The short answer
To set up a venture building unit, take eight steps in order: secure a C-suite sponsor, state the objective, choose where the unit sits, set funding and decision rules, appoint a dedicated leader, build a mixed team, choose a legal vehicle and agree how success is measured. The unit should be close enough to the business to use its brand, customers and partners, and far enough away to move at startup speed.
Before you start: three questions
| Question | Why it matters | What good looks like |
|---|---|---|
| Who sponsors the unit? | Top builders have at least one C-suite sponsor, and CEO and board backing stops a venture being cut after a single quarter (McKinsey) | A named executive with budget authority and time on the calendar |
| What is the objective? | Revenue, strategic options and financial return need different measures | One written objective approved by the board |
| Which assets will the ventures use? | Ventures that use assets competitors cannot copy are easier to defend | A short list of customers, data, distribution, brand and expertise |
Table 1. If any answer is missing, resolve it before hiring. Our guide to why corporate ventures fail shows what happens when they are skipped.
Eight steps to set up the unit
- Secure the sponsor and write the objective. McKinsey found that organizations with five or more support mechanisms built nearly twice as many businesses, with a success rate 2.4 times higher than those with none.
- Choose where the unit sits. Use Table 2. The further the unit sits from the core, the more freedom it has, and the more deliberately it must borrow assets.
- Set funding and decision rules. Release capital in stages against evidence (Table 3), and agree the criteria before each gate.
- Appoint a dedicated leader. The leader should be able to steer the venture through its foundation phase, which McKinsey puts at six to 18 months.
- Build a mixed team. Combine experienced corporate staff with external hires (Table 4).
- Choose the legal vehicle. Decide the entity type early (Table 5).
- Agree incentives and metrics. Define how venture teams are rewarded and how progress is measured (Tables 6 and 7).
- Decide whether to build alone or with a partner. McKinsey reports that companies that consider partners early, especially for go-to-market, see roughly twice the success.
Where the unit should sit
| Option | How it works | Strength | Watch out for | Example |
|---|---|---|---|---|
| Inside a business unit | Reports to a divisional head | Easy access to customers and expertise | Core priorities crowd out the venture; practitioners warn against it | Common starting point |
| Corporate-level unit | Reports to the CEO or a C-suite sponsor | Visible sponsorship; portfolio view | Can feel remote from the business units | bp Launchpad, formed in 2019 as a new legal entity |
| Separate unit with its own entity | Own legal entity and board; parent holds the shares | Speed and clear accountability | Needs deliberate links to the core | Siemens next47, announced in 2016 with EUR 1 billion over five years |
| Holding company for ventures | A holding company houses several ventures, each separate | Each venture has its own brand, governance and board | More legal and admin overhead | Standard Chartered’s SC Ventures holds 15 ventures |
Table 2. Sources: McKinsey (January 2024, March 2021); Global Venturing (12 July 2024); Siemens press release (28 June 2016). Practitioner opinion on in-department units: Global Venturing (5 August 2024).
Funding and governance
| Element | Recommended practice | Source |
|---|---|---|
| Funding period | 12 to 24 months at a time | McKinsey, January 2024 |
| Release of capital | In tranches against evidence, such as a first product and early customers | McKinsey, January 2024 |
| Ring-fencing | 84% of 50 US and UK equity analysts said a dedicated unit with ring-fenced funding improves value growth | McKinsey, March 2021 |
| Steering committee | Agile sessions that remove blockers; decisions at the lowest practical level | McKinsey, June 2023 |
| Decision rules | Agreed criteria for each gate, written before the review | TURN8 go no go decision template |
| Portfolio discipline | Close underperformers; experienced builders have 2.8 successes per underperformer against 1.6 for less experienced ones | McKinsey, November 2023 |
Table 3. Sources are McKinsey publications; McKinsey sells business-building services. Practical budgets vary widely, and no generic benchmark for a unit exists in the sources reviewed.
