Annual pricing, packaging, cost to serve, and the ₹2 crore question

Part of car_sales — the dealership operating layer. Evidence-class legend there.

⚠ No Indian dealer has been asked anything. Every willingness-to-pay figure is 🔎 or ❓ unless tagged otherwise. Competitor list prices and the audited P&L figures are 🌐 verified verbatim. See the index's limitation notice.

0 · The verdict, before the evidence

The previous price book was wrong, and it was wrong in the direction that kills the company quietly

The ₹30,000/yr Starter tier published on this page until 2026-08-23 was below cost to serve under any support assumption that is not fantasy, and 2.6x below the cheapest verified competitor at the headcount it was aimed at. It has been withdrawn, not repriced. QRS-844.

Three things changed. All three came from arithmetic, not from a preference for higher prices.

1. Monthly pricing is removed from every surface. Not softened, removed. The reason is not negotiation psychology, it is that this product's value curve is back-loaded — see §1.
2. The floor rises from ₹30,000 to ₹79,999/yr and the volume tier from ₹72,000 to ₹1,49,999/yr, because the market's own published per-user prices, applied to a real dealership's headcount, land between ₹77,000 and ₹9,07,000 per outlet per year — see §3.
3. The workshop and used-car segments are deliberately abandoned, which is a real loss of reachable logos and the correct trade — see §9.

And one conclusion the owner will not want, stated first because burying it would be dishonest:

🔎 Raising the price does not materially accelerate the path to ₹2 crore, because the binding constraint was never price — it is how many accounts two people can sell and then serve. What repricing does is make every account profitable, and reduce ₹2 Cr from 200 signatures to 13-27 (§10). That is the whole prize, and it is a large one. But August 2027 gets worse on logo count, not better.

1 · Annual commitment only, and the real reason

The owner's instinct is right and the stated reason is the weaker of the two available.

The negotiation argument is true and secondary

🔎 Publishing "₹2,500/mo or ₹30,000/yr" does invite a negotiation about ₹2,500. Every discount conversation then anchors on the small number, and the annual figure becomes the thing you climb back toward.

The structural argument is decisive, and it is a product fact rather than a sales preference

The value this product creates in month 1 is near zero, and it compounds. An attribution spine has nothing to attribute until interactions accumulate; a manager dashboard has nothing to show until reps carry cards; a service-reminder engine needs a vehicle history it does not yet have.

🔎 So a monthly contract systematically terminates before the product works. That is not churn caused by a bad product, it is churn caused by a billing period shorter than the value-realisation period. A month-to-month dealer cancels in month 2 having correctly observed that nothing happened yet.

Therefore the minimum honest contract is 12 months, and the term must start at go-live, not at signature (§7).

What that means in practice, including where it must be constrained

Decision	Rule
List price	Published only as an annual figure. QR Setu Growth — ₹1,49,999/year
A monthly equivalent	Never rendered anywhere — not the site, not the deck, not the invoice, not the in-product billing screen. 📘 This is a product constraint, not only a sales one: QRS-845
Term	12 months minimum, 24 and 36 offered at the same annual price with onboarding waived. Discount the term, never the price (§8)
Collection	⚠ Term and cash flow are different questions. Annual term, quarterly collection by e-mandate is the default
Upfront annual cash	Offered at roughly one month's equivalent off. Cash preference is priced, never demanded

Do not demand annual cash upfront from this buyer

🌐 Popular Vehicles, a listed ₹6,381-crore dealer group, reports FY26 EBITDA of 3.18% and PBT of minus ₹13.30 crore. 🔎 A buyer losing money at the bottom line has a treasurer, and asking for ₹1.5 lakh in one cheque adds a second objection to a deal that already has enough. Annual commitment answers the churn problem; quarterly collection answers the cash problem. They are not the same problem and one mechanism should not be asked to solve both.

⚠ 📘 And do not assume a collection cost: this repo's standing rule is that no branch, threshold or projection may depend on a payment-provider rate, because Digious is on a default plan and renegotiates at volume. Record actual provider-reported figures; never compute from an assumed one.

2 · What ₹30,000 a year actually cost us

🔎 The estimates below are labelled because they are estimates. The conclusion survives across the whole plausible range, which is what makes it usable.

| Cost line | Foundation
10-24 cards | Growth
30-60 cards | Complete
75-100 cards | Basis |

	1 outlet + 9-24 staff	1 outlet + 29-59 staff	1 outlet + 74-99 staff	📘 The outlet card counts (§4)
Supabase marginal — storage + egress	₹600	₹1,500	₹3,000	🌐 $0.021/GB/mo storage, $0.09/GB egress. Genuinely negligible
Observability, error tracking, share of fixed	₹300	₹500	₹800	🔎
Support, steady state	2 h/mo	4 h/mo	6 h/mo	🔎 Password resets, a new joiner needing a card, a leaver to deactivate, "the report looks wrong", a rejected message template
Support at ₹1,500/h loaded	₹36,000	₹72,000	₹1,08,000	🔎 Conservative founder opportunity cost
Cost to serve, year 2+	₹36,900	₹74,000	₹1,11,800
Gross margin at list	50.8%	51.4%	63.7%

The number that withdraws the ₹30,000 tier

₹30,000/yr at…	Cost to serve	Gross margin
2 h/month of support	₹36,900	minus 23%
1 h/month of support	₹18,900	37%, and thin

🔎 A ₹30,000 account is loss-making at two hours a month, and nothing suggests a 30-person dealership in its first year consumes one. Onboarding alone is 20-40 hours before steady state begins.

⚠ Infrastructure is not the argument and must not be used as one. Marginal hosting is under ₹1,500 a year. The cost of this business is human hours, which is exactly why a price that looks generous on an infra spreadsheet is fatal on a calendar.

The corollary that governs the entire commercial strategy

🔎 Cost scales with LOGOS, not with rupees. Two accounts at ₹1.5 lakh cost roughly the same to serve as two accounts at ₹36,000 and earn four times as much. So every rupee of the target should be won from the fewest possible accounts — which is the same conclusion the sales-capacity analysis reaches from the opposite direction, and the two constraints reinforce rather than trade off.

3 · What the market actually charges, at a real dealership's headcount

This is the correction that matters most, and it was hiding inside our own evidence

Market landscape §3 recorded AutoBooom's verified list price and then evaluated it at ten users — concluding "₹1 lakh is 3.4x AutoBooom."

🧮 A ten-user outlet is not the customer. Recomputed at the customer-facing headcount this product is priced for, from the same 🌐 verified figures — ₹10,000 one-time plus ₹2,400/user/yr plus ₹5,000/yr AMC:

Users	AutoBooom year 1	AutoBooom steady state
10	₹39,000	₹29,000
20	₹63,000	₹53,000
30	₹87,000	₹77,000
40	₹1,11,000	₹1,01,000
50	₹1,35,000	₹1,25,000

The cheapest verified dealer-software incumbent in this study already charges ₹77,000 to ₹1,01,000 per outlet per year at 30-40 users. Our ₹30,000 was not aggressive pricing against it. It was 2.6x below it, for a broader product.

🌐 And the rest of the published market, at the same 30-user outlet:

Vendor	Published rate	At 30 users
AutoBooom	₹2,400/user/yr + ₹5,000 AMC	₹77,000/yr
Zoho CRM India	₹800-2,600/user/mo	₹2.88-9.36 lakh/yr
eMsys AutoNet	$30/user/mo	₹9.07 lakh/yr
Rocket ERP	per year	₹1.70 lakh/yr
Eazy DMS	per year	₹8.00 lakh/yr
Cloud telephony — 10 users, calls only	₹5,000-15,000/mo	₹60,000-1.80 lakh/yr
RAMP — workshop, the Layer 4 ceiling	₹175-3,499/mo	₹42,000/yr

The three comparisons that work in a room, and the one that does not

🔎 Use these:

1. Their own advertising and sales promotion line. 🌐 ₹14.0 lakh per sales outlet per year, audited. Growth at ₹1.5 lakh is 10.7% of a line they already spend, on the mechanism that tells them which of that spend produced a named customer.
2. Per-user software applied to their own headcount. They can do this division themselves, and every answer is larger than ours.
3. Cloud telephony. ₹60,000-1.80 lakh a year buys calls. That is one channel, no attribution, no hierarchy, no cards.

Do not use the "a dealership spends ₹20,000 servicing one car" comparison

It is intuitively powerful and it will lose you the room, for two reasons.

1. It compares a software subscription to gross revenue on a job that carries parts and labour cost. The dealer's own controller knows the difference and will say so, and the pitch then reads as somebody who has not looked at their P&L. 🌐 The margin on that job is 41.2%, not 100%.
2. ⚠ It leads the buyer straight to a conclusion we cannot accept: "then charge me per car." Per-transaction pricing makes our revenue a function of their volume rather than our value, and on the finance and insurance side it edges toward commission — 📘 which is licence-gated and already refused (product-scope).

Use the ₹20,000 ticket for the break-even calculation in §6, where it belongs, and never as a price comparison.

4 · The price book — four annual packages

⚠ REVISED 2026-08-23 — annual only, per OUTLET, tiered by who consumes the capability

Supersedes the per-card tiers of 2026-08-22, which superseded a flat per-outlet model. The unit is the outlet; the card count is a bounded ceiling with a priced overflow.

