit volume is consolidating around general-purpose assistants even as the AI tools market as a whole continues to expand, a pattern with direct implications for how specialized software competes for user attention against bundled alternatives.