Empanelment & Routes
Multi-State Empanelment Strategy for Startups
· 9 minute read
A twelve-person firm cannot honestly hold twelve state panels. Choose two corridors, one central channel, and a renewal factory. Everything else is a brochure wall.
A board deck in HSR Layout showed eighteen coloured states 'in pipeline for empanelment'. The company had fourteen people, one DSC token, and a single on-prem engineer who already commuted to a Bengaluru SDC. They won a list in a far state that needed a local office and a two-day site visit on ten days' notice. They missed the visit. They still left the state coloured in. That colour was a lie to themselves.
Multi-state empanelment is a capital-allocation problem. Each RFE costs weeks, each renewal costs a quarter-day forever, each work order costs presence. Startups treat lists like Pokémon. Cells treat no-shows as a reason to distrust the next startup.
This playbook is how to pick a small set — for example a Karnataka–Tamil Nadu southern corridor, or a Maharashtra–Gujarat western corridor, plus Telangana if your product and people already live there — without insulting every other state. Presence later is allowed. Fiction now is not.
Not legal advice. Incorporation and GST in multiple states have tax consequences your CA owns.
Score states, do not colour them
For each candidate state, score four things from 1 to 5. One: named buyers who have talked to you in the last six months, not a summit handshake. Two: delivery capacity — language, a person who can be in the SDC this month, a partner you would actually sign. Three: instrument quality — is there a live RFE or GeM habit, or only a speech. Four: pile-B procurement perks you actually qualify for (see the state-policy method). Add a fifth if you like: payment reputation, from other vendors, quietly.
Bid the top two. Watch the third. Ignore the rest until a buyer in that rest sends a dated problem. A dated problem beats a pretty RFE in a state where you cannot staff a visit.
Corridors beat sprinkles. Two neighbouring states share language overlap, SI pools, and travel. Eighteen sprinkles share nothing but renewal dates.
| State | Buyers | Delivery | Instrument | Perk fit | Play |
|---|---|---|---|---|---|
| Karnataka | 4 — two live conversations | 5 — office and SDC habit | 4 — cell exists, RFEs happen | 3 — if you meet local rules | Primary |
| Telangana | 3 — one serious directorate | 3 — people can go, not live | 3 — institutions strong, panel not automatic | 2 — do not confuse T-Hub with a GR | Secondary if the directorate dates a pilot |
| Tamil Nadu | 2 — summit only | 3 — Chennai is staffable | 3 — StartupTN real, tender hook varies | 3 — check live G.O. | Watch |
| Gujarat | 2 | 2 — need a partner | 3 | 2 — sector lists matter | Only with a signed SI or a named PSU |
| Maharashtra | 4 — PSU and municipal noise | 3 — Pune possible, interiors hard | 4 — volume | 3 — need the GR, not the summit | Primary if cash can stand the cycle |
| A far state with a pretty RFE | 1 | 1 | 5 — RFE is live | 4 — on paper | Skip unless a buyer writes |
Capacity math a founder can do on paper
A serious empanelment bid is two to four focused weeks of a competent person, plus founder time, plus a CA. A renewal is two to five days a year. A work order is a multiple of that forever. If you have one bid person, you have perhaps four serious RFEs a year, not eighteen.
On-prem delivery is the binding constraint, not PDF production. An air-gapped agent in a state SDC needs a named engineer, a media protocol, and travel. If that engineer is already 70% on one state, a second state's 'win' is a customer you will fail.
SIs can extend geography. They cannot multiply your best people. See the SI guide. Count your people, then count the SI's people you have met, then stop colouring.
One central channel plus two states
Most product firms should hold a GeM presence they can actually fulfil, plus at most one central-ish pool (a NICSI-adjacent or ministry RFE if it is real), plus two states. That sentence is an opinion. It is how twelve-person firms stay honest.
DPIIT and Udyam travel across states as papers. They do not replace state enrolment where a perk requires it. Do not incorporate five subsidiaries to chase five reimbursements. That is how you buy compliance cost, not pipeline.
When a third state becomes real — a paid discovery, a letter from a secretary — add it then. Strategy is a sequence, not a map of India in brand colours.
Refuse logo panels
Some lists exist so a cell can say it has startups. There is no work-order path, no budget, no mini-bid template. The RFE is still a lot of work. Ask, before you bid: 'How did the last panel issue work in the last twelve months?' If they cannot describe a path, you are buying a logo. Logos do not pay GPUs.
