Competitor dossier

Click a competitor’s name in the list and you land in their dossier: one page holding everything Vector knows about that company. It’s where you research them, collect their public sources, and decide what gets watched from now on.

​

See how much you know

A dossier is built up over time rather than filled in at once. The counter above the tabs — Dossier · 0/9 filled on a new one — shows how many tabs have something in them: a resource added, monitoring switched on, a help center or changelog source set up, at least one scan or file. A tab with a check mark has content. The List control next to the name moves the competitor between your short and long list — promote them the moment the research justifies it.

​

Learn how their product is built

On 🔬 Deep research, click ✦ Run deep research. You get a structured teardown of the competitor’s product in 13 sections: architecture, AI pipeline, privacy, plans, invariants, trade-offs. It explains how the product works inside rather than how it’s marketed, and runs several hundred lines long. It needs an AI key in Settings and your own product described on My product, because the research is framed against yours. Each run is kept as a version with its date and the model that wrote it; ↻ Re-run research adds a new one. To steer the teardown toward what matters in your market, click edit skill and change the instructions behind it.

​

Collect their public sources

Resources is the competitor’s public footprint, grouped by kind: product signals such as help center, changelog, blog, release notes, pricing and feedback portal; social accounts; app store listings; and review sites like G2, Capterra or Trustpilot.

• Let AI find them. ✦ AI scan sources searches the web for the competitor’s portals, profiles and listings and adds the ones it finds, without duplicates.
• Add one yourself. Click a quick-add button for the kind of source. It adds a row with an example address as a hint — paste the real one.

The Description field at the top holds your own one-line read on the company.

​

Watch them for changes

On 🔔 Monitoring, turn the switch on. Vector starts watching every source it can, and you uncheck the ones you don’t care about. Social accounts, app stores and review sites stay links only — they can’t be watched.

• Set how often. Check sets the schedule: Daily, Weekly, Every 2 weeks or Monthly. Weekly is the default.
• Check right now. ✦ Check now runs a check without waiting for the schedule.
• Pause. Turn the switch off. Your choice of sources is kept for when you turn it back on.

The first check of a source is a baseline: it records what’s there today and never raises an alert. After that, a change means items that weren’t there at the previous check, each with a link you can open and verify yourself. Every check keeps a full snapshot of the source, so history & snapshots on a source shows what it looked like each time. To manage monitoring for all your competitors in one place, see Put a competitor under monitoring.

​

Track their prices

On 💲 Pricing, click ✦ Scan pricing. Vector fetches the competitor’s pricing from the web and saves it. Each scan that finds different pricing is kept as a new version, so the tab builds up a change log of what moved, and a price rise becomes a fact with a date on it. A scan that finds nothing new is only noted under Activity. ✦ Re-scan pricing takes a fresh capture. To have pricing checked on a schedule, watch the competitor’s pricing resource under monitoring.

​

Archive their help center and changelog

📚 Help/Docs and 📝 Change log each save a whole section of the competitor’s site as a copy you can read in Vector. Click Add Help/Docs source (or Add Changelog source), enter the root URL, set Max pages — 500 by default — and click ✦ Crawl now. Crawls need a Firecrawl key on the Web scrapers tab of Settings.

• Trust mode decides how carefully pages are checked. Lightweight is cheap: a page is checked by its hash and a plausibility test. Full corroboration confirms pages against a second source and costs more.
• Auto-index after crawl prepares new pages for matching to your features as soon as they land. It needs an embedding provider in Settings.

Competitor content archive covers reading the saved pages and tracking what changed.

​

Find out which features they have

Two tabs answer this from opposite directions:

• ✦ Feature scan — what the competitor says they have. AI reads their site and lists their features as they present them. Each feature it finds arrives as a candidate for you to review: attach it to a feature you already track, or create a new one.
• 🧩 Features — how they cover a shared list of features. ✦ Evaluate vs comparison checks this competitor against it and fills their column in the comparison. Every evaluation is kept as a version, so ✦ Re-evaluate later shows which statuses changed.

New model: this competitor → opens the competitor’s comparison page, where both answers sit in one table. To run either for several competitors at once, select them in the competitors list and pick ✦ Feature scan or 🧩 Features.

​

Add what you know yourself

Some of what you know isn’t on the public web: battle cards, analyst reports, sales-call notes, trial screenshots, a recording of their demo. Put it on 📎 Files. Text files are read by AI and cited as an internal source in research; images, audio, video and archives are stored for you to open. See Files in a dossier.

​

Where to go next