DifferentiationSection — v2 rewrite strategy

Dual-audience analysis + A/B variant design

Dispatch: marketing-strategist, 2026-06-04 PR: #3232 (feat/getraxx-differentiation-section-3231, fix-forward) Supersedes: v1 section in same PR (copy reviewed by operator 2026-06-04) BLR cross-reference: docs/legal/research/getraxx-differentiation-copy-review-2026-06-04.md (file expected; not yet written at time of this dispatch — copy below is written with BLR redlines in mind but not yet incorporating them. When BLR lands, reconcile any forward-looking or performance-implication language they flag with the variant copy.)

The operator's diagnosis

Current v1 copy is comparative/defensive: "here is why we are different from other platforms." For an experienced trader with an established platform, comparative copy reads as a pitch. A pitch raises the skepticism wall — "convince me" is the opposite of "sign me up." For a new trader, it reads as jargon (backtest, paper-trade, BYOB) without anchoring what any of those things actually mean for them in 30 seconds.

The strategic reframe: lower the activation energy of the free signup, not the bar for understanding every feature. The section's job is not to close the sale — Pillars already stated the product. This section's job is to remove the excuse that delays clicking.

Dual-audience map

Audience A: The skilled trader (primary)

Who: Has thinkorswim, tastytrade, IBKR, Robinhood, or Schwab. Trades options or multi-leg structures at least occasionally. Has a process — however informal. May already paper-trade sometimes. Is not looking for a new platform. Is skeptical of pitches.

Their internal monologue reading v1: "OK so you do backtests and paper trading. So does TOS. What am I actually supposed to do with this page."

What lowers their wall: Seeing the product do something useful in the time it takes to read a paragraph. Not features — output. "Here is what you would actually see." The pitch framing must collapse; the section should feel like a peer showing them something, not a vendor selling them something.

Conversion goal: Remove the activation energy barrier to a free signup. They do not need to be sold; they need to be given a reason that a free sign-up is worth 2 minutes. The ask is not "switch platforms" — it is "run one backtest on a structure you already use."

Audience B: The curious / new trader (secondary)

Who: Knows what options are, may have traded a few calls or puts, is interested in being more systematic. Does not yet have a fortified position. Reads "backtest" and understands the word but not why it would be useful for them specifically.

Their internal monologue reading v1: "This looks serious but I'm not sure if it's for me or for somebody who trades full-time."

What lowers their wall: Concrete time-to-value + explicit "you don't need to be an expert to start." The section should feel achievable. A new trader does not need to understand the full product — they need to believe that the first session will be useful and not overwhelming.

Conversion goal: Confidence that Raxx is real, structured, and designed not to blow up their account. Paper-first gating is actually a strong signal for this audience — the section should name it in terms they can connect to ("you test in paper before anything live moves").

Why v1 failed both audiences

1. "Not another trading tool." — Opens with a comparative negation. This is the classic "we are not like the others" framing that tells the reader they are about to receive a pitch. Kills attention for Audience A immediately.

2. "Most platforms track what you made. Raxx tracks whether you stayed inside the structure." — This is a differentiation claim that requires the reader to already understand what "staying inside the structure" means. It is good internal language and works in Pillars (which built up to it), but here it lands with no setup.

3. D2 and D3 restate Pillars 01 and 02 verbatim in different words. A visitor who scrolled past Pillars already read these. Duplication signals the page has run out of new things to say.

4. D3 "Promote a structure when it earns it" shares the phrase "qualifying round" with Pillar 02, which the operator flagged explicitly.

5. CTA "Ready to test your structure?" assumes they understand what "your structure" means. Audience B does not yet. Audience A knows but is not persuaded by this framing.

Variant design: three distinct hypotheses

The three variants are designed so that each tests a different behavioral hypothesis. This is not three word-tweaks — it is three different conversion mechanisms. A/B results will tell you which mechanism is actually working, not which headline sounds better.

Variant A — "One session. One structure. See for yourself."

Frame: Collapse time-to-value. The section becomes an invitation to a single session, not a feature list. The implicit message: you will know within one backtest whether this is worth anything to you.

What hypothesis it tests: Does reducing perceived commitment (one session, not platform migration) increase click-through from skeptical experienced traders?

Headline: One session. See if it fits.

Sub-headline: You do not need to migrate anything. Run a structure you already use. See what it did in paper. Sign up for free and decide after.

Body content (3 short statements, no card grid — just stacked emphasis lines):

• "Bring the structure you already use. We do not invent one for you."
• "A backtest runs in under a minute. You see the result before any money moves."
• "Paper before live. That is a gate, not a feature you skip."

CTA: Run a backtest in 60 seconds — it is free.

Visual treatment: Minimal. Ink background, three stacked copy lines with a single horizontal moss rule between them. No cards. The contrast with Pillars' numbered-card density signals a change of register. This is a conversation, not a feature list.