🔎 The per-card framing was right that a dealer will buy "a card for each of my staff" more readily than "30 seats". It was wrong to make card count the pricing meter, because that reintroduces exactly the per-seat negotiation it was designed to avoid — the dealer counts staff and divides. Quote one annual number for the outlet. State the card ceiling, and price the overflow.

⚠ The word "rooftop" is gone. It was undefined jargon — see §4a.

⚠ ALL PRICES ARE EXCLUSIVE OF 18% GST

Every figure on this page is ex-GST. 18% GST is added to the quoted price on every invoice — subscription, additional cards, onboarding, add-ons and managed services alike. 🔎 Quote it as "₹1,49,999 per year plus 18% GST", never as an inclusive number and never silently. See §5a.

Package	Annual price ex-GST	Cards included	Additional cards	Mgmt logins	HARD CAP	Sellable from
QR Setu Foundation	₹79,999	10 1 outlet + 9 staff	₹4,999/card/yr	5 cap 10	24 cards 1 + 23 staff	Wave 1 · Dec 2026
QR Setu Growth	₹1,49,999	30 1 outlet + 29 staff	₹4,999/card/yr	12 cap 20	60 cards 1 + 59 staff	Wave 2 · Apr 2027
QR Setu Complete	₹2,99,999	75 1 outlet + 74 staff	₹4,999/card/yr	25 cap 45	100 cards 1 + 99 staff	Wave 3 · Aug 2027
QR Setu Enterprise	quoted, from ₹4,24,999	negotiated	negotiated	negotiated	named in the contract	❓ On demand, never listed
Group Intelligence	₹79,999 per outlet	—	—	—	min 3 outlets	Wave 3 · Aug 2027
Onboarding and activation	₹59,999 / ₹1,24,999 / ₹2,49,999 one-time	—	—	—	by tier	Wave 1
Additional management login	₹2,499/yr	—	—	—	see §4b	Wave 1
Physical QR standees	₹225 each	—	—	—	per unit	⚠ after the card renders images (§5d)
QR Setu AI — Assist / Agent / Agentic	₹29,999 / ₹99,999 / ₹1,99,999	—	—	—	metered AI actions	⚠ Wave 1 drafting only (AI strategy)
Digital growth add-ons	metered or per engagement	—	—	—	never bundled (§5)	Wave 1 onward
Marketplace visibility add-ons	⚠ not sellable yet — §5b	—	—	—	—	❓ unscheduled

The flat ₹5,000 additional-card rate makes the whole book explicable in one sentence, and that is worth more than an optimised rate

🧮 Measured against the tiers it sits beside:

Tier	Implied rate inside the bundle	vs the ₹5,000 card
Foundation	₹79,999 / 10 = ₹8,000/card	above it — a small-account premium
Growth	₹1,49,999 / 30 = ₹4,999/card	exactly it
Complete	₹2,99,999 / 75 = ₹4,000/card	below it — a volume discount

₹5,000 per Setu Card per year is the reference price. Foundation sits above it, Complete below it, and anything beyond a ceiling sits exactly on it.

🔎 That is a sentence a dealer can repeat to their partner after the meeting, which is a property no optimised rate card has. And the margin is excellent — the marginal cost of one more card is effectively zero (§2), so an additional card is close to pure contribution.

⚠ The outlet's own Setu Card COUNTS inside the limit. So quote staff + 1

📘 Owner clarification 2026-08-23. Every card counts against the tier, including the dealership's own organisational card. There is no card that sits outside the limit.

Card	Counts?	Note
The outlet's organisational Setu Card	Yes	One per outlet, always active, so it permanently occupies one slot
Every employee Setu Card	Yes	Sales consultants, team leads, service advisors, reception, F&I, managers
A GROUP-level card for a multi-outlet dealer	No	Ships with Group Intelligence and consumes no outlet's allowance. Otherwise it would arbitrarily eat one outlet's slot

⚠ The quoting rule this creates, and it is the one a salesperson will get wrong

Count customer-facing staff. Add one for the outlet card. THEN pick the tier.

⚠ A ten-person outlet needs 11 cards, not 10 — so it is ₹84,998, not ₹79,999. Quoting Foundation flat to a ten-person shop under-quotes by one card and, worse, leaves one real person without a card. 🔎 That is not a rounding error: 📘 a staff member without a card has a rational reason to share one, which silently destroys the named-employee-to-named-customer chain that is the entire product thesis (QRS-850).

Price by actual headcount — and Foundation reaches further than its ceiling suggests

🧮 Cheapest option at each real customer-facing headcount, ex-GST:

Customer-facing staff	Cards needed	Cheapest package	Annual, ex-GST
9	10	Foundation	₹79,999
10	11	Foundation + 1	₹84,998
14	15	Foundation + 5	₹1,04,994
20	21	Foundation + 11	₹1,34,988
23	24	Foundation at its cap ≈ Growth base	₹1,49,985 ← crossover
24-29	25-30	Growth	₹1,49,999
30	31	Growth + 1	₹1,54,998
40	41	Growth + 11	₹2,04,988
59	60	Growth at its cap ≈ Complete base	₹2,99,969 ← crossover
74	75	Complete	₹2,99,999
99	100	Complete at its cap	₹4,24,974
100+	101+	Enterprise, quoted	from ₹4,24,999

🔎 Two things worth reading off this table. Foundation with additional cards is the cheapest option all the way to 24 staff, so the small-outlet SKU is not as narrow as a 10-card ceiling makes it look — and the band from 25 to 29 staff is a flat ₹1,49,999, which is the cleanest place in the whole book to land a standard franchise outlet.

Why ₹5,000 specifically — the constraints leave exactly one round number

🧮 The rate is bounded on both sides, and only one round figure survives:

Bound	Value	Why
Floor — monotonicity, Foundation → Growth	₹3,500	(₹1,49,999 − ₹79,999) ÷ 20 cards. Below this, overflowing to 30 cards beats buying Growth and the ladder stops working
Floor — monotonicity, Growth → Complete	₹3,333	(₹2,99,999 − ₹1,49,999) ÷ 45 cards
Floor — arbitrage	₹4,000	Complete's own bundle rate. Price overflow below it and the largest accounts buy cards instead of tiers
Ceiling — defensibility	₹5,000	Growth's own bundle rate. Above it, card 31 costs more than card 30 in the same plan, which is a question with no good answer

🔎 So the window is ₹4,000 to ₹5,000, and inside it:

Candidate	Upgrade-cheaper margin, F→G	Verdict
₹4,000	only ₹5,000, about 3%	⚠ Too thin. A dealer stays on Foundation with 20 extra cards and we lose the capability upsell for nothing
₹4,500	₹15,000	Matches no bundle rate, so it needs explaining
₹5,000	₹25,000	✅ Strongest upgrade incentive, equals Growth's bundle rate, roundest number

₹6,000 was rejected for being 20% above Growth's bundle rate, and a declining 6,000/5,000/4,000 ladder for costing three numbers to gain very little. 🔎 The marginal cost of a card is effectively zero, so this is a positioning decision rather than a cost-recovery one — which is exactly why the defensible-sentence test should decide it.

⚠ Three mechanics matter MORE than the rate, and getting any of them wrong is worse than mispricing by ₹1,000

1. The ceiling counts ACTIVE cards, and a leaver frees a slot. 🌐 At 29.53% frontline attrition a 30-card outlet issues roughly 39 cards a year. Billing per card ever issued would charge them for nine phantom people and is the kind of invoice that ends a contract. Deactivating a card must return the slot immediately — and it must be the same action that revokes the leaver's access, never a separate billing step somebody forgets.
2. Additional cards pro-rate to the end of the term. A card added in month 11 costs a month, not a year. Otherwise the honest act of issuing a card to a new joiner in March carries a full-year charge, and the dealer's rational response is to share a card — which silently destroys attribution, the product's entire thesis.
3. The ceiling is SOFT: issue the card, then bill it. Never block. 🔎 A cap that stops a new joiner getting a card on their first morning is a support call at the exact moment the product is working, and it trains the dealer that the ceiling is an obstacle rather than a plan.

📘 All three are billing-surface requirements, and none exists yet — QRS-850. ⚠ Note that (1) and (2) both point the same way: the enemy is a dealer with a rational reason to share a card. Price and mechanics must never create one.

Every package includes mobile AND desktop. It is a headline, not a footnote

📘 The merchant product is one universal codebase (ADR-0011) that ships to Android native, iOS native and the browser from the same source, mounted once at /app (ADR-0028). So multi-surface access costs us nothing incremental and must be stated on every price card and every slide:

Android app · iPhone app · desktop and mobile browser. Same login, same data, every plan.

🔎 It is a genuine competitive line rather than a feature bullet: 🌐 "mobile-first field app" is the named gap in every OEM DMS critique in this study, and a service advisor on the floor is not walking to a desktop to log a follow-up. 📘 Cross-platform parity is a non-negotiable standard in this repo, so the claim is structurally true rather than a roadmap promise.

⚠ What must NOT be claimed, and it is a live gap. Desktop today is the responsive web build — real, working, same codebase, same data. The purpose-built desktop console the design specifies is a separate 16-screen track, and 🧮 6 of 20 designed desktop screens have a route (QRS-815 · npm run check:desktop-parity). Sell "works on desktop", which is true. Do not demo a designed console layout that is 30% built.

