Competitive Brief

Competitive Brief

A competitive brief turns scattered impressions about a rival into a decision. The point is never to list features side by side — it's to answer "where do we win, where do we lose, and what do we do about it?" This skill produces a strategy brief and a sales battlecard, both grounded in dated evidence and ending in a recommendation, not a shrug.

Grounded in: Obviously Awesome — April Dunford: frame competitors by the value customers actually weigh, then build the battlecard. Go deeper (The Product Channel): The Ultimate Guide to Product Strategy

Good competitive work is opinionated and honest. It names where the competitor is genuinely better, distinguishes what you know from what you suspect, and resists the urge to make your product look good on paper. A brief that flatters you is worse than useless — it loses deals quietly.

When to use this

• Informing product strategy or prioritization: deciding where to differentiate vs. reach parity.
• Building or refreshing a sales battlecard for the field team.
• Prepping board, investor, or leadership materials on the competitive landscape.
• A competitor just shipped, repriced, or repositioned and you need to assess the impact.
• Entering a new segment and need to map who already owns it.

Before you start (gather these)

• Competitor name(s) — one for a deep teardown, or a set for a landscape.
• The dimension(s) that matter — what decision is this feeding? (Pricing? A specific feature gap? Enterprise readiness? GTM motion?) A brief with no decision behind it sprawls.
• Your own positioning — your target buyer, your wedge, your current pricing. The brief is relative to you; without this, it's just a profile.
• Evidence sources. If web search is available, this skill can research pricing pages, docs, changelogs, G2/Capterra reviews, job postings, and press. If it's not available — or for anything non-public (win/loss notes, sales-call intel, contract terms) — ask the user to paste it. Always say which mode you're in.

Process

Delegate to the researcher subagent. Send each competitor to a separate researcher invocation. Read together, competitors blur into a feature grid; read independently, the structural differences stay visible. If subagents aren't available on this host, run the step yourself and say so — the output is still useful, just less independent.

1. Define the comparison frame. Pin down the decision and pick 5–9 dimensions that actually move it (e.g. core capability, integrations, pricing model, time-to-value, security/compliance, support, ecosystem). Drop vanity dimensions nobody buys on. State the frame before gathering anything.
2. Gather evidence (web + user-provided). Pull from the live web where available; otherwise work from pasted material. For each claim, capture the source and the date you observed it. Tag every data point as fact (verifiable: a published price, a documented limit) or inference (your read: "likely targeting mid-market because…"). Flag gaps explicitly rather than guessing.

• If the user's intel is secondhand recall (remembered from a call, "I think they…", no source pasted), treat the entire output as provisional: stamp the verification status accordingly and put a prominent ⚠️ PROVISIONAL — re-pull before use banner at the top of both the brief and the battlecard.
• In user-provided / recall mode, where nearly every competitor claim is inference, don't tag every line. State the default once up front — e.g. "Unless marked otherwise, all competitor claims below are inference from user recall, not verified." — and then tag only the exceptions (the few that are verified fact, or the few that are pure guesswork beyond even recall).

1. Map onto a matrix. Score us vs. each competitor across the dimensions. Use plain marks (✅ strong / ⚠️ partial / ❌ absent / — unknown), not invented 1–10 scores that imply false precision.
2. Identify gaps, parity, and differentiation. Sort every dimension into three buckets: we lose (their advantage), we're at parity (table stakes, neutralize cheaply), we win (our wedge, press it). Be ruthless about the first bucket.
3. Derive strategic implications. Translate the map into "so what." Where you lose: is it worth closing, or do you concede and out-position? Where you win: how do you make it un-ignorable? What's the threat trajectory — where are they heading next (read their changelog and job posts)?
4. Build the battlecard. Compress the brief into a field-ready card the sales team can use live: their pitch, your counter, traps to set, objection handling, proof points. If sales can't skim it in 30 seconds, it won't get used.

