October 2025 McKinsey Business Building Practice The way to win in corporate venturing: Serial building and AI The results of our sixth annual survey show that repeat venture builders see the greatest ROI, benefiting from both experience and advanced uses of AI. This article is a collaborative effort by Daniel Aminetzah, Jason Bello, Jerome Königsfeld, and Paul Jenkins, with Corinna Leist and Robin Martens, representing views from McKinsey’s Business Building Practice. Once an executive experiences the entrepreneurial spark, it’s hard to extinguish. The latest McKinsey Global Survey on corporate venture building finds that leaders of companies that have built new ventures want to build Responses show that companies that have launched new ventures—that is, entirely new products, services, or businesses that go beyond incremental upgrades to create entirely new revenue streams—in the past five years are significantly more likely than others to prioritize venture building. Surveyed executives say these efforts are achieving returns quickly and with less investment than in previous years, and the more ventures they build, the more success they report. The findings also suggest that, overall, companies’ venture-building skills are maturing. In our sixth annual survey on the topic, reported new ventures are seeing larger revenues earlier than in the past. Meanwhile, the practices that set successful venture builders apart go beyond the more foundational elements that emerged as success factors in previous years, such as having a C-suite sponsor and sufficient financial resources. In particular, companies with successful ventures are using technology such as AI to enable their creation of new ventures, and many respondents say data- and AI-driven ventures are in their future. The macroeconomic environment isn’t deterring venture builders The current macroeconomic environment—in which geopolitical instability and changes in trade policy vie for executives’ attention—has posed challenges for organizations so far in 2025. Lower-than-average consumer confidence reflects widespread economic Despite these potential obstacles, surveyed business leaders report that companies with experience in building new ventures remain committed to such efforts. In fact, experienced business builders are doubling down. Leaders from companies that have built new ventures in the past five years are 13 times more likely than others to have increased their prioritization of new-venture building in the past year. New-venture building is a top five priority for 58 percent of experienced venture builders, suggesting experience with new-venture building inspires companies to keep at it. Those companies that take the calculated risk to build new ventures often see success. After several years in which 43 to 44 percent of respondents reported successful new ventures that met or exceeded their core organization’s expectations for growth or scale, the share is approaching 50 percent in the latest These new ventures are generating meaningful revenues. This year, we see a larger share of reported new ventures with annual revenues over $10 million. Sixty-one percent of respondents say their new ventures have crossed this mark, up from 45 percent in 2023 (Exhibit 1).4 A closer look shows that the share reporting revenues between $50 million and $100 million has 1 The online survey was in the field from May 13 to May 28, 2025, and garnered responses from 715 senior managers and C-level executives across 66 nations who say they work for companies with $100 million or more in annual revenues and represent the full range of regions, industries, functional specialties, and tenures. To adjust for differences in response rates, the data are weighted based on each respondent’s nation, taking into consideration its contribution to the region’s share of the global GDP (based on purchasing-power parity). Prior to 2025, the survey employed a country-based weighting scheme to align with global GDP (based on purchasing-power parity) and included respondents from companies with all revenue sizes. The new region-based weighting scheme limits over- or underrepresentation of individual countries and improves weighting efficiency of the data set overall. All data have been calculated after removing the share of respondents selecting “don’t know.” 2 “Consumer confidence index (CCI),” OECD, accessed July 15, 2025. 3 In the latest survey, 47 percent say that new ventures launched in the past five years had met or exceeded expectations. The 2023 and 2024 figures differ from previous years’ publications as data have been adjusted to reflect a change in the threshold for companies included in the sample (to include only respondents working for companies with $100 million or more in annual revenue). 4 The 2024 edition of the survey did not capture this information. 