⚠ NO PLAN IS EVER OPEN-ENDED. Every tier states a hard card cap

📘 Owner instruction 2026-08-23: "do not keep any plan open ended with setu card limits." Correct, and the previous version of this page violated it in a way that was easy to miss — the tiers were capped but the overflow was not, so a Complete customer could meter upward forever. An uncapped meter is an open-ended plan wearing a ceiling.

No QR Setu plan says "unlimited". Not on a price card, not in a contract, not in the product. Every agreement names a number.

And the caps are not arbitrary — each one is where its own price meets the next tier's

🧮 At ₹5,000 per additional card:

Tier	Included	Hard cap	Price AT the cap	Next tier's base
Foundation	10	24	₹79,999 + 14 × ₹4,999 = ₹1,49,985	Growth ₹1,49,999	within ₹14
Growth	30	60	₹1,49,999 + 30 × ₹4,999 = ₹2,99,969	Complete ₹2,99,999	within ₹30
Complete	75	100	₹2,99,999 + 25 × ₹4,999 = ₹4,24,974	Enterprise, from ₹4,24,999	within ₹25

🔎 So the cap costs the customer nothing. At the moment a tier runs out of cards, the next tier is available at the same price with more capability — the cap sits exactly where a rational buyer would move anyway. That is what makes a hard limit defensible in a negotiation rather than something to apologise for.

⚠ And below the cap the ceiling stays SOFT — issue the card, then bill it, never block. The two rules are not in tension: soft below the cap so a new joiner is never turned away on their first morning; hard AT the cap so no plan is open-ended. A dealer at 23 cards on Foundation gets card 24 immediately and a conversation about Growth; a dealer already at 24 gets the conversation first. 🔎 In staff terms that is 22 staff waved through, 23 staff talked to — because one slot is always the outlet's own card.

🔎 Worked example of why the middle band matters. A 14-staff outlet needs 15 cards — one for the outlet, fourteen for people — and pays ₹1,00,000 on Foundation rather than ₹1,49,999 on Growth. They buy capacity without being forced to buy capability, and Foundation stays cheapest all the way to 23 staff (full table).

Enterprise exists so that "above the cap" has an answer that is not a meter

🔎 Beyond 100 customer-facing staff at a single outlet, an open-ended per-card meter is the wrong instrument: it prices above Complete's own bundle rate (₹4,000/card), which a sharp buyer will notice and should, and it turns a large account into a spreadsheet argument.

Enterprise is quoted, anchored at ₹4,24,999, and its card cap is written into the contract. ⚠ It is deliberately not listed and has no published ceiling ladder — inventing list prices for a customer shape nobody has met is how the prototype acquired features nobody asked for. What it must always have is a number, because that is the invariant this section now holds.

⚠ Tightening the ceilings pushes the STANDARD franchise outlet out of Foundation, deliberately

🔎 A franchise outlet's customer-facing headcount — sales consultants, team leads, service advisors, reception, F&I, managers — is realistically 15-40 people. At the previous ceiling of 15 cards Foundation covered the low end of that band; at 10 it does not.

So the standard outlet now lands on Growth at ₹1,49,999, and Foundation becomes a genuinely small-outlet SKU. That is commercially better — the volume buyer sits on the volume SKU instead of under-buying — and 🔎 the overflow rate recovers most of what the tighter ceiling would otherwise cost: a 20-card Foundation outlet is ₹1,29,989, not ₹79,999. Stated because it is a consequence of the change rather than something to discover in a quote.

⚠ The price book spans an eighteen-month build, and selling from it today would be selling vapour

🧮 Only Foundation is sellable in 2026. Growth needs hierarchy, attribution and dashboards; Complete needs automation and campaigns; Group Intelligence needs an organizations write path, seats and billing that 🧮 do not exist in any form (index).

So the ladder is also a timeline. In a Dec-2026 pitch, Growth and Complete are named and dated, never demonstrated, and the contract sells Foundation with a written upgrade price. Quoting Complete off a slide in 2026 is how a pilot becomes a refund.

4b · Every kind of access is capped, not just cards

⚠ The owner found a real hole: we metered the CARDS and gave away the LOGINS

📘 Owner instruction 2026-08-23: "we should not create a situation where we strictly charge for Setu Cards but effectively provide unlimited administrative and operational access for free." Correct, and the hole was worse than it looks, because it was in the wrong direction commercially.

🔎 The tier ladder gates on who consumes the capability — Growth is "managers consume it" (§4). So an unlimited management-login allowance gives away the exact thing each tier is for. A dealer could buy Foundation, add twelve manager logins at no cost, and consume the Growth proposition. And support cost tracks logins, not cards: a card is scanned, a login asks questions. QRS-851.

The distinction that has to come first, because conflating it would be a schema error

📘 In this platform the principal is a USER; a Setu Card is an artefact. They are not the same thing and they do not have the same population:

	Has a Setu Card	Has a login
Sales consultant, service advisor, team lead	Yes — it is their public identity	Yes, and it comes with the card
Receptionist	Usually no	Yes — logs walk-ins
CRM head, marketing, accounts, back-office	No	Yes
GM / Dealer Principal	Optional	Yes
The outlet itself	Yes — the organisational card	n/a

The rule, and it deliberately refuses to charge twice for one person

Every Setu Card includes one login for that person. Management and back-office logins — people who log in without needing a public card — are a separate, capped allowance.

🔎 Charging separately for a rep's login when their card is already paid for is double-dipping on the same human, and it would re-create the incentive to share a card. Card-holders' logins are therefore free by construction, and the meter sits exactly where the uncapped hole was.

The limits matrix

	Foundation ₹79,999	Growth ₹1,49,999	Complete ₹2,99,999	Enterprise from ₹4,24,999
Setu Cards included	10 1 outlet + 9 staff	30 1 + 29	75 1 + 74	negotiated
Setu Card hard cap	24	60	100	contract term
Additional Setu Card	₹4,999/yr	₹4,999/yr	₹4,999/yr	contract term
Management logins included	5	12	25	negotiated
Management login hard cap	10	20	45	contract term
Additional management login	₹2,499/yr	₹2,499/yr	₹2,499/yr	contract term
Media storage (photos, video)	5 GB	25 GB	100 GB	negotiated
Additional storage	₹999 per 10 GB/yr	same	same	contract term
Outlets per subscription	1	1	1	multiple
Mobile app + desktop	✅	✅	✅	✅
WhatsApp	recharge	recharge	recharge	recharge
Google Business portal	monitor only	full	full	full
Physical QR standees	₹225 each	₹225 each	₹225 each	volume terms

🔎 Why 5 / 12 / 25, revised the same day. The first figures were 3 / 8 / 20, computed before the Team Leader and Sales Manager cards were withdrawn (QRS-854). Those managers now take logins instead of cards, so a 30-staff outlet needs reception 2 + 4 team leaders + 1 sales manager + GM + CRM head + marketing + accounts + org admin = 12. ⚠ Leaving the allowance at 8 would have billed ₹9,996 for the very seats we had just told the dealer they need instead of cards — moving a charge rather than removing one. 🧮 Net effect of the whole change: minus ₹4,999 per outlet.

⚠ The additional-login price is deliberately LOW, and not because the cost is low

🔎 On cost alone a management login should arguably cost more than a card — a GM refreshing a dashboard is more expensive to serve than a card being scanned, in both support minutes and queries.

It is priced at ₹2,499 anyway, because the risk being managed is not revenue — it is shared logins. A receptionist login expensive enough to be worth sharing destroys the visitor register's attribution (who logged this walk-in), which is the same defect as a shared Setu Card one layer down. 🔎 The rule that generalises across this whole price book: never price an access seat high enough to make sharing rational.

4c · What to meter, and what must never be metered

📘 The owner asks for feature usage limits as well. 🔎 The honest answer is that most of them should not exist, and the reason is not generosity.

	Metered?	Why
Setu Cards	Yes, hard cap	The artefact being sold
Management logins	Yes, hard cap	Where the cost and the tier value both sit
WhatsApp messages	Yes, prepaid	🌐 Genuinely variable cost we pay Meta for
Media storage	Yes, soft	🌐 Storage and egress are the only infrastructure line that scales without bound, and 📘 the owner wants video on the org card
Visitor-register entries	⚠ NEVER	It is the adoption wedge. Metering the one behaviour we are trying to create is self-defeating
Leads, contacts, customers stored	⚠ NEVER	It is the moat. A cap gives the dealer a reason to delete data, which destroys the accumulation the product's value rests on
Catalogue items / vehicles	⚠ NEVER	An incomplete catalogue makes the org card worse, which makes every standee worse
Test-drive bookings, follow-ups, feedback	⚠ NEVER	Same reason. These are the outcomes we are paid to increase
Card scans / views	⚠ NEVER	Capping scans caps the customer's behaviour, not the dealer's

The principle, stated so it survives the next request to cap something

Meter what costs money and scales without bound. Never meter the behaviour you are trying to create.

🔎 A cap on walk-in entries or stored leads would earn a few thousand rupees and cost the product its thesis. Cards, logins, messages and storage are the only four meters that pass both tests.