The team
McKinsey describes three kinds of talent: business-building talent (sales and growth), technology talent (software, data engineering, data science) and product talent (product owners and designers). In its October 2024 survey, product and design roles were the most reported capability gap for both expert and novice builders, followed by commercial and data roles. Experts were 2.5 times as likely to report no talent gaps. Practitioners generally recommend hiring outside leaders because few suitable founders sit inside large corporates, and a 2025 study of 40 practitioners in Germany, Austria and Switzerland found that mixed teams of corporate insiders and external hires worked well.
| Role | What they do | Where to find them |
|---|---|---|
| Venture lead | Owns one venture end to end | External hire or experienced internal leader |
| Product owner and designer | Define and test the product with customers | External hire; the most reported gap |
| Engineer or data specialist | Build the first version | Mix of internal and external |
| Commercial lead | Wins the first customers and tests pricing | Internal sales experience plus external growth skills |
| Finance and legal partner | Keeps the venture inside corporate rules without slowing it | Seconded from the parent |
Table 4. Team design. Typical team sizes are small: a case study of four ventures at a Norwegian venture builder found teams of four to five people, though this is one firm. ADP Ventures reports about 65 product developers and engineers across several ventures at once.
Legal vehicle and ownership
| Option | When to use it | Note |
|---|---|---|
| Business unit | Venture stays close to the core and will be absorbed | Simplest; least independence |
| Joint venture | A partner brings capabilities or customers | Needs clear control and exit terms |
| Separate company or spin-off | Venture is far from the core or may take outside investors | The further from the core, the more a spin-off is favored |
| Holding structure | Several ventures under one parent | SC Ventures uses one holding company for 15 ventures |
Table 5. Based on Bundl (a venture developer, so a vendor view) and Global Venturing (12 July 2024). Take legal advice before choosing. For the Saudi position on the Simplified Joint Stock Company, see our guide to corporate venture building in Saudi Arabia.
Incentives and equity
| Benchmark | Figure | Source and caveat |
|---|---|---|
| Studio share of common stock | 10% to 20% | Fenwick partner quoted by Global Venturing, 12 July 2024 |
| Option pool | 10% to 20% | Same source |
| Co-founders | 60% to 80% | Same source |
| A Norwegian venture builder (one firm) | Founders about 70%, parent about 30%; share allocation delayed 1 to 3 years | arXiv case study, data from 2020 and 2021 |
Table 6. These are studio-style benchmarks, not rules for a corporate unit. Agree equity with legal counsel and the board.
Metrics that work from the start
| Stage | Example measure | Source |
|---|---|---|
| Early | Customer interviews: McKinsey suggests 5 to 15 before writing code | McKinsey, October 2020 |
| Product-market fit | McKinsey rules of thumb: referrals at least 20% of daily new customers; daily actives at least 10% of the base; more than 40% “very disappointed” if the product were removed | McKinsey, October 2020 |
| Unit economics | Cost to acquire a customer under 15% of annual revenue per customer | McKinsey, October 2020 |
| Portfolio | Four measures per project: risk, potential profitability, time spent and cost | Strategyzer, 2020 (a vendor view) |
Table 7. Thresholds are rules of thumb from one firm. Strategyzer warns that demanding a return too early can kill the wrong ideas.
A 12-month plan: TURN8's illustrative sequence
| Months | Focus | Output |
|---|---|---|
| 1 to 2 | Sponsor, objective, asset map, operating model | Approved charter and budget for the first funding period |
| 2 to 4 | Leader and core team hired; legal vehicle chosen | Team in place; entity decision made |
| 3 to 6 | Opportunity scan and customer interviews | Three to five validated opportunities |
| 6 to 8 | First gate review | Go, conditional go or stop on each opportunity |
| 8 to 12 | Build of the first venture; second wave scoped | First product with early customers; lessons recorded |
Table 8. Illustrative sequence. McKinsey reports that the time to reach $10 million in revenue fell from 38 months in 2023 to 31 months in 2025 (survey estimates).
