Every founder who has built a physical product learns the same lesson at roughly the same moment: the prototype that worked on the bench is perhaps a fifth of the way to a product a customer can buy. Between the two lie a second and third design, field units that break in ways the lab never predicted, certification tests, a contract manufacturer who needs tooling paid for in advance, and inventory that has to be bought months before it is sold. A software company can recover from a bad release in a week. A hardware startup in India that orders the wrong tooling can lose a year and most of its seed round.
India is a better place to build hardware than it was a decade ago. Government data shows electronics production rising roughly six-fold in ten years, and new schemes now subsidise component manufacturing. But the policy support is aimed mostly at large manufacturers, and the capital market for early hardware remains thin. This article explains what has changed, what a young hardware company actually has to get through, and sets out a milestone and cash ladder in rupees with every assumption stated, so founders and investors can argue about the numbers rather than about the shape of the problem.
What has changed for a hardware startup in India
The scale of the shift is visible in the official figures. A PIB backgrounder of 3 February 2026 reports that electronics production grew from ₹1.9 lakh crore in 2014-15 to ₹11.3 lakh crore in 2024-25, and electronics exports from ₹38,000 crore to ₹3.27 lakh crore over the same period. Mobile phones drove much of it: production rose from ₹18,000 crore to ₹5.45 lakh crore, and the number of manufacturing units from two in 2014 to more than 300.
Three policy instruments sit behind this.
The production-linked incentive (PLI) schemes reward incremental sales by manufacturers who meet investment thresholds. The same PIB document puts the overall PLI outlay at ₹1.97 lakh crore across 14 sectors, and reports that the PLI for large-scale electronics manufacturing and IT hardware had attracted ₹13,107 crore of investment and ₹8.56 lakh crore of production by June 2025. Across all PLI schemes, a Lok Sabha reply of July 2026 reported ₹2.40 lakh crore of investment and 14.15 lakh jobs.
The Scheme for Promotion of Manufacturing of Electronic Components and Semiconductors (SPECS), run by MeitY, reimburses 25% of capital expenditure for eligible goods. As SS Rana's summary of the guidelines notes, eligible goods are grouped into seven sub-heads with minimum investment thresholds starting at ₹5 crore and rising to ₹1,000 crore.
The Electronics Component Manufacturing Scheme (ECMS) was notified on 8 April 2025 with an outlay of ₹22,919 crore, raised to ₹40,000 crore in the Union Budget 2026-27 according to PIB. By January 2026, 46 applications across 11 states had been approved, with ₹54,567 crore of investment.
For a startup, the honest reading is that these schemes help indirectly. They build the supplier base, contract manufacturers and component availability that a young company needs, and they make it more likely that a printed circuit board, an enclosure or a battery pack can be sourced domestically. They rarely fund the startup itself, because the thresholds assume a factory, not a team of twelve with a prototype. A startup that becomes a component maker may eventually qualify; one that designs a device and outsources production usually will not.
The capital gap hardware founders face
The investor side has not moved as fast as the factory side. The Nasscom-Zinnov Indian Tech Start-up Report 2025 counted more than 4,200 deeptech startups and $2.3 billion of deeptech funding in 2025, with AI taking 91% of deeptech funding. That leaves a small pool for robotics, devices, sensors, energy hardware and advanced materials. The IVCA Bharat DeepTech Report 2026, as reported by YourStory, found that investor participation drops sharply after seed.
Hardware also costs more before it proves anything. BCG and Hello Tomorrow's 2019 study of deep tech startups found that the average first prototype among biotech participants in the Hello Tomorrow Challenge cost almost $1.3 million (about ₹12.4 crore at ₹95 to the dollar), against about $200,000 (about ₹1.9 crore) for blockchain ventures. Hardware sits between those poles, and the gap between a working prototype and a shippable product is where most of the money goes.
The practical consequence is that hardware founders in India should plan for a longer bridge between seed and Series A than software founders, and should treat grants, customer-funded pilots and non-dilutive programmes as part of the capital stack rather than as extras. A separate PanScience Research piece maps which kinds of capital exist at each stage of technology readiness in India and where they run out.
Certification: the step most plans underestimate
Certification is where a product meets the state, and in India the route depends on what the product is.