⚠ And four meters is already a lot to build for two people, so ship them in order

🧮 Every meter needs a counter, a display, an enforcement point and a support conversation. Recommended sequencing: cards and logins as hard meters in Wave 1 (they gate the commercial model); messages are already metered by the recharge; storage stays a monitored SOFT limit — we watch it, we do not enforce it — until a real dealer actually exceeds it. 📘 Building enforcement ahead of a single breach is process built ahead of need. QRS-851.

4a · What an "outlet" is, and why "rooftop" was the wrong word

⚠ "Rooftop" appeared 71 times across this section and was NEVER ONCE DEFINED

🧮 Measured 2026-08-23. It is US auto-retail jargon for one physical dealership site — standard in American dealer-software pricing and absent from Indian dealer vocabulary. 🌐 Every Indian source in this study says outlet: FADA's own release counts "over 15,000 Automobile Dealerships having over 30,000 dealership outlets", and the audited advertising figure is "₹14.0 lakh per sales outlet per year."

So the concept was right and the word was wrong. Using a term the buyer does not use, as the unit of the pricing model, with no definition anywhere in 71 appearances, is a self-inflicted objection in the first meeting. Renamed throughout. 🔎 The general lesson: jargon inherited from a foreign market's vendor literature reads as expertise while you are writing and as confusion while the customer is reading.

The concept, which is load-bearing and stays

An outlet is one place of business that has its own customers and its own numbers. A showroom. A service centre that runs its own P&L. A used-car yard operating as its own business.

📘 It maps exactly onto a workspace in the platform's tenancy tree (ADR-0022/0023), and that ADR already carries the test verbatim:

If a place needs its own card and its own P&L, it is an OUTLET. If it is only an address, it is a LOCATION.

Why a dealer needs the distinction, and what it saves them

🔎 This is the part with actual customer value, and it answers the question every multi-site dealer asks in the first ten minutes — "so what do I pay for my second service point?"

The dealer has	It is	Do they pay?	Why
A second brand's showroom across town	An outlet	Yes	Own customers, own targets, own team, own public card. A separate business wearing the same owner
A service centre with its own manager and P&L	An outlet	Yes	Same reason
A second collection or delivery point, no staff of its own	A location	No. Free	An address on the existing outlet's card, not a business
A stockyard	A location	No. Free	Nobody is served there
A temporary festival or mall kiosk	A location	No. Free	Staffed by the outlet's own people, carrying their own cards

⚠ This is the opposite of how per-site software usually prices, and it belongs in the pitch: a vendor charging per address bills you for your stockyard. We bill for businesses, not pins on a map.

Why it is the right pricing unit for this product specifically

Reason
It is what the customer already counts	A dealer principal knows exactly how many showrooms and service centres they run. They do not know their customer-facing headcount without asking HR
It scales with the value delivered	Each outlet has its own walk-ins, its own follow-up debt, its own reps to attribute. Two outlets is genuinely twice the product
It matches the schema, so billing is not a bolt-on	📘 One workspace, one subscription. Per-address billing would need a concept the tree deliberately does not have
It survives the group upsell	Group Intelligence prices per outlet consolidated, so the moat's price grows with the moat's value
It avoids the per-seat objection	🌐 Per-user pricing is unsellable against a 10-40 person Indian outlet, and the card ceiling plus a ₹5,000 overflow handles size without becoming a headcount licence

Why ₹79,999 per outlet for Group Intelligence rather than a flat group fee

🔎 The previous flat ₹1,50,000/group priced a 3-outlet group and a 15-outlet group identically, which inverts the value: consolidation is worth more the more brands there are to consolidate. Per-outlet scaling also fixes a proportion problem — a flat ₹2.4 lakh would have been 35% of a small group's total bill, which reads as a tax rather than a tier.

The gating principle: gate on WHO CONSUMES IT, never on what it cost to build

Tier	Who consumes the capability	Why the upgrade happens naturally
Foundation	The front line and the customer	Everything the dealership shows the outside world
Growth	Managers — team leader, sales manager	Bought the day a manager starts asking who produced what
Complete	The business	Bought when the dealership wants the system to act rather than report
Group Intelligence	The group	Bought when a second brand arrives
Enterprise	A single very large outlet, or an org-wide agreement	Quoted, never listed. Its card cap is a contract term

🔎 This is worth more than an arbitrary split for one reason: it explains itself in a pitch, and it maps to the org chart the dealer already has. A feature ladder built on implementation cost cannot be explained to a buyer at all, and every negotiation becomes bespoke.

⚠ Three product constraints that a three-tier ladder creates, and they are not sales concerns

1. 📘 Feature gating controls access, never discovery (CLAUDE.md fifth rule). Every capability above the customer's tier stays visible and locked with its value stated — not hidden, not greyed out.
2. ⚠ An unbuilt capability is availability-off, not entitlement-off, and they must not present the same way. A Wave-3 feature shown with "Upgrade to unlock" is a lie — no plan unlocks it, because it does not exist yet. It must read as not ready yet. 📘 ADR-0021 resolves effective = applicability AND entitlement AND availability; a price book spanning an 18-month build makes that distinction load-bearing rather than academic. QRS-846.
3. 📘 No in-app purchase CTA on native (ADR-0002 / Apple 3.1.3(d)). Discovery is unconditional on every surface; the action differs — a real CTA on web, "plans are managed on the web" on iOS and Android.

5 · Digital growth add-ons, and the WhatsApp contradiction this section had to resolve

⚠ Two pages of this section gave OPPOSITE answers on an IRREVERSIBLE decision

📘 index.md D5 read "WhatsApp recharges and dealer-branded sending are mutually exclusive. Pick recharges first" — which is ADR-0030 Option A. 📘 This page's §11 recommended Option B, dealer-owned, under which credits cannot be sold at all.

🌐 Option C attachment is irreversible once attached, and Option A pools messaging limits across every tenant. A section cannot hold both positions on a decision that cannot be undone.QRS-848.

Resolved in favour of Option A — and the first draft of this section got it backwards

⚠ I initially resolved this to Option B on the argument that "dealer branding IS the product". That was wrong, and the page that already held the right answer is operating model §8

🔎 Two things I had not weighed, either of which decides it alone:

1. The branding that matters is scoped to MARKETING, and Wave 1-2 sends UTILITY. Enquiry acknowledgement, test-drive confirm and remind, feedback, service due and insurance renewal are all 🌐 utility category at ₹0.1150 — sent to a customer who already has the dealership in their contacts, with the dealer's name in the message body. A verified sender in the chat header earns its cost when messaging strangers; it earns much less when confirming a test drive.
2. ⚠⚠ B is the ONE option that forecloses the endgame. 📘 Converting A → C is clean because we own the accounts. B → C means recreating every dealer account and forfeiting its accumulated quality history. So "start on B because branding matters" buys branding now and destroys the migration to the option that provides branding and credits later.

And the credits are not the reason to choose A. 🔎 At ~₹959/mo of message spend per marketing-active dealer, 30 dealers at a 20% markup is ₹69,000 a year — immaterial against ₹12-25 lakh of software ARR. So Option A is chosen for the migration path and for a working Wave-2 capability, not for the margin.

Test	Answer
Decision	Option A now — QR Setu's account sends, recharges are sellable, dealer name in the message body
Non-deferrable condition	⚠ 🌐 Limits pool per portfolio, so a per-workspace send cap must exist before the first campaign runs, or one dealer's festival blast throttles another dealer's paid service reminders. 📘 feature_grants.limit_value and on_exceed already support it
Graduation trigger	The first dealer who refuses to proceed without a verified sender starts Option C. Never pursued speculatively — 🌐 "a lengthy process", and irreversible once attached
Option B	Refused, and not merely deferred. It is the only branch with no path onward
When margin becomes a reason	🔎 ~₹6.9 lakh/yr at 300 dealers; ₹86.3 lakh at 750. Revisit at 300, decide at 750

🔎 This also matches the owner's original intent — WhatsApp as a metered recharge add-on outside the core subscription — which my first resolution had overruled on the weaker of the two available arguments. QRS-848.

The add-on book

Add-on	Price	Buyer inside the dealership	Verdict
Onboarding and data setup	₹59,999-2,49,999 one-time	Dealer Principal	Sell from day one. Cost recovery, and it seeds the data everything else needs
WhatsApp platform fee	₹17,999/outlet/yr + prepaid recharge packs of ₹2,000 / ₹5,000 / ₹10,000 drawn down per message	Marketing / CRM head	Yes, as software plus metered usage. Templates, opt-in ledger, scheduling, delivery reporting. ⚠ Never bundled: 🌐 a festival campaign can be 50x a normal month, and bundling transfers that variance onto us for no margin
Managed campaign operation	₹14,999/campaign or ₹20,000/mo retainer	Marketing	Yes, capped at ~15 accounts. ⚠ Labour, not software. A bridge, not a destination
Google Business portal	included from Growth; managed responses ₹59,999/yr	Marketing	Yes, and the TRIGGER is the differentiator — see §5c · 🌐 Podium reached $100M revenue by 2019 and a $3B valuation here (teardown)
Physical QR standees	₹225 each + 18% GST	Dealer Principal / Marketing	Yes, and ⚠ not before the org card renders images — see §5d
Festival and re-engagement campaigns	per campaign, as above	Marketing	Same line. Seasonal, and 🔎 concentrated enough to be worth capacity planning
Telephony integration	🌐 ₹1,300-1,999/user/mo	Sales Head	Integrate; resale margin only. Never own it
Dealer website or microsite	❓ not public	Marketing	OEM-dependent. 🌐 Hyundai hosts dealer microsites on its own domain; Maruti dealers run theirs. Ask per brand
SMS campaigns	pass-through	Marketing	Thin, and 🌐 inside TRAI's DLT regime, unlike WhatsApp. Low priority
Marketplace lead resale	❓ no provider publishes	Sales Head	No. Opaque business you cannot enter
Agency retainers, creative, imaging	❓	Marketing	No. Labour that does not scale past two people
Insurance or finance commission	licence-gated	F&I manager	No. 📘 product-scope

The strongest evidence for a digital-presence add-on is primary, embarrassing and dated

🌐 Popular Vehicles' own annual report discloses that popularmaruti.com and popularnexa.com were malware-compromised in November 2025, with Google's index serving fake URLs, root-caused by their vendor to a WordPress theme "nearly ten years old."

