Acme Metrics: Positioning Diagnosis and 6-Week Test Plan
I haven't seen your site copy, competitor pages, or funnel data, so the diagnosis below is a hypothesis built from your description. Week 1 is designed to confirm or kill it.
Diagnosis
This is a positioning problem, not a product or awareness problem. You match two triggers in the skill's decision tree: awareness is strong but conversion is weak, and your messaging sounds like three competitors. Buyers know who you are but can't tell why to pick you, so they stall or default to whichever vendor they already shortlisted.
Likely contributing causes:
- Commoditized claims. "Usage analytics for finance teams" probably leads with table-stakes language such as "insights," "visibility," "real-time" and "easy." If all four vendors say these, none of them differentiates.
- Category-level description instead of an outcome. You describe what the product is (analytics) rather than the decision it enables or the pain it removes.
- Possibly the wrong persona. Finance teams include the FP&A lead, the CFO, RevOps and the finance-systems owner. They care about different things, and one message for all of them lands with none.
- No claim you can defend. You need to find out whether you have a structural advantage, such as data model, integrations, billing-system depth, audit-readiness or pricing model. Without one, any claim is a feature that competitors can copy.
Step 1: Find the position (Week 1 work)
- Competitor audit. Collect headlines and subheads from your 3 direct competitors, plus 2–4 adjacent ones. Mark the claims they share. Those claims are off-limits as differentiators.
- Customer interviews. Talk to 8–10 of your best customers and 3–5 recent losses. Ask why they chose you, what they'd say if a peer asked what you do, and what they almost bought instead.
- Structural advantage check. Use the skill's defensibility hierarchy. A structural advantage beats category ownership, which beats a feature. A position that a competitor could copy in one sprint isn't worth testing.
Pick two candidate angles, not more. The skill warns against testing many at once. Draft the Layer 2 positioning statement for each: who, what problem, how you're different, and proof.
| Angle type | Illustrative example (replace with what your interviews support) |
|---|---|
| Outcome/persona | "Usage-to-revenue reconciliation for FP&A, so month-end close doesn't wait on data" |
| Structural | "Audit-ready usage data your auditors will accept" (only if the product truly supports it) |
Also check wording. The skill's key lesson is that one word can change deal progression ("autonomous" became "teammate"). Look for scary or vague words in your copy, such as "AI-powered," "automated," or "replaces manual work." Finance buyers tend to react to loss of control and audit risk, so "you stay in control" framing is worth testing.
6-Week Plan (Crawl → Walk, with Run as a decision gate)
The skill's methodology is Crawl (1–2 weeks), Walk (2–3 weeks), then Run. Run is the rebrand itself, so I've kept it outside these six weeks. Don't run the phases in parallel.
Weeks 1–2: Crawl (validate)
| Week | Actions |
|---|---|
| 1 | Competitor audit. 10–15 customer and loss interviews. Choose 2 angles (A, B) and write a positioning statement and headline/subhead for each. Pull baseline metrics for current messaging: homepage CTA click-through, outbound reply rate, demo-to-trial rate. |
| 2 | Launch homepage A/B/C: current vs. angle A vs. angle B. Launch two outbound sequences (about 100+ prospects per variant, same ICP, same sender profile). Run the "would you still recognize us?" check with 5–8 existing customers. |
Go/No-Go (end of week 2): At least one variant beats the incumbent by 20% or more on outbound reply rate. Existing customers don't say "that's not what you do." If results are tied or the test is too short, extend it a week or try a different angle. Don't force a winner.
Sample size caveat: 100 prospects per variant is small. A 20% relative lift on a low base reply rate (say 3% to 3.6%) is statistical noise. Either use larger sends or treat the result as directional and confirm it with qualitative reply content and booked meetings. Website CTR is also only moderate-strength evidence.
Weeks 3–5: Walk (align, with no public rebrand)
| Week | Actions |
|---|---|
| 3 | Rewrite homepage and key pages for the winning angle. Draft updated pitch deck, call script, and email templates. Add 2–3 use-case examples. |
| 4 | Split the sales team. Half use the new script and half use the old one, then compare demo-to-trial conversion. Train CS on the narrative so existing customers aren't confused. |
| 5 | Collect AE feedback ("is this easier to use?"). Segment who responds by persona and company size. Check that nothing in the docs contradicts the new story. |
Go/No-Go (end of week 5): Sales says the new messaging is easier to use. Engagement and CTR beat the incumbent. No notable customer confusion.
Week 6: Decision gate (Run readiness)
- Review pipeline volume, demo-to-trial, and early win-rate data by positioning angle.
- Decide: proceed to Run (dedicated landing pages per use case, full material update, announcement), iterate with a new angle, or keep current positioning and fix something else.
- Win rate and retention need a longer window than six weeks. Treat week 6 as a decision on leading indicators, and keep tracking win rate and CAC during Run.
Pitfalls to avoid
- Deciding by internal consensus. Your team isn't the buyer.
- Claiming you're "better" at what everyone already does.
- Positioning on a feature a competitor can ship in a quarter.
- Running one message for every finance persona.
- Waiting for the product to catch up. The skill says product should follow positioning, but don't claim anything you can't deliver today.
What I need from you
To turn this into something concrete, send me any of these:
- Your current homepage headline and subhead, plus the competitors' headlines.
- Current funnel numbers: CTR, reply rate, demo-to-trial.
- Who your best customers are (role, company size) and why they bought.
With those I can draft the two candidate positioning statements and the A/B copy.
Real run in an isolated folder with only this skill installed. Only the Skill and Read tools were enabled, so nothing was fetched from the web and no file was written. The example is fictional. The model answered from SKILL.md alone and said it had no real site or funnel data.
できること
弱いポジショニングの診断と、安全な変更の計画をモデルに支援させます。競合のメッセージ監査、防御性の評価(構造的な優位性、市場での地位、製品機能のどれか)、3 層のポジショニング構造(市場背景、ポジショニングステートメント、ナラティブ)、見出しとサブ見出しの型、言葉選びのテストを扱います。1 語の変更で法人案件の進みが変わった実例も含みます。
動き方
- モデルが製品、買い手、現在のメッセージを確認します。
- 自社の表現が競合とどれだけ似ているか、真似されにくい強みは何かを監査します。
- 「Crawl-Walk-Run」の段階導入を提案します。A/B テストで検証(1〜2 週間)、製品と営業の足並みをそろえる(2〜3 週間)、拡大の順で、各段階に進めるかどうかの判断基準があります。
向いている場面
他社と同じに聞こえるメッセージ、認知は高いのに転換が弱い、再ポジショニングや新製品の発売。
指示のみのパッケージです。スクリプトはなく、ネットワーク接続、ファイルの書き込み、アカウントは不要です。例や閾値(たとえば 20% の改善を進行の目安とするなど)は作者自身の経験に基づく目安で、普遍的なベンチマークではありません。自社の買い手のデータで検証してください。モデルが作成した自社製品の表現は、公開前にご自身で確認してください。