How DawnTrend works
DawnTrend is a research site that measures where public attention is moving across AI companies, startups and markets. It uses monthly Google search demand for each company’s own name, from Google search estimates, and compares it with website traffic where that is available. It covers the S&P 500, the Nasdaq-100, the Fortune 100, every Y Combinator company and 80 curated AI companies and topics, grouped into sectors and industries with a fixed formula: the Search Score, from 0 to 100, blends scale, twelve-month momentum and staying power. Browsing is free. A free account adds a daily search allowance, bulk research, website lookups and an API. Search demand measures interest, not revenue, users or adoption. Every figure is an estimate shown with its source, the months it covers and the date it was collected, and none of it is investment advice.
What is the mental model?
Attention comes before adoption, and adoption before revenue. DawnTrend measures the first two public layers and says plainly where public data stops.
| Layer | What happens | Measured here? | Source or note |
|---|---|---|---|
| Awareness | People hear about a company: launches, press, social posts. | No | Loud but hard to attribute. Not measured here. |
| Intent | People search for the company by name. | Yes | Google searches per month |
| Visits | People arrive on the company’s website. | Yes | Site visit estimates for Y Combinator companies |
| Adoption | People sign up, download and keep using the product. | No | Only the company knows this. Pair DawnTrend with product analytics. |
| Revenue | People pay. | No | The outcome everyone cares about, and the last to show up. Not measured here. |
How should I read the charts?
- Compare shapes, not levels. A small company doubling matters more than a giant staying flat. Every chart can be read as an index that starts at 100.
- A brand search is not a market size. It counts people looking for one name. It says nothing about how many people need what the company sells.
- Google reports volume in buckets. Small brands often read exactly 0% because the bucket did not change. For groups of companies we therefore count how many are rising or falling, instead of trusting a median.
- Every number carries its dates. Each page states the months it covers and when the data was collected. Queries are reused from cache for seven days, so the same question never costs twice.
How is the Search Score calculated?
Scale 40% (log of monthly searches: 1,000 scores 0 and 100 million scores 100), plus momentum 35% (twelve-month growth: −50% scores 0 and +200% scores 100), plus staying power 25% (the latest month as a share of the twelve-month peak). The scales are fixed, so a score does not change when other companies are added.
How are indexes built?
Each industry or sector is a basket of companies’ branded searches, rebased to 100 over the first three months. Equal-weighted averages every company’s own rebased series; attention-weighted sums the searches, so giants dominate. When the two disagree, the leader is the story.
Frequently asked questions
What is DawnTrend?
DawnTrend is a research site that shows where attention is moving across AI companies, startups and public markets, using how often people search for each company by name and, for many startups, how many people visit its website.
Where does the data come from?
Search demand comes from Google monthly search estimates. Website visits and traffic channels are third-party estimates. Company lists come from Wikipedia (S&P 500, Nasdaq-100, Fortune 100), with official websites from Wikidata, and from the Y Combinator directory.
How fresh is it?
Google publishes search estimates monthly. DawnTrend reuses any query it has answered in the last seven days, so most pages show the latest collection. The months covered and the collection date are printed on every page.
What is the Dawn Index?
A monthly 0 to 100 attention score for every brand we track: public companies, Y Combinator companies and AI products. It weighs search demand (40%), purchase intent (15%) and site visits (45%), and links every product to the company that makes it, so Claude counts toward Anthropic. Signals we have not measured are left out and the score is pulled toward 50 rather than guessed.
What is the difference between the Dawn Index and the Search Score?
The Search Score describes one search term from its own twelve months of demand. The Dawn Index describes a brand: it adds up the brand and its products, sorts those searches by intent, and adds site visits and loyalty.
What is the Search Score?
A number from 0 to 100 that blends three things: scale (40%, how many people search for the name), momentum (35%, twelve-month growth) and staying power (25%, how close the latest month is to the peak). The scales are fixed, so a company’s score does not change when other companies are added.
Is a rising trend a reason to buy a stock or invest?
No. DawnTrend measures attention, not revenue, users or profit, and is not investment advice. Use it to decide where to look, then check the company’s fundamentals.
Why does a company show no data, or exactly 0%?
Very small brands fall below Google’s reporting threshold, and volumes are reported in buckets, so a real change can stay inside one bucket. Companies with too little search volume are excluded from group statistics and are not listed in search engines.
Do I need an account?
No. Browsing, company pages and the main charts are free. A free account adds a daily search allowance for new queries, bulk research of up to 100 terms, website lookups, related searches, the full Markets analysis, CSV export and an API key.
Can I get the data through an API?
Yes. A free account includes an API key. See the API documentation for endpoints, limits and examples. Answers already in the cache are free and do not use your allowance.