🔎 That is a listed ₹6,381-crore dealer group. If their web presence is in that condition, the supply of competent dealer digital services in India is worse than any market-size report suggests — and it is an argument that can be made with a citation rather than an opinion.

5a · GST — every price on this page is exclusive of 18%

📘 Owner instruction 2026-08-23. Quote ex-GST, add 18% on the invoice, never present an inclusive number.

Line	GST treatment
Subscription — Foundation, Growth, Complete	18%, added to the quoted annual price
Additional Setu Cards at ₹5,000/yr	18%
Group Intelligence	18%
Onboarding and activation	18%
Managed campaigns, review hygiene	18%
WhatsApp platform fee	18%
WhatsApp messages themselves	⚠ Meta bills as an imported service, so 18% under reverse charge on the pass-through. Different mechanism, same rate

Ex-GST quoting is honest here rather than a presentational trick, and the reason is a sales point

🔎 A dealership is GST-registered and claims input tax credit on business software. So the 18% is recoverable — it is a cash-flow item in their books, not a cost. Which means:

1. Quoting ex-GST is the correct comparison, because ex-GST is what the purchase actually costs them.
2. ⚠ And it is worth saying out loud, because a dealer who mentally adds 18% as real cost is comparing ₹1,76,999 against a competitor's ex-GST number. "₹1,49,999 plus GST, which you claim back" removes an objection that only exists through omission.
3. Never quote inclusive. 🔎 An inclusive figure looks like a discount you have already given, and it makes the input-credit conversation impossible to have.

⚠ Two product consequences, so 18% does not become a literal in the code

• The rate is statutory and can change. It must be a configured value on the invoice/plan surface, never a hardcoded 18 in a branch or a template. 📘 Same discipline this repo already enforces for payment-provider rates, for the same reason: a rate baked into code is a rate nobody can find when it moves.
• Place of supply matters for interstate dealers. ⚠ CGST+SGST versus IGST is a function of the customer's state, and a group with outlets in several states may hold several GSTINs. 📘 platform_tax_identity is already time-keyed for the platform's own side; the customer side needs a GSTIN per billed entity before the first invoice, which is a billing-surface requirement rather than a pricing one.

5b · Marketplace visibility — a paid add-on, and NOT sellable yet

📘 Owner instruction 2026-08-23: marketplace visibility, featured placement and recommendations are paid add-ons rather than bundled into subscriptions. 🔎 The packaging instinct is right. The prerequisite is not met, and one of the three collides with a refusal this section already made.

⚠ THIS CONTRADICTS AN EXISTING REFUSAL, AND THE REFUSAL WAS SPECIFIC

📘 product-scope §5 refuses "Advertising or profiled placement" with the reason: "Needs a consumer base that will not exist, and brings DPDP profiling duties." 📘 And ADR-0004's promo_slot fails closed forever pending a compliance_profile column that does not exist, precisely so a compliance-restricted vertical cannot carry a placement it may not legally carry.

So this cannot simply be added to the price book. It has to be split, because the three things named are not one product.

The three things, separated — because they have different answers

	What it is	Verdict	Why
(a) Marketplace LISTING — the dealer and their vehicles are present and findable	Presence	⚠ INCLUDE it in every plan. Do not charge	🔎 An empty marketplace is worth nothing to a buyer and nothing to a dealer. Charging for presence suppresses the supply side that makes the marketplace worth anything at all. Charging for the input to your own network effect is backwards while the network is being built
(b) FEATURED placement — paid priority in browse and search results	Advertising	Paid add-on. Priced, not built, not sold	Legitimate for this vertical and it needs measured consumer traffic to be a real product. 📘 Mechanically it is promo_slot, which is inert platform-wide
(c) RECOMMENDATIONS	Depends entirely on one word	Split it. Non-personalised: fine, later. Personalised: refuse for now	📘 "Vendors near you" or "most viewed this week" carries no profiling duty. Ranking by an individual's inferred interests is profiling under DPDP, and 📘 personalised recommendation is already deferred in the consumer scope

⚠ The one thing that must not happen: selling placement before there is traffic

🔎 Every other unvalidated item in this plan risks wasted effort. This one risks taking money for something we cannot deliver — a dealer who pays for featured placement in a marketplace nobody visits has been sold an impression count of roughly zero. That is a different category of exposure from a missed roadmap date, and it is the first item in the whole vertical with that property.

So the gate is a measurement, not a date:

No visibility add-on is quotable until the marketplace records a measured monthly consumer session count for the dealer's own city, shown to the dealer. Price it against that number, never as a flat fee.

⚠ 📘 And note the measurement itself is currently broken: the public card's analytics beacon 404s (QRS-734), so "we can tell you how many people saw you" is false until that is fixed. It is Wave 1's first item for exactly this reason. QRS-849.

What to do with it commercially in the meantime, which is not "nothing"

🔎 Sell the listing as included value, and name the add-on as a dated roadmap item with no price."Your outlet and your stock are listed at no extra cost. Featured placement will be available when we can show you the traffic it would buy." That is a stronger position than a price list, because it is the one claim in the pitch that visibly refuses to overpromise — and 🔎 it converts the marketplace from a promise into a reason to be an early customer, since early listings accumulate the history that later ranking uses.

5c · Google Business Profile portal — and why OUR version is different

📘 Owner direction 2026-08-23. 🔎 This is the strongest add-on idea in the thread, and the reason is not the review management — it is the trigger.

The differentiated part, in one sentence

🌐 Podium reached $100M revenue by 2019 from a 2014 founding and a $3B valuation in 2021, on review generation and messaging, and India has no equivalent incumbent. ⚠ This page previously said "$60M ARR in four years" — wrong, and it understated the category (QRS-857). 🌐 Comparable tooling runs $18-49/location/month (Synup) or ₹2,000-6,000/month (SocialPilot).

Every review tool has to GUESS when to ask. We know.

🔎 A generic tool fires a review request on a schedule or a manual trigger. The Setu Card recorded the interaction — this customer, this rep, this test drive, finished twelve minutes ago. Asking at that moment, naming the rep the customer actually met, is a materially different product from a bulk SMS blast, and it is only possible because the attribution spine exists underneath it.

⚠ 📘 Note what this does to the product argument overall: product-scope rates the review funnel "copyable: high" and says bundle it, do not lead with it. That still holds for review management. The trigger is not copyable without the card layer.

⚠⚠ The compliance trap, and the owner's example walks straight into it

Selecting only happy customers for a review request is REVIEW GATING, and Google prohibits it

📘 The direction reads: "Positive customer experience can lead into a Google review request." 🌐 Google's review policy prohibits review gating — soliciting reviews selectively based on how positive the feedback is expected to be. Building the funnel that way risks the dealer's Google Business Profile, which is a far more valuable asset than anything we sell them.

❓ Re-read the current policy wording before building. I have been wrong about a platform policy in this section before (the reported Meta repricing), so this is marked as needing verification rather than asserted.

The compliant design, which is barely more work and is a better product:

	Non-compliant	Compliant
Who gets the review link	Only customers who rated well	Everyone
What the internal rating does	Decides whether to ask	Routes a service-recovery task to a manager, in addition
What the dealer gets	A filtered review stream and a policy risk	Reviews from everyone, plus a queue of unhappy customers to fix

🔎 The second column is worth more to the dealer: an unhappy customer routed to a manager within the hour is retention, and it is exactly the kind of proactive action item this platform is supposed to produce.

What ships in which tier, and what stays a paid service

Capability	Tier	Note
Review monitoring and alerts	Foundation	Read-only. Cheap, and it makes the portal visible from day one
Business-information management	Growth	Hours, contact, photos, pushed from the outlet's Setu Card so it is entered once
Review response from inside QR Setu	Growth
Card-triggered review requests + feedback routing	Growth	⚠ The differentiated capability. Needs the interaction spine, so it cannot precede Wave 1
Review analytics, per-outlet and per-rep	Growth	🔎 Per-rep review attribution is a thing no review tool can do
Multi-outlet review rollup and comparison	Group Intelligence	One GBP location per outlet maps cleanly onto the workspace tree
We write the responses	Paid add-on, ₹59,999/yr	Labour, capped like every other managed service

⚠ Two build prerequisites that are not ours to schedule

• 🌐 Google Business Profile API access is approval-gated, not self-serve: a Cloud project plus a review process. ❓ Neither the lead time nor the current criteria have been verified. Establish access before promising a date, because this is the second capability in the vertical whose timeline depends on another company's queue.
• 📘 A dealer may not have management rights on their own listing. Many Indian dealer listings were created by the OEM or an agency. Onboarding has to include a claim-and-verify step, and 🔎 for some dealers that will be the single most valuable thing we do for them — and it takes days, not minutes.