For a wide range of electronics and IT goods, the relevant regime is the Bureau of Indian Standards Compulsory Registration Scheme (CRS), established under the Electronics and Information Technology Goods (Requirement for Compulsory Registration) Order, 2012, notified by MeitY in October 2012. Products on the list cannot be sold in India without registration. Nemko describes the scheme as based on self- declaration of conformity, supported by testing in a BIS-recognised laboratory, after which the manufacturer receives a registration number to use with the Standard Mark. TÜV Rheinland lists 63 electronics and IT products in scope, spanning audio and video equipment, IT equipment such as laptops and printers, UPS and inverters, LED products and certain batteries, plus five solar photovoltaic products notified by MNRE. TÜV notes that sample testing is required and factory inspection is not, and that a licence is granted initially for up to five years and may be renewed.
Telecom equipment follows a separate path. The Department of Telecommunications' Mandatory Testing and Certification of Telecommunication Equipment (MTCTE) regime, under Part XI of the Indian Telegraph Rules notified on 5 September 2017, requires equipment that connects to licensed networks to be tested and certified against essential requirements set by the Telecommunication Engineering Centre; the current procedure is version 3.0, issued in 2024. A connected device with a cellular module can therefore face both CRS and MTCTE, and products with radios may need further approvals.
Other sectors bring their own regulators: medical devices, automotive components, industrial equipment for hazardous areas and defence products each have separate regimes. A founder who discovers this at the end of development faces months of delay and redesign. The cheaper approach is to identify every applicable standard at the start of the engineering prototype stage, design to those standards, and run informal pre-compliance tests before formal submission.
Three points are worth building into any plan:
1. Certification is per model. A change of component, enclosure or firmware can trigger retesting. Freezing the design before submission matters. 2. Test laboratory slots are a queue. The time from booking to report, plus any failure and retest cycle, should be in the schedule as a range, not a single date. 3. Export markets bring their own certifications. A product meant for the United States or Europe should be designed to meet those standards from the start, because a second redesign is far more expensive than designing once to the stricter set.
Manufacturing: build, buy or partner
Very few hardware startups should own a factory. The usual sequence is in-house prototypes, then a contract manufacturer for pilot runs, then either continued outsourcing or, much later, owned capacity for the part of the process that carries the company's advantage.
The questions that decide the choice are straightforward. Does the company's advantage lie in how the product is made, or in what it does? If the former, some manufacturing must eventually be brought in-house, and the IP strategy should treat the process as a trade secret. A separate PanScience Research piece covers how to decide what to patent and what to keep secret. If the advantage lies in design, firmware, data or service, outsourcing is usually right, and the company should concentrate on design for manufacture, supplier qualification and quality control.
Three costs arrive in this phase that software founders rarely expect. Tooling, especially injection moulds for enclosures, is paid largely in advance and is specific to one design. Minimum order quantities for components can force the purchase of a year's supply to build a few hundred units. And the cash conversion cycle turns negative: components are paid for when ordered, products are sold weeks or months later, and enterprise or government buyers may pay later still.
A hardware milestone and cash ladder
The table below is the original contribution of this article. It sets out the milestones a typical connected-device or industrial hardware startup in India passes through between a lab result and repeatable sales, with an illustrative cash requirement for each stage. It is a model, not an observation, and every number is driven by the assumptions listed after the table.
Assumptions behind the ladder.
Team: four to six people in stage 1, rising to 10 to 14 by stage 3 and 18 to 25 by stage 6. Average fully loaded cost of ₹20 lakh to ₹28 lakh per person per year, reflecting a mix of senior hardware, firmware and software engineers with operations and sales staff added later. Bill of materials: ₹40,000 to ₹1.5 lakh per unit at prototype volumes, falling to ₹15,000 to ₹60,000 at pilot production volumes. Prototype units cost several times the production cost because of small-batch fabrication and hand assembly. Stage 3 includes travel, installation, spares and one round of redesign after field failures. Assume at least one significant failure mode is found in the field. Stage 4 includes testing fees, pre-compliance work, one retest after a failure, and the engineering time to freeze the design. It assumes certification for India only. Adding a major export market would raise this stage by a similar amount again. Stage 5 includes tooling (₹25 lakh to ₹80 lakh for enclosures and fixtures), components bought ahead of sale, and quality control at the contract manufacturer. It assumes the startup does not own a production line. Stage 6 assumes working capital to fund 90 to 120 days between paying for components and receiving customer payment, plus sales and support staff. Revenue offsets part of the burn; the figure shown is net. No owned facility, no clean room, no regulated medical or defence product. Any of those can multiply the figures.