Common first-year mistakes
| Mistake | Why it hurts | Better practice |
|---|---|---|
| Building everything in-house | Slows go-to-market | Consider partners early, especially for sales and distribution |
| Underestimating go-to-market | Good products stall with no customers | Test sales and pricing from the first phase |
| Stakeholders write the feature list | The product reflects internal opinion, not customer need | Let customer evidence set the roadmap |
| Chasing a perfect first version | Delays learning and adds technical debt | Launch a minimum product and improve |
| Not closing weak ventures | Capital and talent are stuck | Use gates with clear stop rules |
| Corporate rules apply unchanged | Marketing restrictions and rigid HR rules slow the team | Agree exceptions in the charter |
Table 9. Source: McKinsey (June 2023, January 2024, November 2023).
Build alone, partner, or both
A corporate with no venture team today can start with a partner who builds with its assets while it develops its own capability. TURN8 works in three ways: Operate For You, where TURN8 designs, builds and operates; Operate With You, a hybrid of build and operation alongside your teams; and Build to Own, where TURN8 stands up the capability, trains your people and hands over operations. See venture building at TURN8 and the TURN8 method. If you are new to the topic, start with what corporate venture building is.
Planning a venture building unit?
TURN8 can scope the structure, funding and first venture. Discuss a Mandate or Request a Private Briefing.
Frequently Asked Questions
How do you set up a venture building unit?
Secure a C-suite sponsor, state the objective, choose where the unit sits, set staged funding and decision rules, appoint a dedicated leader, build a mixed team of corporate and external talent, choose a legal vehicle and agree how success is measured.
Where should a venture building unit sit in the company?
It should sit close enough to use the parent’s brand, customers and partners, and far enough to move at startup speed. McKinsey links success to how separate the venture is from the core, and practitioners warn against keeping it inside a single department.
How much should a venture building unit be funded?
There is no reliable generic benchmark. McKinsey advises funding of 12 to 24 months at a time, released in tranches tied to milestones. Global Venturing reports that 65% of programs plan to spend under $1 million per venture over its lifetime.
What roles does a venture building team need?
A venture lead, product and design, engineering or data, commercial and a finance and legal partner. McKinsey reports that product and design roles are the most reported capability gap, so plan to hire them first.
Should a corporate build ventures alone or with a partner?
McKinsey reports that companies that consider partners early, especially for go-to-market, see roughly twice the success, and that building everything in-house is a common error. Many corporates start with a partner while building their own capability.
What legal structure should a new venture use?
The options are a business unit, a joint venture, a separate company or spin-off, or a holding structure. The further the venture sits from the core, the more a spin-off is favored. Take legal advice before choosing.
Important information
This article is general information about corporate venture building and is not financial, investment, legal or tax advice. Figures are taken from the sources listed below and were checked in October 2026; they may change. Examples and thresholds are illustrative unless stated. Past results do not predict future performance.
Sources
- McKinsey, Now is the time to create new corporate ventures (22 January 2024)
- McKinsey, How to build businesses faster and better with AI (March 2026)
- Global Venturing, Venture building (12 July 2024)
- Global Venturing, How ADP builds ventures that stay inside the business (16 October 2025)
- Garidis, Rossmann and Murray, ECIE conference paper (2025) (40 practitioner interviews, DACH region).
- Tkalich, Moe and Ulfsnes, case study of a venture builder (arXiv) (one firm; data from 2020 and 2021).
- Siemens, next47 press release (28 June 2016) (older than 2020).
- Strategyzer, Taking the guesswork out of innovation (12 August 2020) (vendor view).
- TURN8, Our Approach and Venture Building (engagement models).
Partner · TURN8
Ahmed Hassan is a Partner at TURN8, the venture operating partner for GCC corporates and sovereign-linked institutions. He leads the firm’s corporate venture offerings, the venture studios, accelerators, and CVC platforms. TURN8 builds and runs for its partners, and owns go-to-market across every engagement.