Copy notes: - No jargon: "backtest" is explained inline ("you see the result before any money moves") - "Gate, not a feature you skip" lands for Audience A who has bypassed paper modes elsewhere. For Audience B it is reassuring ("the platform makes me do this safely"). - CTA is behaviorally specific — not "join the waitlist" but the specific action. (Still routes to #waitlist since live app is not yet accessible.)

Variant B — "It works on what you already have."

Frame: Zero-migration. The section's entire job is to eliminate the "but I'd have to move my account / learn a new platform" objection before the reader articulates it. Raxx is additive to the setup they have, not a replacement.

What hypothesis it tests: Does emphasizing zero-migration / additive-not-replacement lower the activation barrier specifically for Audience A who has sunk costs in an existing platform?

Headline: Keep your broker. Add structure.

Sub-headline: Raxx connects to the account you already have. No migration. No new account. One structure layer on top of what you are already doing.

Body content (two columns, three points per column would be too dense — use four short items in a 2x2 grid):

CTA: Try it on a structure you already use. Free to start.

Visual treatment: 2x2 card grid, ink-2 cards, moss border accent. Mirrors the existing DiffCard layout so it does not visually restate anything. Short, punchy — no paragraph prose in cards.

Copy notes: - "Keep your broker" is the clearest possible signal to Audience A. The entire pitch anxiety dissolves in three words. - "Add structure" frames Raxx as a layer, not a competitor. Psychologically this is the "yes, and" move — you do not have to choose. - No broker names per feedback_no_backend_branding. "Account you already have" is sufficiently concrete without naming.

Variant C — "What you actually see in a session."

Frame: Show the output, not the feature. The section becomes a narrated product moment — a single concrete scenario that makes the product tangible in 20 seconds. Less UI-interactive than the proposed "Variant A — show don't pitch" frame (no screenshots yet), but uses copy structure to simulate the experience.

What hypothesis it tests: Does making the product concrete through a specific scenario narrative (not a feature list) reduce bounce from Audience B who does not understand what they would actually do in the app?

Headline: Here is what a first session looks like.

Sub-headline: No guided tour. No demo mode. You put in the parameters for a structure you want to test. You get the historical result.

Body content (a three-step narrative, not a feature list):

Step 1 — "You define the structure." Entry condition, premium target, exit rule. Takes about two minutes.

Step 2 — "Raxx runs it against real history." You see what this structure returned across real market bars. Not a simulation — the actual recorded result.

Step 3 — "You decide." If the result looks worth testing in paper, you paper it. When it has earned promotion, you run it live.

CTA: Start with step 1. It is free.

Visual treatment: Three numbered steps, large mono numerals (matching Pillars' 01/02/03 numbering language but in a horizontal row or horizontal-scrolling single column). Ink-3 background to differentiate from Pillars' snow background and the ink-2 cards. The numbered-step format signals sequence and progression — reduces the "where do I even start" anxiety for Audience B.

Copy notes: - "Not a simulation — the actual recorded result" directly addresses the skepticism that backtests are fake. Retrospective framing is legally cleaner than "here is what you will earn." - Audience B gets a literal answer to "what would I actually do?" without needing to decode product language. - Audience A gets a peer-level description of the tool's output without a pitch.

What each variant distinctly tests

Recommended starting split: Run A and C first (A tests the experienced trader, C tests the new trader). B is a strong test but is most meaningful once you have baseline data on whether the blocking objection is "I don't understand what I'd do" (C wins) or "this sounds useful but I already have a setup" (B wins).

Recommended success metrics (Clarity engagement events)

All events should be filed as a product-manager card for wiring up. The variant identifier should be passed as a property on every event.

Primary metric: diff_cta_click rate by variant (denominator: diff_section_view). Secondary metric: Waitlist conversion rate conditioned on diff_section_view (via session replay stitching in Clarity). Guardrail metric: Scroll depth past section vs. bounce — a variant that improves CTA click but kills scroll-to-pricing is net negative.

Recommended starting variant

Variant A for the first 30 days, on the basis that: 1. Audience A (experienced trader) is the higher-value conversion target for waitlist. 2. Commitment reduction is the most directly testable and most immediately reversible copy lever. 3. The minimal visual treatment (no cards) cleanly differentiates the section from Pillars and from the pricing teaser, reducing redundancy risk.

If variant A does not meaningfully outperform the v1 section on diff_cta_click in 30 days, rotate to C.