5d · Physical QR standees — the bridge, and the best adoption mechanic in the plan

📘 Owner direction 2026-08-23: standees throughout the dealership at ₹225 + 18% GST each, charged separately from the subscription.

⚠ THIS SOLVES THE SINGLE BIGGEST RISK IN THE VERTICAL, AND I HAD NOT SEEN IT

📘 Value proposition §6 names the central problem: the person who benefits is not the person who must change. Attribution needs the rep to hand over a card instead of a phone number — the lowest-paid, highest-churn staff (🌐 29.53% attrition) changing a habit so their manager can measure them.

A standee requires no behaviour change from anyone. It sits on a table and the customer scans it.

🔎 So standees generate interactions, catalogue views, enquiries and feedback without any staff member doing anything differently — which means the data the whole product depends on starts accumulating on day one instead of after a behaviour-change programme. On the behaviour-change ranking that reordered the roadmap, a standee scores better than the visitor register, because reception at least has to type.

⚠ And it makes the subscription tangible. A dealer paying ₹1,49,999 for software they cannot see is a renewal risk; a dealer who walks past twenty QR Setu standees every morning is not asking what they pay for.

What makes a standee worth ₹225 rather than a printed sticker

⚠ Each standee must carry its OWN tracking code, or you are selling printing

🔎 A standee scan is unattributed by construction — nobody handed it over, so there is no rep to credit. That is fine, provided the standee itself is the unit of attribution. Each one needs a distinct code resolving to the outlet's card plus its own placement identity, so the dealer learns:

"The waiting-area standee was scanned 47 times last week and produced 6 enquiries. The one on the test-drive desk was scanned twice."

That is an analytics product and it justifies a per-unit price. A QR code that just opens the card is printing, and printing is not worth ₹225. ⚠ It also means placement is a first-class data field (reception · waiting area · table · vehicle · sales desk · test-drive bay), not a label on a box. QRS-852.

⚠⚠ The hard prerequisite, and it is currently failing

DO NOT SELL A STANDEE UNTIL THE ORG CARD RENDERS IMAGES

🧮 Measured and stated in this repo's own operating manual: there is no public media URL anywhere in the codebase. get_public_catalogue projects storage_key rather than a URL, mediaUrl.ts fails closed and returns null until a base is configured at boot, and nothing configures one because no bucket is provisioned.

So every image on the public Setu Card is unresolvable today. A customer who scans a standee, expecting the vehicle photos and videos the owner's own description promises, gets text and broken-image icons — and they read that as the dealership being broken, not as a missing feature. 🔎 For a showroom standee that is worse than no standee at all.

📘 Add to that QRS-734: the card-view beacon 404s, so "47 scans last week" cannot be reported either. Two hard prerequisites, both Wave 1, both already tracked. Standees are the right idea and they are not sellable before the card is worth scanning.

The physical-goods reality, which is a different business from software

Price	📘 ₹225 + 18% GST per 6-inch standee
Fulfilment	🔎 Print on demand through a Pune vendor. Never hold stock. Inventory, damage and returns are a two-person team's worst use of attention
Margin	🔎 Thin per unit at small volumes once print and courier are paid. The money is in quantity — an outlet wanting reception, waiting area, ten tables, twenty vehicles and the test-drive bay is 30+ units
⚠ GST classification	❓ 18% is assumed, not verified. A printed acrylic or foam display is goods, not a service, and its HSN classification may not attract 18%. 📘 This repo's standing rule is never to assume a rate — have a CA confirm the HSN before the first invoice
Replacement	🔎 Standees get damaged, moved and taken home. Expect reorders, and treat that as a feature of the line rather than a nuisance

💡 A growth mechanic worth deciding deliberately rather than discovering later

🔎 Every standee is a surface a customer scans, so the org card it opens is the platform's largest organic-discovery surface. A discreet "powered by QR Setu" on the card turns each dealership into a distribution channel — 📘 exactly the self-propagating loop the strategy section's virality page cares about.

⚠ But it is the dealer's premium brand surface, so it must be their choice, not a default. Offer it as a priced option: standees at ₹199 with QR Setu attribution, ₹225 white-label. The dealer picks, the trade is explicit, and neither party is surprised. 🔎 The ₹26 is irrelevant as revenue — it exists to make the choice legible.

6 · Break-even the dealer computes, never an ROI we project

Never present revenue or conversion improvement as an outcome

📘 A standing constraint in this repo and a standing instruction from the owner. 🔎 It is also the stronger sales position: a break-even threshold is arithmetic, a projection is a claim the dealer can reject.

🌐 Using the dealer's own service ticket of ₹20,000 at the verified 41.2% service gross margin = ₹8,240 of contribution per additional retained job:

Package	Annual price	Recovered by	Per month
Foundation	₹79,999	10 additional retained service jobs a year	0.8
Growth	₹1,49,999	18 additional retained service jobs a year	1.5
Complete	₹2,99,999	36 additional retained service jobs a year	3.0

🔎 Hand the dealer their own numbers and let them do the division. We assert nothing about whether they will get 18 jobs; we state what 18 jobs would be worth against what the platform costs, and they decide whether their own reminder discipline is currently better or worse than that. 🌐 With a 20:1 serviced-to-sold ratio, 18 jobs is a small fraction of one month's service throughput at a mid-size outlet — which is the point, and it is a statement about magnitude, not a forecast.

⚠ Do not run the same calculation on the sales side. Attribution changes who gets credit for a lead; it does not create demand, and 📘 D2 forbids selling lead generation. A break-even argument built on incremental car sales would be exactly the claim the product cannot support.

7 · Onboarding is a priced engagement with a contract, not a courtesy

🔎 The owner's adoption journey is correct and it has a commercial consequence the old fee did not reflect.

Package	Realistic effort	Cost at ₹1,500/h	Fee
Foundation	20-40 h	₹30,000-60,000	₹59,999
Growth	40-70 h	₹60,000-1,05,000	₹1,24,999
Complete	80-150 h	₹1,20,000-2,25,000	₹2,49,999

⚠ The old ₹25,000 low end was below its own cost, and this page published it

QRS-847. Org setup, employee setup, card issuance, hierarchy configuration, catalogue population, template authoring and two or three on-site training sessions is 20-40 hours minimum for a small outlet. A fee below cost is a discount you did not decide to give, and it also attracts the buyer who will not do the data work — which is the buyer who churns.

The activation contract, which is what makes annual pricing honest

Mechanic	Rule	Why
The annual term starts at GO-LIVE, not at signature	Written into the contract	Answers the owner's real concern directly. A dealer who spends six weeks on setup should not spend it out of their paid year
Activation is 45 days, with named milestones	Cards issued · hierarchy configured · catalogue populated · reception trained · first 50 interactions logged	An open-ended activation is how a pilot dies without anyone cancelling it
Day-45 health check	🔎 If fewer than 60% of issued cards are active, we intervene	📘 Proactive by default. Waiting for renewal to discover non-adoption is the reactive failure this platform's own product principle forbids
Onboarding waived on a 36-month term	Instead of a price cut	Converts a fee objection into a longer commitment, which is what the back-loaded value curve needs anyway

8 · Discounting policy, written before the first negotiation

Never discount the price. Discount the term or the scope

🔎 Dealers in one city talk to each other, and 🌐 FADA chapters are organised precisely so they do. A price cut on logo #3 becomes the market price permanently, and it is unrecoverable — you cannot raise a price a buyer has already been quoted.

The dealer says	Do not	Do
"₹1.5 lakh is too much"	Cut to ₹1.2 lakh	Move them to Foundation. Fewer capabilities for less money is a coherent offer; the same capabilities for less money is a price cut
"Give me a discount for 3 years"	Cut the annual figure	Same annual price, onboarding waived, price locked for the term
"My other outlet too, so give me a better rate"	Improvise a bundle	Outlet 2 and 3 at list, and Group Intelligence becomes available at 3 — the volume reward is a capability, not a rate
"Your competitor is ₹29,000"	Match it	₹29,000 is 🌐 AutoBooom at ten users. At their headcount it is ₹77,000-1,01,000 for a narrower product (§3)
"Let me try it monthly first"	Agree	A paid 3-month pilot at ₹25,000 that converts at list. Time-boxed, named, and written as expiring

Founding-customer pricing must be time-boxed, named, and expire in writing

🔎 The first five dealers will need a price concession, because we have no reference customer — the single largest objection in value proposition §2. Give it as "Founding Dealer, first 5, expires 31 March 2027", in the contract, with the renewal price stated. An unnamed discount is indistinguishable from the real price, and it leaks to every prospect in the city.

9 · Willingness to pay, close rate, and who to disqualify

The 50-60% close-rate target is a qualification artefact, not a pricing outcome

🔎 A close rate is a ratio whose denominator you choose. Qualify hard enough and 60% is easy on a tiny denominator; talk to everyone and 10% is good. Pricing a product to hit a close-rate target is backwards.

The number that matters is closes per month, and it is capacity-bound. And the diagnostic runs the other way from intuition: 🔎 if you are closing 60% of everyone you meet in year one, with no reference customer, you are too cheap.