Output templates

Template 1 — Competitive Brief

# Competitive Brief: [Us] vs [Competitor(s)]
Prepared: [date] · Author: [operator's name/role, or the persona if none given — otherwise leave blank; don't invent one] · Evidence dated as of: [date]
Research mode: [live web search / user-provided only / mixed]

## Overview
- Who they are, stage, size, momentum (funding, headcount, growth signals) — [fact/inference]
- Why they're on our radar / what triggered this brief

## Positioning
- Their stated positioning (their words): "[…]"
- Their real target buyer (our read): [segment] — [fact/inference + why]
- Our positioning by contrast: […]

## Feature comparison matrix
| Capability / dimension | Us | [Competitor] | Gap (who wins, how big) |
|---|---|---|---|
| [Core capability]   | ✅ | ⚠️ | Us — meaningful |
| [Integrations]      | ⚠️ | ✅ | Them — closeable |
| [Pricing model]     | ✅ | ❌ | Us — structural |
| [Security/compliance] | — | ✅ | Them — unknown on our side |
(✅ strong · ⚠️ partial · ❌ absent · — unknown)

## Pricing & packaging
- Their model, tiers, and headline prices [date observed] — [fact]
- Hidden costs / gotchas (overages, seat minimums, add-ons)
- How a real deal compares vs. ours — compute a concrete per-deal dollar figure at a representative seat count, both sides (e.g. "at 40 seats: $X/yr us vs. $Y/yr them, before overages"); state the seat count and any assumptions

## Strengths / Weaknesses
**Their strengths:** […]
**Their weaknesses:** […]

## Where to differentiate (press our advantage)
- [Dimension] — why it's defensible, how to make it the deciding factor

## Where to reach parity (neutralize, don't lead)
- [Dimension] — minimum bar to take it off the table, rough cost/effort

## Threats
- Where they're heading next (from changelog / job posts / funding) — [inference]
- What would hurt us most if they shipped it

## Recommended moves
1. [Action] — owner, rough timing, expected effect
2. […]
3. […]

Template 2 — Sales Battlecard

# Battlecard: vs [Competitor]  ·  updated [date]
**Intel as-of:** [date] · **Verification:** [verified live / user-provided / unverified — re-pull before use]

**Their pitch (what reps will hear):** "[…]"

**Our counter (one breath):** "[…]"

**Landmines to set** (questions that expose their weak spots):
- "Ask them about [X]…" → they struggle because […]
- "Have them show you [Y] live…" → […]

**Objection handling**
| They'll say | We say |
|---|---|
| "[Competitor] is cheaper" | "[reframe to TCO / value…]" |
| "They have [feature]" | "[parity claim or out-position…]" |
| "[Competitor] is the market leader" | "[…]" |

**Proof points** (dated, specific, defensible):
- [Metric / customer / benchmark] — [date]
- […]

Quality bar

• [ ] Every claim is evidence-backed and dated — no "I think they cost more."
• [ ] Fact vs. inference is marked throughout; guesses are labeled as guesses.
• [ ] The frame ties to a real decision, not a generic profile.
• [ ] At least one dimension where the competitor genuinely wins is named honestly.
• [ ] Marketing copy is treated as a claim to verify, not as truth.
• [ ] Unknowns are flagged as unknowns (—), not papered over.
• [ ] It ends with a recommendation and owners — not just a comparison table.

Avoid (anti-patterns)

• A feature grid. It always concludes you're ahead, because you chose the rows. Compare on the value customers actually weigh.
• Built from their marketing site. Their site describes their ambition. Your lost deals describe their product — start there.
• No honest weakness. A brief with no place you lose is a brief nobody in sales will trust, and they'll go back to improvising.
• Analysis with no decision attached. If it doesn't change positioning, pricing, or the roadmap, it's a reading exercise. State the decision it informs up front.
• Treating "no decision" as not a competitor. Inertia wins more deals than any named rival and is systematically under-recorded.
• Stale by default. Competitive facts rot in weeks. Date it and set a review, or it will be confidently wrong.

Tips

• Date every data point. "Pricing as of [date]" is the difference between a brief that ages gracefully and one that gets someone fired in a deal.
• Competitors change fast. A brief is a snapshot, not a monument — note when it should be re-pulled, and re-verify before any high-stakes use (board deck, big deal).
• Never trust marketing copy as truth. "Enterprise-grade," "AI-powered," "real-time" are claims to test, not facts to repeat. Prefer docs, changelogs, pricing pages, and actual reviews over the homepage.
• Reviews are gold, with a grain of salt. G2/Capterra/Reddit reveal real weaknesses, but skew toward churned or angry users — weight by recency and volume.
• Read the changelog and the job board. What they ship and who they hire tells you where they're going better than any press release.
• Resist symmetry. You don't owe each competitor equal coverage. Go deep on the one that actually threatens the deal or the roadmap.
• One competitor, one teardown; many competitors, one landscape. Don't write five shallow profiles when the decision needs one sharp comparison.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: Sidsaladi9
• Source: Sidsaladi9/persona-os
• License: MIT
• Homepage: https://sidsaladi.substack.com

Install and usage instructions live in the source repository linked above.