If you already hold a logo panel, do not spend founder time on its conference. Spend it on the state that issues orders.
Objections you will hear — and what to do with them
These are the lines that stall the file. Answer them in the room, then put the answer in the note. A spoken answer without paper will be forgotten by the next officer.
Investors want a national footprint.
Give them a national story of the product — on-prem, Hindi, no training on customer data — and a sequenced go-to-state plan. A lie in eighteen colours will surface in diligence when renewal letters cannot be found.
RFEs are practice. Bid them all.
Practice on the corridor. Random far-state RFEs teach you that state's portal, not your product. The opportunity cost is the bid you needed at home.
A local director in every state is cheap.
A sleeping director is a compliance object. Cheap until it is not. Prefer travel and a real SI on a named account.
We will staff after we win the lists.
Cells now ask who will deliver. Empty org charts are a technical fail, not a later hiring plan.
A 30-day multi-state reset
If your deck has more than three coloured states and fewer than twenty people, run this.
- Week 1: list every panel, RFE, and 'pipeline state'. Attach a letter or mark it as fiction. Uncolour fiction.
- Week 2: score remaining states on the four (or five) axes. Pick two primaries and one watch.
- Week 3: assign people and SIs. Kill bids that have no owner. Calendar renewals for the keepers.
- Week 4: rewrite the board slide as a sequence with dates. Brief sales that new states need a dated buyer problem.
How this shows up in the file
Internal policy: 'We bid state empanelments only in [A] and [B], plus a named exception signed by the founder when a buyer in another state issues a dated problem and a delivery owner exists. We will not incorporate a subsidiary solely to harvest a brochure perk.'
Review the policy every six months. Sequence can grow. Fiction should not.
This article is informational field guidance for Indian public institutions and their vendors, not legal, tax, procurement or engineering advice. Confirm the live circular, RFE, GCC, GeM term, state G.O. and your counsel before you file anything. Incentives, ceilings and portal screens change.
How to run the route without confusing the letterhead
“Multi-State Empanelment Strategy for Startups” is a route problem. A P5 Startup/Vendor should know which legal person they are talking to — NIC, NICSI, a state IT corporation, iDEX, or a GeM buyer — and which paper that person can actually issue. Searching “multi state empanelment” is not the same as being on a panel that can receive a work order.
A twelve-person firm cannot honestly hold twelve state panels. Choose two corridors, one central channel, and a renewal factory. Everything else is a brochure wall. Empanelment letters are not purchase orders. DPIIT recognition is not a technical score. Reserved startup seats, if a notice writes them, are local to that notice. IndiaAI compute empanelment is not NICSI application-software empanelment.
Keep a warm evidence pack: CIN, GST, DPIIT, Udyam, financials, work-completion letters, architecture one-pager, DPA draft. Renewals are lost by people who treat the panel as a trophy.
- Screenshot the live RFE paragraph you are relying on, dated.
- Match the bidding entity name across every certificate.
- Do not mix iDEX, TDF, NICSI and GeM clocks on one tracker cell.
- Record the validity end date 90 days before it dies.
Close this loop before the next CAB
Put “Multi-State Empanelment Strategy for Startups” on the next change-advisory or bid-opening agenda as a single line item with an owner. If it cannot earn a line item, it will not earn a control. The owner should be a P5 Startup/Vendor, not “the vendor.”
Revisit the item when the model, the GeM term, the region, or the SI changes. “multi state empanelment” is not a one-time workshop. It is a watch item. Date the last check. Unsigned watch items are souvenirs.
Questions this usually raises
- How many state empanelments should a startup hold?
- As many as you can renew and deliver. For a dozen people, two states plus a central channel is a defensible default, not a law.
- Which state is best for AI vendors?
- The one with a named buyer and a staffable SDC. Brochure energy is not the same as purchase orders.
- Do we need a local company in each state we bid?
- Only if eligibility or tax reality requires it. Do not multiply CINs for colour.
- Should we follow every new state AI policy with a bid?
- Read it. Bid only if the score holds. Policies are pile A more often than they are a panel.
- Is GeM a substitute for state panels?
- It is a channel many state buyers can use. It is not a substitute when the state cell only draws from its own list. Hold GeM anyway if you can fulfil it.