Honest expectation: 🔎 15-25% of hard-qualified prospects in year one, rising toward 40-50% once three named Pune dealers can be cited. The reference customer moves this number far more than the price does.

The qualification bar — six signals, and all six should be present

Signal	Why it predicts a close
3+ outlets under one owner	📘 D0. The only buyer with the cross-brand problem, and one signature covers three outlets of revenue
A GM or DP who runs a recurring review meeting	The visibility product needs a consumer. Without this person, nobody wants the dashboards
Already pays for telephony, a CRM or a lead tool	Proves a software budget line exists and reveals its size. The single best predictor in the list
Reception logs walk-ins on paper today	The zero-behaviour-change wedge has a target (value proposition §6)
Dealer Principal reachable in one hop	🔎 At ₹1.5-7 lakh this is still a discretionary signature. A two-hop path turns it into a committee and a 6-month cycle
Runs their own domain and marketing	Not fully OEM-templated, so digital presence is theirs to improve and theirs to buy

Disqualify deliberately, and accept the lost logos

Segment	Why not	Cost of refusing
Independent workshops	🌐 RAMP caps the category at ₹42,000/yr. Below our floor, and the highest support load per rupee	Real. It is a large, reachable segment
Independent used-car dealers	🌐 Cars24, CarTrade and Mahindra First Choice give dealer tooling away as customer-acquisition cost. You cannot out-price free	Real, and it looked like the most open segment
Single-outlet dealers with no GM	Nobody in the building consumes the visibility half. Foundation only, and never a group upgrade	Acceptable. Foundation is still profitable
Groups whose OEM mandates a specific CRM	You will be replaced by a mandate, not by a competitor	Acceptable, and it is discoverable in one question

⚠ Abandoning the low end is a deliberate loss and it must be stated as one

🔎 The previous version of this page reached 25-40 logos by assuming a ₹36,000 wedge would sell to workshops, used-car dealers and small outlets. Withdrawing the ₹30,000 tier withdraws that path.

The trade is favourable and it is not free: those logos would have consumed the same support hours as a ₹1.5 lakh account while paying a quarter as much, and 🔎 support hours are the binding constraint on the whole business (§2). We are trading reachable logos for a serviceable book. Say so plainly rather than presenting a smaller number as an improvement.

10 · The ₹2 crore question, revised

200 dealerships at ₹1 lakh by August 2027 fails FIVE independent tests, and repricing fixes only one of them

🔎 The previous version listed four. Cost-to-serve adds a fifth, and it is the one nobody plans for, because it constrains serving rather than selling.

(a) The denominator · (b) Pune · (c) Sales capacity · (d) The product does not exist

🌐 FADA: "over 15,000 Automobile Dealerships having over 30,000 dealership outlets" — 200 is 1.33% of the entire organised national dealer universe, in year one, from Pune, with two people. 🌐 INFOMAN, at 25 years old, reports 600 customers. 🌐 Pune metro holds ~215-403 dealer entities, of which ~45-70 are franchised PV, so 200 is 3-4x the entire local franchised universe. 200 closes over 50 weeks is 4 per week with zero ramp, which is 🔎 a 3-5 person inside-sales function. And 🧮 five of seven declared primitives have zero tables, RBAC is 0% built, and there is no write path to organizations anywhere in the product.

(e) Cost to serve — the fifth test, and it is new

Book	Support at 3 h/outlet/month	Pure-support headcount
30 outlets	90 h/mo	0.6 FTE
81 outlets — 27 groups of 3, the ₹2 Cr group path	243 h/mo	1.5 FTE
200 outlets — as briefed	600 h/mo	3.8 FTE

🔎 The 200-logo plan needs roughly four support people, and it needs them independently of the 3-5 salespeople test (c) already demanded. Neither existed in the plan. ⚠ And note that the good path is not free either: 27 groups still needs a support hire. Any path to ₹2 Cr requires hiring; the difference is whether it is one person or nine.

What the new book does to the arithmetic — this is the prize

Customer shape	Software	Group Intelligence	WhatsApp	Annual ARPU
Small outlet, Foundation	75,000	—	18,000	₹93,000
Standard franchise outlet, Growth	1,50,000	—	18,000	₹1,68,000
Large outlet, Complete	3,00,000	—	18,000	₹3,18,000
3-outlet group, Growth + GI	4,50,000	2,40,000	54,000	₹7,44,000
6-outlet group, Growth + GI	9,00,000	4,80,000	1,08,000	₹14,88,000
15-outlet group	22,50,000	12,00,000	2,70,000	₹37,20,000

Path to ₹2 Cr ARR	Logos	Signatures/week	Assessment
As briefed — 200 single dealerships at ₹1 lakh	200	4.00	Fails all five tests
Single Growth outlets only	119	2.38	Fails on capacity and support
27 groups of 3 outlets	27	0.54	Reachable, with 2-3 sales hires and 1 support hire
13 groups of 6 outlets	13	0.27	The best shape in the study. 🌐 One listed group alone runs 208 showrooms
One OEM network deal	1	~0	Lottery ticket, and ⚠ mutually exclusive with the dealer-trust position

⚠ The repricing INVERTS this section's own harshest criticism of itself, and that must be said

📘 The previous version of this page carried a danger callout arguing that this brief was "the same goal made strictly harder on every axis" than the dealership motion the repo rejected in August. That plan was 30 seats x Rs 1,000/month = Rs 3.6 lakh ACV, about 35 deals for Rs 1.25 Cr; the brief was Rs 1 lakh ACV and 200 deals for Rs 2 Cr - 5.7x more signatures for 3.6x less per account.

🧮 At the corrected book, Rs 1.25 Cr needs ~17 deals at Rs 7.44 lakh. That is fewer signatures AND more revenue per account than the plan that was rejected for being too hard. The arithmetic objection is gone.

⚠ It does not un-reject the plan, and reading it that way would be the expensive mistake: the org-admin surface is still unbuilt, seats and billing still do not exist, and there is still no reference customer. What changed is that the shape is now defensible where it previously was not - which means the remaining objections are execution objections, and those are answerable.

This is what repricing actually buys, and it is worth more than the extra margin

🔎 ₹2 crore drops from 200 signatures to 13-27 — a 7-15x reduction in signatures, onboardings and support load, for the same revenue. 0.27-0.54 signatures a week is a number two people can plausibly reach once they have references. 4 a week never was.

⚠ But it does not arrive by August 2027, because Group Intelligence needs Wave 3 and Wave 3 lands in August 2027.

What August 2027 should commit to

Model	Assumption	Logos	ARR	One-time
Adversarial	15% win rate, Foundation only, no reference customer until Q2	4-8	₹4-8 L	₹2-5 L
Base — recommended commitment	Selling from Dec 2026, 1 logo/month, mix of Foundation and early Growth	8-15	₹12-25 L	₹6-15 L
Optimistic	Two 3-outlet groups convert early to Growth	15-20	₹30-45 L	₹15-25 L
₹2 Cr arrives	13-27 groups, Wave 3 complete, 2-3 sales hires from Q3 2027 + 1 support hire	13-27	₹2 Cr	FY 2028-29

⚠ The recommended logo count went DOWN and the ARR band overlaps the old one. Read that honestly

The previous commitment was 20-30 logos / ₹20-35 lakh, assuming the ₹36,000 wedge. The new one is 8-15 logos / ₹12-25 lakh, plus ₹6-15 lakh of onboarding cash the old model was charging below cost for.

🔎 Total cash is comparable. The book is not. 8-15 accounts at ₹1.2-1.7 lakh are serviceable by two people and every one is profitable; 25-40 accounts at ₹36,000-90,000 are serviceable by neither, and the cheapest of them lose money on every support call. The right question is not which number is bigger but which book survives to 2028.

Pre-committed revision rule — restated at the corrected price

If fewer than 3 of 5 pilot dealers pay a real invoice at ₹25,000 for three months by 31 October 2026, the ₹2 crore target is restated to ₹35-50 lakh of ARR by August 2028, and this vertical becomes a single bet rather than the company's direction.

🔎 Note that the ₹25,000-for-three-months pilot already annualises to ₹1,00,000/yr — above Foundation and below Growth. The company's own validation experiment was already testing the corrected price while this page listed ₹30,000, which is independent evidence from inside the documents that the floor was wrong.

📘 A target without a revision rule is a number that gets missed and then quietly forgotten, which this repo measures at roughly zero completion for deferred reconciliation (QRS-180).

11 · Meta India rate card, for the pass-through clause

🌐 The demand evidence is the best in this study: Meta/Kantar via FADA found 48% of new Indian auto buyers contacted dealerships via WhatsApp and 47% want service reminders by messaging. And WhatsApp sits outside TRAI's TCCCPR/DLT regime — no DLT headers, no telco template registration, no NDNC scrubbing.

🌐 Cards effective 1 July 2026, ex-GST, charged on delivery only. Add 18% GST on Meta as an imported service.

Category	Rate	Volume discount?
Marketing	₹0.8631	Never. Confirmed in Meta's own documentation
Utility	₹0.1150	Yes, but pooled per business portfolio
Authentication	₹0.1150	Yes, same pooling
Service	Free	—

Why there is no arbitrage, stated precisely because it is invisible in partner marketing

🌐 Volume discounts apply to utility and authentication only, never marketing, and they pool at the business portfolio level. If each dealer owns their own account and portfolio — which is what dealer-owned branding requires — volume never pools, every dealer sits in the lowest tier permanently, and you buy marketing messages at exactly the ₹0.8631 any dealer can buy them for.