2The way to win in corporate venturing: Serial building and AI nearly doubled, with a corresponding reduction in new ventures with reported revenues less than $1 million. The findings also suggest that new ventures are generating value faster than in previous years. The average age of new ventures that respondents say have revenues over $10 million has decreased to about two and a half years (31 months), compared with more than three years (38 months) in the earlier The findings also suggest that when it comes to building new ventures, more is better; a portfolio of multiple ventures leads to larger reported gains. Fifty-nine percent of surveyed business leaders who launched three or more new ventures in the past five years say their companies are seeing more than 10 percent of total enterprise-wide revenue from their venture-building efforts, compared with 32 percent of respondents from companies that launched just one new venture. We also see differences in success when looking at the underlying focus of new ventures. Ventures built around digital products and services—such as software, mobile apps, and cloud- based platforms—generated the highest average revenue, while hardware or physical-product ventures generated lower revenues. 5 Companies may be younger than the stated age. These figures were calculated using the weighted average from the maximum age possible given the respondent’s answer selection in the survey question, “In which year of operations is the new business currently?” For example, if a respondent answered that the business was in its second year since launch, we used 24 months when calculating the average. Exhibit 1 Web <2025> <VentureBuilding> Exhibit <1> of <8> Annual revenue of new ventures at respondents’ companies, % of respondents1 Note: Figures may not sum to 100%, because of rounding. 1Survey asked about the current annual revenue of new ventures launched in the past 5 years. Figures were calculated after removing the respondents who selected “don’t know.” 2023 data adjusted due to change in threshold for companies included in sample (≥$100 million in annual revenue). Source: McKinsey Global Surveys on the state of new-venture building, 2023 and 2025 The latest survey �ndings show a larger share of corporate ventures crossing key revenue thresholds. McKinsey & Company >$100 million $51 million–$100 million $11 million–$50 million $1 million–$10 million <$1 million 2023 (n = 509) 22 45 61 8 15 34 21 11 29 26 15 20 2025 (n = 424) 3The way to win in corporate venturing: Serial building and AI Exhibit 2 Web <2025> <VentureBuilding> Exhibit <2> of <8> Investment into new venture before it broke even, % of respondents1 Weighted average investment, $ million1 1Question: “What was your organization’s approximate investment in the new business before it broke even?” Figures exclude respondents who said “don’t know” or “not applicable,” and the 2024 data have been adjusted to account for a change in threshold for companies included in the 2025 sample (≥$100 million in annual revenue). Source: McKinsey Global Surveys on new-venture building, May 21–July 2, 2024, and May 13–29, 2025 The weighted average investment that survey respondents say was required for ventures to break even has decreased since 2024. McKinsey & Company ≥$1 billion or more $201 million–$999 million $101 million–$200 million $11 million–$100 million <$10 million Has not broken even yet 4 4 1 1 3 10 53 15 35 39 7 28 2024 (n = 417) 2025 (n = 420) 125 77 Companies are limiting their exposure to risk and breaking even within two years Our research shows that companies are being judicious about their investments in new-venture building. The findings suggest that they are creating new ventures based on proven concepts or existing assets, largely staying within the industry they know, and using AI to build ventures more efficiently at a time when AI technologies can create significant value for businesses with just a few employees. As a result, they are breaking even after lower levels of investment—and that’s disproportionately true for the companies building the most ventures. In last year’s research, we found that the weighted average investment required before a new venture broke even6 was about $125 million, whereas this year it has dramatically decreased to $77 million (Exhibit 2). This translates to an average of about 2 percent7 of the core organization’s annual revenue required as an investment to break even in 2025. 6 Break-even investment numbers refer to cash flow breakeven. The survey question was, “What was your organization’s approximate investment in the new business before it broke even?” 7 This 2 percent figure was calculated by dividing the weighted average investment required for the new venture to break even (about $77 million) by the weighted average annual revenue of the core organization (about $ billion) in 2025. 