How to read the ladder. Three features matter more than the specific numbers.
First, cumulative cash before repeatable sales lands in the range of roughly ₹7.5 crore to ₹15 crore under these assumptions. A seed round sized for a software company often covers only stages 1 to 3.
Second, the most expensive stages come after the technology has been proven. Stages 5 and 6 together account for about 60% of the total at both ends of the range. This is why investors who back the prototype are often not the ones who can fund the scale-up, and why the bridge between a successful pilot and a Series A is where many hardware startups stall.
Third, each stage has a defined exit test. A company that moves to pilot production without a frozen, certified design, or to scale without pilot yields, usually pays for the step twice.
Ways to shorten the ladder
The ladder is not fixed. Founders in India have five ways to compress it.
Use commodity hardware where possible. If the advantage lies in software or analytics, building on certified off-the-shelf modules and sensors can remove much of stages 2, 4 and 5.
Get customers to pay for pilots. A paid field pilot, even at cost, shortens stage 3 and tests willingness to pay. A separate PanScience Research piece covers how deeptech ventures sell to government buyers, including procurement rules that relax prior turnover and experience requirements for recognised startups.
Design for certification from day one. Choosing pre-certified radio modules and power supplies, and designing enclosures and layouts to the relevant standard, reduces retest risk in stage 4.
Use the manufacturing ecosystem the schemes are building. The expansion of component and contract manufacturing under PLI and ECMS should, over time, reduce lead times and minimum order quantities for domestic sourcing.
Plan non-dilutive capital into stages 1 to 3. Research grants and mission-linked funding can carry the earliest stages, preserving equity for the more expensive later steps.
Conclusion
Building a hardware startup in India in 2026 means operating in a manufacturing base that has grown several times over in a decade, with policy now reaching down into components. It also means facing a capital market that has concentrated its deeptech money in AI and thins out after seed. The gap between those two facts is where hardware companies live or die.
The milestone and cash ladder above is meant to make that gap explicit. Under stated assumptions, a connected or industrial device needs roughly ₹7.5 crore to ₹15 crore before it has repeatable sales, with the largest share spent after the technology works. Founders who plan for that, design for certification early, and use customers and grants to fund the first stages have a far better chance of reaching the point where the new manufacturing ecosystem starts to work in their favour.
Methodology and sources
This article combines official data on Indian electronics manufacturing and certification with an original, illustrative cost model. The milestone and cash ladder is a PanScience Research model built from the stated assumptions; it is not an observation of any company's spending and should be adjusted to each product. Production, export, PLI and ECMS figures are from the PIB backgrounder "Electronics Components Manufacturing Scheme: Union Budget 2026-27 Raises Scheme Outlay to ₹40,000 Crore", 3 February 2026 (static.pib.gov.in/WriteReadData/specificdocs/documents/2026/feb/doc202623777901.pdf). Aggregate PLI investment and jobs are from a Lok Sabha reply of 21 July 2026 as reported by Morung Express (editor-verified). SPECS thresholds are from SS Rana's summary (ssrana.in/corporate-laws/manufacturing-industries/specs-incentive-scheme-india-electronic-semiconductors-manufacturing/). CRS details are from Nemko (nemko.com/product-certification/bis-compulsory-registration-scheme) and TÜV Rheinland (tuv.com/market-access- services/en/certification-filter/bis-compulsory-registration-scheme.html). MTCTE details are from C-DOT's MTCTE page (msso.cdot.in/aboutMTCTE). Startup funding figures are from the Nasscom-Zinnov Indian Tech Start-up Report 2025 (zinnov.com) and the IVCA Bharat DeepTech Report 2026 as reported by YourStory (both editor-verified). Prototype cost figures are from BCG and Hello Tomorrow, "The Dawn of the Deep Tech Ecosystem", March 2019 (media-publications.bcg.com/BCG-The-Dawn-of-the-Deep-Tech- Ecosystem-Mar-2019.pdf), converted at ₹95 to the US dollar.