What this implies for BLR review

When docs/legal/research/getraxx-differentiation-copy-review-2026-06-04.md lands, reconcile the following specific copy lines in each variant:

• Variant A: "a backtest runs in under a minute" — confirm BLR is comfortable with specific time claim (may need to be softened to "quickly" if they flag it as a performance-implication).
• Variant C: "Not a simulation — the actual recorded result" — confirm BLR is comfortable with this contrast claim. It is descriptive of the methodology, not a return claim, but BLR should confirm.
• All variants: "Paper before live. That is a gate, not a feature you skip" — confirm BLR is comfortable with the strong assertion about the paper requirement (it is architecturally true but could be read as a warranty claim if paper outcomes are treated as predictive of live outcomes).

Implementation notes for feature-developer

• DifferentiationSection.jsx accepts a variant prop ('A' | 'B' | 'C'). Default is 'A' per this strategy recommendation.
• Variant selection via REACT_APP_FLAG_GETRAXX_DIFFERENTIATION_VARIANT env var. The component reads the env var if no variant prop is passed.
• Outer flag gate (REACT_APP_FLAG_GETRAXX_DIFFERENTIATION_SECTION) still controls whether the section renders at all. Variant selection is inside that gate.
• Analytics event wiring is a separate card for product-manager — the components expose data-variant and data-section="differentiation" attributes so Clarity auto-capture can pick them up without code instrumentation. Explicit PostHog events are a stretch goal; file a follow-up card.

Handoff note for product-manager

Cards to file from this strategy doc:

1. Wire REACT_APP_FLAG_GETRAXX_DIFFERENTIATION_VARIANT env var to Cloudflare Pages dashboard (or set to 'A' as default).
2. File Clarity event tracking card: diff_section_view, diff_section_scroll_complete, diff_cta_click, diff_waitlist_conversion — with variant property.
3. When BLR redline doc lands: review the three flagged copy lines above and determine if any require copy edits before flag flip.
4. After 30 days of variant A data: review diff_cta_click rate; decide whether to rotate to variant B or C based on results.

Shape 1 rewrite — 2026-06-05

Operator direction: Staged sentiment roadmap locked. Shape 1 = Personal Sentiment Journal ships in v1. Shape 2 (aggregated per-ticker) and Shape 3 (structure leaderboard) are roadmap items only — not v1 launch claims.

Strategic rationale for the rewrite: The prior variants (activation-energy / zero-migration frame) addressed the "why sign up" question. The Shape 1 rewrite addresses a stronger question: "why is this different from anything I have seen?" Almost nobody does this at retail. Bloomberg has market sentiment data. Public.com has community feed. Neither shows you what happened to your own structure when you were in a specific emotional state. That is the differentiator. It extends Raxx's structure-gap thesis from trade execution to emotional state.

Operator framing translated: "Backtest your strategy, and understand other's emotions to make the right trades" — the operator's spirit. Legal translation for v1: instead of "understand others' emotions" (requires external market-sentiment data, data-vendor cost, and higher legal exposure), it is "understand your own emotions in real-time, then see what they did to your trades historically." Shape 2 covers the "others" dimension once it ships.

Shape 1 legal constraints (BLR parallel check in flight)

The BLR Shape 1 compliance check is running in parallel. Assume the following constraints will be confirmed:

1. User always asserts first. Raxx never auto-classifies sentiment or emotional state. All copy uses first-person user-initiated language: "you mark", "you label", "you record". Raxx stores and surfaces what the user asserted; it does not infer.

2. Backtest outputs are auth-gated. No backtest result filtered by emotional state appears on the marketing page. All copy describes the capability without showing an example result. The user sees their own result only after logging in.

3. Retrospective framing only. All language describes what happened in past trades: "what your structure looked like", "what it returned", "what it did". No forward- looking inference ("if you label Disciplined, your trades will perform better").

4. No improvement claims. The copy surfaces a record; the user draws conclusions. "No 'improve' / 'better results' / 'guaranteed' / 'trade better'" language appears anywhere.

5. Emotional-state taxonomy is under design. Copy uses illustrative labels (Disciplined, Confident, Panicked). These are examples, not the final taxonomy the data-scientist is designing. Labels in copy should not be treated as product promises.

Variant A (Shape 1) — "Backtest yourself."

Retention hypothesis: Self-knowledge is rare. Experienced traders know they have an emotional weak spot; they have never had a tool that shows it to them on their own data. The question "what did your structure do when emotion got a vote?" makes them want to know the answer before they have thought about what it would cost to find out. New traders read it as: this is a thing I can learn about myself here, before it costs me anything.

Headline: Backtest yourself.

Sub-headline: You marked your emotional state before the trade. Raxx shows you what your structure looked like in those moments — on your own historical data.

Section kicker: Personal Sentiment Journal

Three stacked emphasis lines: 1. Before the trade, you commit a sentiment label. Bullish, Bearish, Neutral, High-Uncertainty. That is your pre-commit record. 2. After the trade closes, you record your emotional state. Disciplined, Confident, Panicked. That is your emotional record. 3. Your backtest is queryable across both. What did your structure actually look like when you marked Disciplined? When you marked Panicked?

