competitors

/competitors builds a competitive landscape you can act on: who’s out there, how mature they are, where the gaps sit — then keeps it current as the field moves. Competitors are treated as validation, not threat: an empty landscape is usually a warning sign, and the gaps are the point.

​

When to reach for it

• “Who else is doing this?” — first-time discovery
• “I know a few competitors — flesh them out and find more”
• “Dig deeper on X — I think they’re a real threat”
• “What do users actually think of X?” — review mining
• “What changed since we last looked?” — the watch pass
• “This one pivoted away, archive it”

​

Discovery: mapping the landscape

The first-time flow starts with a short brief — do you already know competitors, what to focus on (pricing, integrations, segments), anything to exclude — asked one question at a time. Then it researches in two deliberate phases:

1. Scout — a fast pass for the top 2–3 direct and 2–3 indirect players, enough to confirm the direction is right. While research runs in the background, the conversation continues — typically about what makes your approach different, captured before external findings can anchor your thinking.
2. Expand (optional) — after you’ve confirmed the scout results, a deeper search for up to five more. Explicitly flagged as more time and tokens; you can always defer.

You review one-line summaries, confirm or correct, and the skill writes individual competitor profiles plus a landscape map. Each competitor is classified direct (same core problem, same audience) or indirect (adjacent tool or partial substitute), and by maturity — incumbent, scaleup, or startup — inferred from funding, age, and market presence.

​

User feedback: what real users say

Offered as an opt-in pass after discovery, or ad-hoc for a single competitor. It mines review sites (G2, Capterra, Trustpilot), app stores, Reddit, and niche communities for recurring themes — one-off rants don’t make the cut — and writes a ## What Users Say section into each profile:

• What Users Love — table stakes you’ll be measured against
• Complaints — openings for differentiation
• Unmet Needs — gaps users wish existed; interview and positioning gold

It’s heavier than discovery (more sources per competitor), which is why it’s opt-in.

​

Watch: keeping the landscape current

Competitors ship features, change pricing, pivot, and quietly die. When you ask to re-check the landscape, the watch pass researches what changed for each competitor since its last check — plus a scan for new entrants — and applies everything in one batch:

• Changed profiles refreshed in place, with a dated ## Change Log line
• Dead competitors archived (with the reason), new entrants added as full profiles
• The landscape map re-synced, and a dated entry (Changed / New / Gone / No change) prepended to the rolling digest in competitor-watch.md

The plugin runs locally with no background jobs, so a watch pass happens when you ask for it — a natural habit is running one before any big decision or investor conversation.

​

What it writes

Frontmatter fields (type, maturity, status, last_checked) are Obsidian/Dataview-queryable.

​

Good to know

• Research runs on the live web with confidence flags; single-source claims are marked. Early-stage competitors often have thin public footprints — that absence is reported as signal, not padded over.
• Works best after your idea is defined in core.md — research quality follows project-definition quality. If the project isn’t initialized, it will point you to /whats-next first.
• ## What Users Say and ## Change Log are machine-maintained — hand-edits there get overwritten on the next pass. Your space is ## Notes.
• Archiving is reversible; removing a competitor during review actually deletes the file.
• If your existing profiles are thin stubs, the skill notices and offers to fill them before doing anything else.

​

Related