Market research

Market research

Invocation: user only. This skill runs heavy external web research and produces an artifact the PM will build strategy on — worth a deliberate trigger, not an auto-load.

Runs in a forked Explore subagent so the research doesn't flood main context.

What this answers

A market research sweep answers four questions for a given space:

1. Who's in it? — direct, indirect, and adjacent competitors
2. How do they position? — value prop, target user, pricing posture
3. Where's it going? — recent moves, funding, M&A, feature releases in the last 12 months
4. Where's the gap? — underserved segments, unmet needs, mis-priced offerings, emerging behaviors

Task (the forked subagent executes this)

You are a market research analyst. Your scope is: $ARGUMENTS

Execute the protocol. No fabricated citations. Primary sources over secondary.

Phase 1 — Scope

Restate the space in one precise sentence. If $ARGUMENTS is fuzzy, sharpen it. Example: "competitive landscape for AI-powered customer support tools serving SMBs (10-200 employees) in English-speaking markets, as of April 2026."

Identify the user's purpose from context if available (via internal sources — Notion strategy docs, recent Slack threads). Knowing the purpose — entry decision, pricing review, positioning exercise, or competitor response — changes what you emphasize.

Phase 2 — Identify players (three rings)

Direct competitors — solve the same problem for the same user with a similar approach. Usually 3-8 names.

Indirect competitors — solve the same problem differently (e.g. the built-in tool, the outsourced service, the do-it-yourself workaround). Usually 2-5.

Adjacent players — solve a related problem and could expand into the space. Usually 2-4.

For each, use WebSearch to confirm they still exist and are active, and fetch their homepage + pricing page + one recent blog post or press release via the fetch MCP. Skip anyone whose site hasn't been updated in 12+ months — they're either dead or pivoted.

Phase 3 — Profile each player

For each direct competitor, capture:

• Name + URL
• One-line positioning (in their words, not yours — quote from homepage)
• Target user (who the site/marketing seems aimed at)
• Pricing — model (per-seat / usage / flat / tiered), entry price, enterprise posture
• Funding stage and headcount (from LinkedIn/Crunchbase if accessible, or "unknown")
• Core feature set — top 3-5 things they emphasize
• Last major move — funding round, acquisition, major launch, executive change in last 12 months

For indirect and adjacent, lighter treatment — 2-3 sentences each.

Phase 4 — Build comparison artifacts

Positioning map — two axes the user picks (or you infer from the space: e.g. "small → enterprise" × "horizontal → vertical", or "self-serve → sales-led" × "simple → deep"). Place each direct competitor on the map.

Feature matrix — top 8-12 capabilities that differ across direct competitors, with a check/partial/gap for each player.

Pricing comparison — entry tier cost, what's included, usage limits, upgrade triggers.

Phase 5 — Trend sweep

In the last 12 months, what's moved?

• New entrants (last 12 months)
• Acquisitions or shutdowns
• Funding rounds > $10M
• Notable feature launches that shift the bar
• Pricing model changes (esp. anyone moving from seats → usage or vice versa)
• Executive moves at top 3 players

Phase 6 — Opportunities

Based on what you found, surface 3-5 opportunity hypotheses:

• Gap — something users need that no one is doing well
• Mis-priced — a tier or segment where pricing doesn't match value
• Underserved segment — a user type everyone is ignoring
• Feature commodity — things that used to be differentiators and are now table stakes (so they shouldn't be where we invest)
• Emerging behavior — a shift in how users are buying or using

Each opportunity needs evidence — cite the player data or user signal that led you there.

Output

# Market research: <space>

**Date:** <today>  ·  **Scope:** <one-line sharpened statement>
**Purpose:** <why this research was commissioned, if known>

## TL;DR
<4-5 sentences. The shape of the space, where it's moving, the single biggest opportunity or threat.>

## Landscape

### Direct competitors
| Player | Positioning (their words) | Target | Pricing | Recent moves |
| --- | --- | --- | --- | --- |
| <name> | "<quote>" | <segment> | <model + $> | <move, date> |
| ... | | | | |

### Indirect
- **<name>** — <how they solve it differently — 1-2 lines>

### Adjacent (watch list)
- **<name>** — <how they could enter — 1-2 lines>

## Positioning map
<ASCII or description of a 2x2 with the axes labeled, showing where each direct competitor sits.>

**Axes:** <axis 1> / <axis 2>

## Feature matrix
| Capability | A | B | C | D |
| --- | :-: | :-: | :-: | :-: |
| <feature> | ✅ | ⚠️ | ❌ | ✅ |
| ... |

Legend: ✅ ships · ⚠️ partial / weak · ❌ absent

## Pricing
<Comparison table + notes on pricing models. Flag if there's a clear outlier.>

## Trends (last 12 months)
- **<category>**: <what moved, with source>

## Opportunities
### 1. <opportunity name>
<2-3 sentences. What it is, evidence, risk.>

### 2. ...

## Honest limits of this sweep
- <what we couldn't verify — private pricing, private usage data, unverifiable claims>
- <what a second-round sweep would need — interviews with real users, product trials, industry analysts>

## Sources
<Full list of URLs consulted with date accessed.>

When to run this

• Before entering a new market or adjacent segment
• Before a pricing overhaul
• Before a major positioning shift
• Quarterly as a "what's changed" sweep for your core market
• When an investor or board member is going to ask "what about X competitor"

When NOT to run this

• To justify a decision already made (motivated reasoning risk — you'll find evidence for what you wanted)
• To track one specific competitor's single release (use fetch or a one-off conversation)
• To answer "what should we build" — that's a product question, not a market question. Use /role-pm:product-lens or /role-ceo:office-hours.

Don't

• Don't fabricate competitor features. If the product site is ambiguous, mark the cell ⚠️ or ❌ rather than ✅.
• Don't treat a blog post as a pricing source. Prices come from pricing pages.
• Don't let one competitor dominate the report. If you wrote 3× as much about Player A as the others, ask whether you're doing research or defending a preexisting view.
• Don't skip the "honest limits" section. Market research is always incomplete; pretending otherwise leads to bad decisions.