Roadmap kicker (Shape 2, honestly framed): "Coming next: see what other Raxx users are marking alongside you on the same tickers. Your record stays yours."

CTA: Start your emotional record — free

What this does NOT say: It does not imply marking your emotion improves your trading. It does not show any results. It does not describe what the "right" emotional state is. The user sees their own data and draws their own conclusion.

Variant B (Shape 1) — "Your trades, your patterns."

Retention hypothesis: The differentiator is specific: you can filter your own backtest by emotional state. Most traders who have backtested before have never asked "what happened when I was Panicked?" because no tool kept that record. Surfacing the filter as the specific capability — not the concept — makes this concrete for both audiences. The four-card grid walks both audiences through what each of the two records is, then what they enable together.

Headline: Your trades, your patterns.

Sub-headline: See what your structure did when you were Confident. See what it did when you were Panicked. Two records from the same trade — yours to query.

Section kicker: Personal Sentiment Journal

Four cards (2x2 grid): 1. You mark sentiment before the trade. (Bullish / Bearish / Neutral / High-Uncertainty) 2. You record your state after it closes. (Disciplined / Confident / Panicked) 3. Your backtest becomes queryable. (filter by emotional state) 4. Your data. Nobody else's. (not market sentiment — your personal record)

CTA: Build your emotional record — free to start

What this does NOT say: It does not say one emotional state produces better results. It does not claim emotional awareness improves trading. It describes a record and a query.

Variant C (Shape 1) — "Two records. One trade."

Retention hypothesis: The structure-enforcement-of-emotion thesis is the cleanest single-sentence explanation of what is new. "The trade gets your structure. You get the record of whether emotion got a vote." This framing works for both audiences: experienced traders immediately understand the extension of Raxx's existing structure-gap thesis; new traders understand that something is being recorded before and after, and that the backtest knows about both. The three-step format (before / after / in the backtest) is concrete enough to work as an onboarding frame.

Headline: Two records. One trade.

Sub-headline: The trade gets your structure. You get the record of whether emotion got a vote — and what it looked like in your history when it did.

Section kicker: Personal Sentiment Journal

Three steps: 01. Before the trade, you commit. (sentiment label: Bullish / Bearish / Neutral / High-Uncertainty, attached to the trade permanently) 02. After it closes, you record. (emotional state: Disciplined / Confident / Panicked — your words, your assessment, stored with the trade) 03. In the backtest, you see the split. (query your historical results by emotional state — the answer is in your own data)

CTA: Start your record — free

What this does NOT say: No inference about what the split means. No implication that one state is better. The user interprets their own data.

Shape 2 roadmap claim framing (all variants)

Shape 2 (aggregated per-ticker sentiment from other Raxx users) ships in v2. The honest roadmap claim that can be used now:

"Coming next: see what other Raxx users are marking alongside you on the same tickers. Your record stays yours."

This is included in Variant A as a mono-styled kicker. Variants B and C do not include the roadmap kicker in v1 — the four-card and three-step formats are already complete. The kicker can be added to B and C once Shape 2 is closer to ship.

Do not frame Shape 2 as a current capability in any variant. The kicker must remain a "coming next" statement until the feature is live and tested.

BLR coordination notes for Shape 1 copy

When the BLR Shape 1 compliance check lands, reconcile the following specific lines:

• All variants: "Disciplined, Confident, Panicked" — these are illustrative labels. The data-scientist is designing the final taxonomy. If BLR flags these as product promises, soften to "your own words" or "the labels you choose" until taxonomy is locked.
• Variant A: "Your backtest is queryable across both" — confirm BLR is comfortable with "queryable" as a feature description on the marketing page. It is an accurate description of a retrospective query on user data; not a result claim.
• Variant C: "In the backtest, you see the split" — confirm BLR is comfortable. "Split" describes a query result on user's own historical data. No inference about what the split means. Should be clean, but verify.
• Shape 2 roadmap kicker: "Coming next: see what other Raxx users are marking" — confirm BLR is comfortable with this as an honest roadmap statement. It does not promise a ship date. It does not claim v1 has this capability.

Recommended starting variant for Shape 1

Variant A ("Backtest yourself.") remains recommended for the first 30 days.

Rationale: The headline is a question the experienced trader has never been able to answer. It creates curiosity before it requires understanding. "Backtest yourself" is two words that contain the entire thesis. For new traders, "your emotional state" is accessible without jargon — everyone understands they have emotional states; the novelty is that a trading tool keeps that record.

Rotate to C if A underperforms after 30 days. Variant C's "Two records. One trade." headline is the second-strongest single-sentence thesis statement and may convert better with Audience B (concrete, sequenced, immediately parseable).

Variant B tests the specific "your patterns" framing and is most useful after you have data from A or C establishing baseline interest in the sentiment feature.