Most Indian corporates that try to build new ventures do not fail for lack of ideas or money. They fail because the decisions that determine a venture's fate are taken in the wrong order, by the wrong people, or not at all until it is too late. A corporate venture building playbook is a way of putting those decisions in sequence, so that each one is made when the information to make it exists and before the cost of getting it wrong has compounded.
The case for doing this deliberately is strong. McKinsey's 2024 survey of 1,176 executives found that about half of CEOs rank venture building among their top three priorities, and that experienced builders are roughly twice as likely to succeed as first-timers, while novices are about three times as likely to fail to reach product-market fit. Experience, in other words, is a large part of the outcome, and a playbook is how an organisation turns one team's experience into a repeatable practice. The same research base also explains the cost of improvisation. McKinsey's 2021 work on corporate and startup collaborations found that only 28% of startups were satisfied with their corporate partnerships, and that many corporate venture capital units stopped investing after two or three years.
This piece synthesises the twenty-four preceding articles in PanScience Research's cluster on deeptech and corporate venture building into a single sequence, from thesis to spin-out. Each stage names the question it answers, the decision it requires, the artefact it should produce, and the signal that should end or advance the effort. Each stage also points to the separate PanScience Research piece that treats it in depth, so this article stays at the level of sequence and does not repeat them.
Why corporate venture building needs a sequence
Venture building inside an established company differs from building a startup in one important way: the parent controls most of the resources and most of the decisions, and its normal processes are designed for a different kind of business. That creates three recurring sequencing errors.
The first is committing to a structure before committing to a thesis. A company sets up a venture studio, a corporate venture capital fund or a joint venture because peers have one, then searches for something to put in it. The structure then shapes the thesis rather than the other way round.
The second is deciding governance after the first crisis. Decision rights, funding cadence and kill rules are easy to agree when nothing is at stake and very hard to agree when a venture is late, over budget and championed by a senior executive.
The third is leaving the venture's destination open. Whether a venture will be reintegrated into a business unit, kept as a subsidiary, or spun out with outside capital changes how its team should be paid, how its IP should be held and which legal entity it needs. Deciding late means redoing work.
The sequence below is designed to prevent all three.
The corporate venture building playbook: eight stages
The playbook has eight stages. They are presented in order, but stages overlap in practice and some ventures will loop back. What matters is that no stage's decision is skipped.
Stage 1 thesis and mandate
The first question is what the parent wants from new ventures that it cannot get from its existing businesses, research and development, or acquisitions. The answer should be written down as a thesis: the markets, technologies and customer problems where the parent has an unfair advantage, and the outcomes it expects (new revenue, strategic options, capability, or financial return).
The thesis then determines the route. "CVC, venture building, M&A or internal R&D: four ways a corporate buys the future", a separate PanScience Research piece, compares those routes on control, speed, cost, learning and failure mode. "The corporate venture studio: how established companies build new ones" sets out four studio models and what each requires of the parent. For family groups, "Indian family businesses and conglomerates as venture builders" describes the archetypes common in India and their governance implications. For public sector undertakings, "Public sector undertakings and innovation: how PSUs can build and buy new technology" maps the routes available within procurement and vigilance rules.
For deeptech theses specifically, two further pieces set the frame: "What a deeptech venture studio does differently: longer clocks, harder science, different money" and "The deeptech ecosystem in India in 2026: labs, money, missions and buyers".
Stage 2 governance before ventures
Before the first venture is funded, the parent should agree who decides what, how money is released, and when a venture is stopped. "Why corporate ventures die inside the parent: the governance problem" provides a checklist covering decision rights, funding cadence, kill rules and talent. The single most important rule is that kill criteria are agreed when a venture is funded, not when it is in trouble.
Governance also includes the metrics the board will use. "Measuring corporate innovation: the metrics that predict value, and the vanity ones" distinguishes leading from lagging indicators by venture stage. A venture in its first year should be judged on evidence of customer pull, not revenue.
