Paris Rooftop Solar: Pressure Test

Rule: Every assumption in the TurboTax wizard is questioned below. If we can't answer the pressure test, it's a red flag. No optimism allowed.

1. Data Access Assumptions

We can find a syndic's managed buildings from their email domain.HIGH RISK

Pressure test: Syndics often use generic domains (@gmail.com, @orange.fr), especially smaller ones. Even with a @company.fr domain, France's public syndic registries (Registre du Commerce et des Sociétés, ANAH copropriété database) may not map email domains to building addresses in a machine‑readable way. What percentage of syndic emails are on recognizable domains?

If false: The "we found 3 buildings" opening is dead on arrival for 60%+ of users. They'll see "We couldn't find your portfolio" and bounce.

Backup plan: Ask for one building address in Step 1 (with BAN autocomplete) and expand from there by matching their name/company in the copropriété registry. Use the "found 3 buildings" as a wow‑moment only when we have high confidence.

IGN cadastre + 3D roof data is available via API for real‑time lookup.MEDIUM RISK

Pressure test: IGN Géoportail API exists, but coverage of 3D building data (BD Topo / Bâti3D) is partial. Access requires registration, API keys, and rate limits. Real‑time roof geometry extraction from cadastre is non‑trivial — does the API expose roof orientation, pitch, and usable surface? Or just 2D footprint?

If false: Our solar calculations are garbage. Bad data → bad savings estimates → angry customers who lose trust.

Backup plan: Use PVGIS for irradiation + standard roof assumptions (30° pitch, South orientation as default) + manual override. Calibrate against real installer quotes once we have data. Disclose confidence intervals: "Estimated savings: €8,200/year (±15%)."

We can detect ABF (historic zone) and PLU restrictions from geoportal layers.HIGH RISK

Pressure test: ABF zones are mapped, but the rules differ by building (e.g., visible vs. non‑visible roofs). PLU restrictions are per‑municipality and change frequently — some PLUs ban panels on certain streetscapes. Is there a single API that returns "solar panels allowed: yes/no/conditional"? Or do we need to parse PDF urban planning documents?

If false: We tell a syndic "no ABF required" and they later learn their building is in a protected zone. We look incompetent or worse, liable.

Backup plan:>/strong> Always flag "Conditional" unless the building is 100% outside historic zones. Link to the arrondissement's PLU PDF. Add a disclaimer: "Regulatory status is an estimate — confirm with your mairie before proceeding."

Solar savings estimates are accurate enough to base AG motions on.HIGH RISK

Pressure test: Real solar production depends on: actual shading (trees, neighboring buildings), roof condition (age, structural integrity), inverter choice, panel degradation, consumption patterns of the copropriété. PVGIS gives theoretical max. Real output can be 30–50% lower. An AG motion with optimistic savings that fails to materialize = lawsuit risk.

If false: Copropriétaires vote yes based on €8,200/year, get €4,500/year. Syndic loses credibility. We get blamed.

Backup plan: Use conservative estimates (70% of PVGIS theoretical). Add installer validation: "These figures will be confirmed by the installer during site visit." Never quote a number without a confidence range.

2. Legal / Delegation Assumptions

Syndics can legally delegate AG motion preparation to a third‑party software service.HIGH RISK

Pressure test: The AG motion is a legal document proposing a multi‑thousand‑euro investment to a copropriété. Does French law require this to be drafted by the syndic themselves, or can a third party (non‑notaire, non‑avocat) generate it? If the motion is challenged in court, does "generated by software" create liability for us? Do we need to be registered as a legal document preparer?

If false: The core value prop (instant AG docs) is legally void. We become a template seller, not a concierge.

Backup plan: Frame AG docs as "draft templates for review" not "official legal documents." Add disclaimer: "This motion is a starting draft — review with your legal counsel before submission." Partner with a notaire or copropriété lawyer for final review (€50/motion, passed to customer).

Syndics can delegate ABF permit filing and Enedis connection requests to us.HIGH RISK

Pressure test: ABF permits require building ownership details, architectural plans, and often the syndic's signature. Enedis connection requests require the mandataire to be the property owner or have a signed power of attorney (mandat). Can a software company legally file these on behalf of a copropriété? Do we need to be registered as a mandataire? What liability do we have if the permit is rejected due to our error?

If false: "We handle everything" becomes "We prepare drafts, you file them." The concierge promise is broken.