There is no wholesale price. There is only a markup the dealer can remove by going direct. 📘 This independently confirms ADR-0030's finding.

	A · Platform sends — CHOSEN NOW	B · Dealer owns it — refused	C · Solution Partner — the graduation
Whose account	Yours	The dealer's, Embedded Signup	The dealer's, your credit line attached
Meta bills	You	The dealer	You
Can you sell credits	Yes	No	Yes
Dealer branding	Name in the message body — sufficient for utility messages	Yes, verified	Yes, verified
Isolation	⚠ None — limits pool. Answered by a mandatory per-workspace send cap, not by choosing B	Full	Full
Friction	Low	Dealer completes verification and attaches a card	🌐 "A lengthy process", irreversible once attached
Path onward	A → C is clean; we own the accounts	⚠ B → C recreates every account and forfeits its quality history. This is why B is refused	Terminal, and that is fine

Two corrections a fact-checker forced, both material

• The widely-reported "1 October 2026 repricing" is reported, not confirmed. 🌐 Meta's own pricing documentation, read 22 Aug 2026, still states "Utility template messages sent within an open customer service window are free" with no deprecation notice. ❓ Price with a pass-through clause anyway, because a fixed bundle would carry the risk.
• Messaging margin is not trivial at scale; the first arithmetic was 10x out. At ~₹959 per marketing-active dealer per month, 750 dealers is ₹86.3 lakh a year, not ₹8.6 lakh. 🔎 Which is why the recharge line is worth keeping even though it is immaterial today — it becomes material on exactly the path Option A leaves open.

🌐 Two operational limits to design around now: a Meta Tech Provider onboards 10 new business customers per rolling 7 days, rising to 200 only after full verification — start verification before you need it. And marketing frequency is capped at roughly two marketing templates per user per 24 hours across all businesses, which puts a hard ceiling on any campaign-blast pitch.

❓ The favourable regulatory position may be temporary: TRAI's April 2026 draft spam rules propose bringing OTT messaging into scope. Treat "no DLT" as a current advantage, never a moat.

11a · The seven revenue lines, kept separate on purpose

📘 Owner direction 2026-08-23. 🔎 Separating them is not presentational: each line has a different cost shape, and bundling any two hides the one that loses money.

Line	Shape	Our cost	Why it must stay separate
1 · Annual SaaS subscription	Fixed, annual, per outlet	🔎 Support hours, ~50-64% margin	The anchor. Everything else attaches to it
2 · Additional Setu Cards	₹4,999/card/yr	~zero marginal	Near-pure contribution, and the tier ladder's pressure valve
3 · Additional access — management logins, storage	₹2,499/login/yr · ₹999/10 GB	Support and storage, real	⚠ The line that did not exist until today (§4b)
4 · Physical QR standees	₹225 + GST per unit	Print + courier — a goods cost	⚠ A goods business inside a SaaS company. Bundling it would put a physical COGS inside a software margin
5 · WhatsApp and marketing usage	Prepaid recharge	🌐 Paid to Meta per message, variable and unpredictable	🌐 A festival campaign can be 50x a normal month. Bundling transfers that variance onto us for no margin
6 · Google and digital services	Included from Growth; managed work priced	Labour	Capped at ~15 accounts. It does not scale past two people and must never be sold as though it does
7 · QR Setu AI	Packages + action recharge	🧮 Real COGS, ~₹0.44 per action	⚠ The only line with a per-use marginal cost that scales with success. Metered in business units, never tokens (AI strategy)

What this structure buys, beyond tidiness

🔎 It makes the loss-makers visible. Line 4 carries a physical cost of goods and line 6 carries labour; both would be invisible inside a bundled annual price, and both are exactly the kind of line that quietly turns a 50% software margin into a 20% blended one. 📘 This section already withdrew one loss-making SKU it had published (§0) — the six-line split is how that stops happening by construction rather than by audit.

12 · What changed on 2026-08-23, in one table

	Before	Now	Why
Billing period shown	₹2,500/mo and ₹30,000/yr	Annual only, everywhere	Back-loaded value curve; monthly churns before the product works
Floor	₹30,000/yr	₹79,999/yr	🔎 Loss-making at 2 h/month support; 🧮 2.6x below the cheapest verified competitor at real headcount
Volume tier	₹72,000/yr	₹1,49,999/yr	🌐 AutoBooom is ₹77,000-1,01,000 at 30-40 users for a narrower product
Top tier	₹1,08,000/yr	₹2,99,999/yr	🌐 Per-user comparables land at ₹2.88-9.07 lakh at the same headcount
Pricing unit	per card	per outlet, cards as fair-use ceiling	A per-card meter reintroduces the per-seat negotiation it was meant to avoid
Group Intelligence	₹1,50,000 flat per group	₹79,999 per outlet, min 3	Flat pricing inverted the value and taxed small groups
Onboarding	₹25,000-75,000	₹59,999-2,49,999	🔎 The low end was below its own cost
Low-end segments	wedge strategy	Deliberately abandoned	Cannot out-price free; worst support-hours-per-rupee
WhatsApp	contradicted across four pages	Option A now, C as graduation, B refused	Utility messages need only body-branding, and B → C forfeits every account's quality history. Recharges stay a metered add-on
₹2 Cr shape	70 groups	13-27 groups	Higher ARPU per logo, same revenue, 7-15x fewer signatures
Aug 2027 commitment	20-30 logos / ₹20-35 L	8-15 logos / ₹12-25 L + ₹6-15 L one-time	Honest consequence of abandoning the low end

Second revision, same day — five owner directions

	Before	Now	Why
Card ceilings	15 / 40 / open-ended	10 / 30 / 75 included, hard caps 24 / 60 / 100, and the outlet's own card counts inside the limit	📘 Owner: limits should be controlled, and no plan open-ended. ⚠ Pushes the standard 15-40-person outlet onto Growth, deliberately. Each cap lands on the next tier's base price (why)
Additional cards	not priced	₹4,999/card/yr + GST	📘 Owner rate, and 🧮 it is the only round number in the ₹4,000-₹5,000 window the constraints allow (why)
Mobile + desktop	unstated	Headline on every plan	📘 One universal codebase (ADR-0011) reaches Android, iOS and browser, so it costs nothing and answers 🌐 the named gap in every OEM DMS critique. ⚠ Sell "works on desktop", not the 6-of-20-built designed console
Marketplace visibility	not in the model	Split three ways: listing INCLUDED, featured placement priced-but-not-sellable, personalised recommendation refused	📘 Collides with an existing refusal, and 🔎 selling placement before traffic exists is the first item in this vertical that would take money for something undeliverable (§5b)
GST	absent	Every price ex-GST, 18% added	📘 Owner instruction. 🔎 Recoverable as input credit for a GST-registered dealer, so ex-GST is the honest comparison — and worth saying, since a dealer who mentally adds 18% is comparing the wrong two numbers
"Rooftop"	the pricing unit, 71 times, never defined	"Outlet", defined, with the outlet-versus-location test	🌐 FADA and the audited filings both say outlet. US vendor jargon read as expertise while writing and as confusion while reading (§4a)

Third revision, same day — psychological pricing, access caps, Google, standees

	Before	Now	Why
All prices	round numbers	₹79,999 / ₹1,49,999 / ₹2,99,999, cards ₹4,999, GI ₹79,999/outlet, standees ₹225	📘 Owner: use psychologically stronger price points. 🧮 Verified that the cap identities survive — they hold to within ₹14-30, so the structure is intact (§4)
Foundation hard cap	25 cards	24 cards	🧮 The crossover moved: ₹79,999 + 14 × ₹4,999 = ₹1,49,985, which is ₹14 from Growth's base
Management / admin logins	⚠ uncapped	5 / 12 / 25, caps 10 / 20 / 45, extra at ₹2,499/yr	📘 Owner found the hole. 🔎 An unlimited login allowance gave away the exact thing each tier gates on (QRS-851)
Media storage	unstated	5 / 25 / 100 GB, soft limit	🔎 The only infrastructure line that scales without bound, and the owner wants video on the org card
Other usage limits	—	Deliberately none	🔎 Never meter walk-ins, leads, catalogue or scans: those are the behaviours the product exists to create (§4c)
Google Business portal	a managed add-on	Included from Growth, managed responses priced	📘 Owner direction. 🔎 The differentiator is the trigger, not the review management — ⚠ and the owner's "ask happy customers" flow is review gating, which Google prohibits (§5c · QRS-853)
Physical QR standees	did not exist	₹225 + GST, per-standee tracking codes	📘 Owner direction, and 🔎 it solves the biggest adoption risk in the vertical — a standee needs no behaviour change from anyone. ⚠ Not sellable until the org card renders images (§5d)
Revenue structure	one price plus add-ons	Six separate lines	🔎 Each has a different cost shape; bundling hides the two that carry goods and labour costs (§11a)

Related

• Value proposition — why a dealer would and would not buy this at all
• Go to market — the single roadmap these packages become sellable against
• Market landscape — the price benchmarks
• 📘 Strategy: revenue model — the platform-level target this sits inside