The Software People Actually Research Looks Nothing Like What the Press Covers

If you read the tech press, you would think the entire software market is three chatbots and a coding assistant. Then you look at what people actually research when they sit down to choose a tool, and the list is almost unrecognizable. Over 30 days we logged 121,435 tool-profile page views across 10,501 distinct tools on Toolradar. The tools people look up most are practical, private, and often self-hosted. The AI products that own the headlines are barely in the picture.

The dataset

This report is built from tool-profile page views on Toolradar over a rolling 30-day window: 121,435 views across 10,501 distinct tools, aggregated per tool across its profile, pricing, and comparison pages. A page view here means someone opened a specific tool's profile to look at it, which is a research action, not a passing headline. This is first-party behavioral data from Toolradar, released under CC BY 4.0.

We are measuring demand, what people choose to look up, not our opinion of what is good. The two often disagree, which is exactly why the data is useful.

The tools people actually research

Look at the shape of this list. Half of it is note-taking and personal knowledge management (Obsidian, Joplin, Simplenote, Notion, Logseq just outside the top fifteen). A big chunk is privacy and self-hosting (Nextcloud, Proton Mail, Mega, Cloudflare Tunnel). The AI names that do appear are utility tools people are evaluating for a job (NotebookLM for research, HeyGen and ElevenLabs for media), not the general-purpose chatbots that dominate coverage. Duolingo tops the list because millions of people compare it to alternatives, not because the tech press writes about it weekly.

Research demand versus press coverage

Put this next to what the press covers and the gap is the whole story.

Almost no overlap. Coverage tracks the AI news cycle. Research tracks the actual work people are trying to get done: taking notes, owning their data, running a home server, picking a productivity app. If you build content, tools, or a media strategy around what the press covers, you are optimizing for a small, crowded, hype-driven slice of demand and ignoring the much larger, quieter market where buyers are genuinely shopping.

The categories buyers compare in

Category-level demand tells the same story. These are the most-viewed best-of guides in the window.

AI video and voice are here, which is real. But so are SMS marketing, push notifications, and headless CMS, categories the tech press almost never writes about. Demand is spread across the working software market, not concentrated on the four tools everyone is talking about.

What this means

• For buyers: the tools worth your time are not always the ones you have heard of. The most-researched software skews toward practical, privacy-respecting, and self-hostable options that never trend. If you only shortlist what you have seen in headlines, you are missing where the market actually is.
• For anyone building or covering software: research demand is a far better guide than press coverage for what to prioritize. It points at real, underserved categories instead of the same handful of AI launches. Note-taking, privacy tooling, and self-hosting are quietly enormous.
• For the AI-is-everything narrative: it holds in the press and collapses in the research data. People are curious about AI and are actively choosing note apps, email, and workflow tools. Both things are true, and only one of them makes headlines.

Methodology

Source: 121,435 tool-profile page views on Toolradar over a rolling 30-day window ending August 2026, aggregated per tool across profile, pricing, and comparison pages, covering 10,501 distinct tools. Category figures are best-of guide page views over the same window. A profile view is treated as a research action. This is descriptive first-party behavioral data, released under CC BY 4.0. Reuse it with a link back to this page.

Frequently asked questions

Why is Duolingo the most-researched tool?

Because millions of people compare language apps, and Duolingo is the reference point they check first. Research volume measures how many people look a tool up when making a decision, and consumer-scale tools with lots of alternatives draw the most comparison traffic. It is a demand signal, not a quality ranking.

Why do privacy and self-hosting tools rank so high?

Nextcloud, Proton Mail, Joplin, Simplenote, Mega, and Cloudflare Tunnel all sit in the top twenty. That reflects a real, sustained trend: people actively shopping for alternatives to default cloud and productivity stacks over data ownership and cost. It rarely makes headlines, but it shows up clearly in what people research.

Does high research demand mean a tool is good?

No. This is a demand signal, how many people look a tool up, not an editorial ranking. A tool can be heavily researched because it is popular, controversial, or has many alternatives to compare. For quality we use external review strength and editorial review, described on the how we rate page.

How is this different from press coverage?

Press coverage measures what journalists write about, which is dominated by AI launches. Research demand measures what buyers look up when choosing, which is dominated by practical, everyday software. The two lists barely overlap, which is the main finding of this report.