Backup plan: Two options: (a) Act as mandataire with formal power of attorney per project, or (b) Pre‑fill all paperwork and provide DocuSign links for the syndic to submit in their name. Option (b) is weaker but legally safer. Test with a notaire.

We can negotiate and sign contracts with installers on a syndic's behalf.HIGH RISK

Pressure test: Installer contracts are typically 5‑figure commitments. Can a third party (us) select the installer and commit the copropriété to a contract? This likely requires a signed mandat. Without one, the syndic must sign. Which means our "one‑click selection" is actually "one‑click recommendation, syndic signs contract." Is that still concierge?

If false: The installer selection step requires human sign‑off on a €30–50k contract. We can't fully automate it.

Backup plan: Reframe: "We negotiate terms with the installer and prepare the contract. You review and sign electronically." Still adds massive value. We become the proxy, not the principal.

€1,200 covers the full coordination effort until commissioning.MEDIUM RISK

Pressure test: How many hours does ABF filing + Enedis chase + subsidy paperwork + installer coordination actually take? In Paris, ABF delays are 2–6 months. Enedis grid studies add 3–9 months. Subsidy filings are complex. If one project requires 20+ hours of human coordination, €1,200 may be unprofitable unless we scale.

If false: We lose money on every project or cut corners, damaging reputation.

Backup plan: (1) Test on 5 pilot projects to measure actual time. (2) Build automation for repetitive steps (form pre‑filling, status tracking, email follow‑ups). (3) Price at cost + 30% margin after pilots. €1,200 may become €1,500–2,000.

3. Flow / UX Assumptions

Syndics prefer a wizard over talking to their existing property manager or installer.MEDIUM RISK

Pressure test: Most Paris syndics already have a gestionnaire de copropriété (property manager) who handles day‑to‑day. Why would the syndic use us instead of asking their existing manager to handle solar? What’s the switching cost? Are we competing with inertia, not just other solar services?

If false: The target customer prefers to delegate to their existing relationship. We're a novelty, not a necessity.

Backup plan: Don't replace the property manager — enable them. White‑label the wizard for gestionnaires. Or frame us as "the specialist their manager doesn't have time to be."

Step order: installer quotes should come after AG vote.MEDIUM RISK

Pressure test: Wait — our wizard offers quotes before the AG vote. But copropriétaires might want to see a concrete installer quote before voting. Do we need quotes in the AG motion to get the vote? Or do we get the AG motion first, then get quotes? Which order actually works in practice?

If wrong: We build a flow that syndics abandon because it asks them to get quotes too early (wasting installer time) or too late (AG motion lacks financial specifics).

Backup plan: Research 5 real Paris AG solar motions. See if quotes are included. Ask 3 syndics: "Would you want quotes before or after the AG vote?" Build both paths into the wizard: "Include rough estimates [fast]" vs. "Include firm quotes [2‑week delay]."

Syndics will trust software to recommend an installer.MEDIUM RISK

Pressure test: Installers are a relationship business. Syndics typically use whoever their property manager recommends, or the installer who serviced their boiler. Trusting a faceless algorithm to pick a €30k contractor feels risky. What happens if our recommended installer goes bust or delays?

If false:>/strong> Syndics ignore our recommendation and shop independently. We lose the coordination value.

Backup plan: Transparency wins: Show why we recommend each installer (certifications, local references, copropriété experience). Add third‑party validation: "Verified by [inspection body]." Let them override our pick but explain the trade‑offs.

"Free to explore" converts to "pay to start" at meaningful rates.HIGH RISK

Pressure test: If 100 syndics generate AG docs for free, how many pay €1,200 to start the project? If conversion is <5%, we have a business model problem. Free tools attract tire‑kickers. The real test is whether the wizard creates enough trust/urgency to convert.

If false: We become a free AG‑motion factory with no revenue. The €1,200 price point may be too high or the value too abstract.

Backup plan: A/B test pricing (€999 vs. €1,200 vs. €1,500). Add urgency: "Subsidies decrease 10% next quarter — lock in now." Consider a "success fee" model instead: free upfront, €500 at permit approval, €700 at commissioning.

4. Technical / Infrastructure Assumptions

We can integrate with installer APIs or email systems for quote collection.LOW RISK

Pressure test: Paris solar installers are small businesses. Do they have APIs? No. Do they respond to structured email quote requests? Sometimes. Will they engage with a third‑party software company sending them leads? Maybe, if the lead quality is high. But each installer has a different quote format. Normalizing their responses into a comparison table requires parsing unstructured PDFs/emails.