Stage 3 choosing how to engage each opportunity
Not every opportunity in the thesis should become a venture the parent builds. Some are better bought from a startup, some better pursued with a partner, and some better built.
"Venture clienting: when a corporate should buy from a startup instead of investing in it" sets out when buying is the better route. "Corporates and deeptech ventures: anchor customer, investor or acquirer" describes a six-step relationship ladder from paid pilot to acquisition, with what each step should cost and exclude. "Technology joint ventures in India: when two companies should build one" offers a five-question test for when a joint venture beats a licence, contract or investment, and the five clauses that decide whether one works. For enterprise AI specifically, a separate PanScience Research piece in the enterprise AI cluster covers the choice between building, buying and co-building.
Stage 4 the business case and its gates
Once an opportunity is chosen for building, it needs a business case that a board will approve and that does not overclaim. "Writing the corporate venture business case that a board will approve" proposes a stage-gated template that says what may reasonably be claimed at each gate. Early gates should require evidence about the problem and the customer; later gates evidence about the product and unit economics.
For deeptech ventures, the gates should track technical maturity alongside commercial maturity. "From TRL 3 to first revenue: where deeptech companies actually die" maps the commercial milestone that should accompany each technology readiness level, and "The seven- year company: planning milestones, runway and morale for a deeptech venture" provides a milestone template with go and no-go criteria. Hardware ventures need their own cash plan, set out in "Building hardware in India: prototypes, certification, manufacturing and the cash it takes".
Stage 5 team and incentives
The team decides more than the plan does, and McKinsey's finding that experienced builders are about twice as likely to succeed puts a premium on attracting and keeping them. "Intrapreneurs: incentives, equity and career risk for people who build inside companies" sets out an incentive design table covering cash bonuses, phantom equity, real equity in a venture entity and career guarantees, with the Indian mechanics of each. The central design rule is to choose the incentive by the venture's destination, and to pair any upside with a written return ticket for the intrapreneur.
Stage 6 IP and assets
Most corporate ventures start with something the parent already owns: technology, data, brands, customer relationships or licences. How those are transferred shapes the venture's freedom to raise capital and sell to the parent's competitors. "Licensing corporate IP to new ventures: terms that let the venture succeed" provides a term sheet guide covering field of use, exclusivity, royalties and improvements. "IP strategy for deeptech ventures: patents, trade secrets and what to file when" covers the venture's own filing decisions.
Stage 7 capital and recognition
As a venture matures, it may need capital beyond the parent's budget. "Deeptech funding in India: why patient capital is still the missing layer" maps which money exists at each stage, noting the IVCA Bharat DeepTech Report 2026 finding, reported by YourStory, that investor participation drops sharply after seed. "India's new Deep Tech Startup category: what the 20-year, ₹300 crore definition changes" explains the February 2026 DPIIT recognition rules. Corporate builders should note one provision in particular: G.S.R. 108(E) excludes entities formed by splitting up or reconstruction of an existing business, which may affect ventures carved out of a parent. "Government as first customer: how deeptech ventures sell to the Indian state" covers public procurement as a source of early revenue.
Stage 8 destination
The last stage is deciding where the venture ends up: reintegrated into a business unit, held as a subsidiary, or spun out with outside capital. "Spinning out a corporate venture in India: carve-out mechanics, people and IP" sets out the carve-out sequence covering entity, assets, IP, people, contracts and capital. Destination should be provisionally decided at stage 4 and confirmed here, because it determines the incentive and IP choices made at stages 5 and 6.
The master table
The table below is the original contribution of this piece: a one-page synthesis of the sequence. For each stage it names the question, the decision, the artefact the stage should produce, the signal to stop or advance, and the separate PanScience Research piece that treats it in depth. It is an analytical framework assembled from the cluster, not an observed process from any single company.
Three principles that run through the sequence
Three principles recur across the cluster and hold the sequence together.