If false:>/strong> We can't automate quote comparison. We become a manual service again.

Backup plan:>/strong> Start with a quote request template we send to installers, and a human normalizes the first 20 responses while we build a parser. Lock installers in with an exclusivity incentive: "First 3 leads per month guaranteed if you use our quote form."

Real‑time project tracking (ABF, Enedis, subsidies) is possible.HIGH RISK

Pressure test: ABF status: no public API. You call or email the Architecte des Bâtiments de France. Enedis: has a portail, but no API for third‑party tracking. MaPrimeRénov' has an online tracker, but access is for the applicant (the syndic), not a third party. Can we actually build a "live timeline" or is it manually updated based on emails?

If false: The "I'll track everything" promise is a manually updated spreadsheet pretending to be software.

Backup plan: Honest about limitations: "We track milestones based on our follow‑ups with authorities. You'll get an update each Monday." Build a mandataire model where we have login access to their Enedis/MaPrimeRénov' accounts. Proactive, not real‑time.

5. Regulatory Realities We Can't Ignore

⚠️

Copropriété law requires AG votes for work >€25k. Solar installations almost always trigger this. Our AG motion must comply with Article 24‑2 and 25 of the Loi ALUR. Are we sure our template covers all required clauses?

⚠️

Unanimity or ⅔ majority? Some copropriété statutes require unanimity for structural work or roof modifications. Our wizard must ask this before generating the motion.

⚠️

Syndics have fiduciary duties. Delegating decisions to software could be challenged as failing their duty of care. We're not liable — but if they lose a court case, they'll blame us.

⚠️

GDPR. We're processing building addresses, syndic contact info, copropriétaire data if we handle AG docs. Do we need a DPO? Are we transferring data to installers? Consent management?

⚠️

Subsidies require the applicant to be the property owner. MaPrimeRénov' and CEE are claimed by the copropriété, not by us as a service provider. We can assist with paperwork, but the syndic must sign and submit.

⚠️

Insurance. If we recommend an installer who damages the roof, are we liable? Do we need professional liability insurance (respossabilité civile professionnelle) as a mandataire?

6. Brutal Honesty: What Will Likely Break First

🔴 Most Likely Failure: Building Discovery

We can't auto‑find buildings from email domains at scale. The "I found 3 buildings" wow moment works for 30% of syndics, dead‑ends for 70%. Result: syndics bail at Step 2.

Fix: Ask for one address immediately. Use it to power the rest of the wizard. Treat auto‑discovery as a bonus, not the entry point.

🟡 Likely Failure: AG Motion Quality

Our automated AG motion is decent for 80% of buildings, wrong for 20% (unique PLU restrictions, copropriété statutes, special majority rules). A bad motion gets rejected or challenged.

Fix: Add a legal review gate for the first 50 motions (partner with a copropriété lawyer). Learn edge cases. Build a "motion confidence score" — if <90%, flag for human review.

🟡 Likely Failure: Installer Relationships

Installers don't trust our leads at first. They ignore quote requests or give inflated prices because they don't know us.

Fix: Pre‑qualify 10 installer partners before launch. Guarantee them exclusive arrondissement access for the first 20 projects. Build trust before automation.

🟠 Slow Burn Failure: Unit Economics

€1,200 assumes we automate 80%+ of coordination. In reality, the first 20 projects need heavy human hand‑holding. At 10 hours/project, that's €120/hour — decent but not scalable. If automation only reaches 50%, we need €2,000+ to profit.

Fix: Measure actual time per project. Adjust pricing after 10 pilots. Don't lock in €1,200 until we know the real cost.

7. Questions We Must Answer Before Coding

?

Can a non‑lawyer third party legally generate an AG motion in France?

?

What formal mandat do we need to file permits and requests on behalf of a copropriété?

?

What percentage of Paris syndic emails are on recognizable corporate domains vs. generic Gmail/Orange?

?

Do real Paris AG solar motions include installer quotes, or just financial estimates?

?

What is the actual average time (and cost) of an ABF → Enedis → subsidy → installation project in Paris?

?

Would 3 real Paris syndics actually use this, or do they trust their existing property manager more?

Bottom line: The TurboTax wizard is brilliant UX. But the legal, data, and relationship assumptions are brittle. The MVP should validate can we build this and should we build this in parallel, not sequentially. Start with the questions above.