Decide early what is expensive to decide late. Kill rules (stage 2), destination (stages 4 and 8) and IP terms (stage 6) are cheap to agree at the start and expensive to renegotiate. Most of the sequencing errors described above are failures of this principle.
Price every step of every relationship. Whether the relationship is between the parent and a startup, between two joint venture partners, or between the parent and its own intrapreneurs, the cluster's pieces repeatedly find that trouble comes from rights granted without a price. Exclusivity without volume, equity without a valuation event, and pilots without a named owner are versions of the same mistake.
Build experience deliberately. McKinsey's finding on experienced builders points to an organisational choice, not only a hiring one. A parent that runs one venture every few years will remain a novice. A parent that runs a portfolio with consistent governance, documented gates and a stable core team accumulates the experience that the research associates with success. BCG reports a 66% success rate across more than 200 ventures it built with partners, against 20 to 30% for traditional venture capital and corporate venture capital benchmarks. That is the firm's own figure for its own model, and should be read as such, but it illustrates the scale of advantage that repeated practice is claimed to produce.
How the playbook applies to deeptech
Deeptech ventures stretch every stage. Timelines are longer, technical risk persists into stages where software ventures would already be selling, and capital after seed is scarce in India. The playbook adapts in three ways.
At stage 4, technical and commercial gates run in parallel. A deeptech venture should not be killed for lacking revenue at a gate where the agreed evidence was a technical milestone, and should not pass a gate on technical progress alone when the agreed evidence included a paying pilot.
At stage 5, incentives must survive a longer clock. A bonus tied to annual targets will not hold a team through a seven-year build. Real or phantom equity with a clear valuation event, plus career guarantees, are better suited.
At stage 7, the parent may need to remain the main funder for longer than it planned, or to bring in strategic co-investors on terms that do not narrow the venture's market.
PanScience, a New Delhi deep-tech and AI venture studio, reports that it has co-founded more than 24 ventures across nine verticals, with shared in-house product, technology, go-to-market, operations, finance and legal functions (company-reported). The studio model is one way of institutionalising experience across stages 3 to 8; the corporate equivalent is a standing venture unit with a stable core team and consistent governance, rather than a sequence of one-off projects.
Conclusion
A corporate venture building playbook for India is less a set of new ideas than a discipline about order. The parent writes a thesis before choosing a structure, agrees governance before funding ventures, chooses for each opportunity whether to buy, partner or build, gates the business case on evidence, designs incentives and IP terms around the venture's intended destination, plans capital beyond the parent, and decides at the end whether the venture belongs inside or outside. Each stage has a separate PanScience Research piece behind it. The sequence is what turns those pieces into a practice, and practice, on the evidence, is what makes experienced builders roughly twice as likely to succeed.
Methodology and sources
This article is a synthesis of the other twenty-four articles in PanScience Research's Cluster 3, cited by title as separate pieces. It uses only figures from the editor-verified list. McKinsey's 2024 findings on CEO priorities and experienced builders come from "How CEOs are turning corporate venture building into outsize growth" (22 October 2024; https://www.mckinsey.com/capabilities/mckinsey-digital/our- insights/how-ceos-are-turning-corporate-venture-building-into-outsize-growth). McKinsey's 2021 findings on startup satisfaction and CVC lifespans come from "Collaborations between corporates and start-ups" (https://www.mckinsey.com/capabilities/strategy-and-corporate- finance/our-insights/collaborations-between-corporates-and-start-ups). BCG's success rate is from "A Proven Model for Corporate Venturing" (15 June 2022; https://www.bcg.com/publications/2022/corporate-venturing-models) and is the firm's own figure. The DPIIT G.S.R. 108(E) provisions of 4 February 2026 are as summarised by Nishith Desai Associates, Lakshmikumaran and Sridharan and TaxGuru. The IVCA Bharat DeepTech Report 2026 finding is as reported by YourStory (yourstory.com, August 2026). The eight-stage sequence, the master table and the three principles are the author's analytical synthesis, not an observed process. PanScience facts are company-reported.