4The way to win in corporate venturing: Serial building and AI Respondents from the most experienced companies—that is, those that respondents say have built at least three ventures in the past five years—report the strongest outcomes. These respondents report generating times the revenue for every dollar invested until breaking even, compared with at companies that have launched just one or two Our latest research also looked for the first time at how long it takes new ventures to break even, and more than eight in ten respondents report breaking even within three years. In fact, the majority break even within the first two years of operations (Exhibit 3). Companies headquartered in Asia and Latin America get their ventures to break even more swiftly. Among the 22 percent of respondents whose new venture broke even within their first year of operations, a majority are based in Asia, while most ventures breaking even in the second year of operations are built by companies headquartered in Latin America. The timing and level of investment required to build a value-generating venture differ based on the type of asset at its core. In our survey, nearly one-third of respondents say their new ventures focusing on data or intellectual property broke even with investments of less than $1 million, while a much smaller share says ventures focusing on physical products required investments under $1 million before breaking even (see sidebar “Data monetization: A lower-cost option for new-venture building”). What’s more, industrial consumer-focused products each have a longer 8 These figures were calculated by dividing the weighted average revenue by the weighted average investment required until breaking even in each category. Exhibit 3 Web <2025> <VentureBuilding> Exhibit <3> of <8> Year of operation in which new venture broke even, % of respondents (n = 354)1 1Question: “Approximately when did the new business break even?” Asked of only the respondents who said they were familiar with a new venture launched by their organization in the past 5 years that had broken even. Results exclude respondents who answered “don’t know.” Source: McKinsey Global Survey on new-venture building, 715 senior managers and C-level respondents at companies with ≥$100 million in annual revenue, May 13–29, 2025 Most new ventures that survey respondents say have broken even did so within the �rst two years of operation. McKinsey & Company 1st year 2nd year 3rd year ≥4th year 22 38 24 15 5The way to win in corporate venturing: Serial building and AI break-even period—a weighted average of more than two years, compared with new ventures built around data and intellectual property. This can be expected, in our experience, given that digital businesses are easier to scale than those requiring physical production and distribution. Often, the best new ventures stem from existing assets not yet living up to their potential, and we found in last year’s research that nearly nine in ten business leaders said their organizations had at least one asset with unrealized commercial potential. This year, respondents who say their companies had above-average growth rates compared with industry peers in the past 12 months—that is, high-performing companies—are much more likely than other respondents to say their companies have scaled new ventures out of assets that had unrealized potential. Overall, 72 percent of respondents who say their companies built such ventures report above- average growth. Often, high-performing companies are not creating entirely new playbooks from scratch; rather, they use proven business concepts from other markets (that is, from outside the organization). Of respondents who say their companies used proven concepts, two-thirds report an above- average organic rate of growth for their industries. Adopting a successful start-up business model is often a successful strategy we see for incumbents to scale their new ventures quickly. For example, JPMorgan Chase introduced online bank Chase UK, using the digital-first approach popularized by neobanks such as Monzo and Revolut to attract tech-savvy customers and expand its market presence. Additionally, we find that companies are increasingly launching new ventures that are within their primary industry. While in 2024, about half of respondents said the new venture they were most familiar with was built in the same industry as their company’s primary industry, this year, about six in ten reported staying close to home. Organizations have already increasingly embraced data monetization when reimagining their business models. A notable example is in “commerce media,” in which commercial companies leverage customer data and media platforms to connect brands with their customers. A leading grocery retailer in Canada launched a commerce media network to address margin pressures and increase alternative revenue growth. The retailer built a unified platform to enable advertisers to access a suite of targeted advertising solutions. Early efforts on the platform included securing 12 brand campaigns, implementing ad-serving technology, and creating custom audience segments for more precise targeting. The initiative is expected to generate more than $150 million in annual revenue for the retailer by its fifth year. This type of media network, which started in retail, has now expanded into travel, financial services, telecommunications, and other B2C and B2B businesses. Furthermore, incumbents in this industry are capitalizing on their customer data to launch fintech ventures that provide tailored financing solutions, enhance customer access, and generate new revenue streams. Data monetization: A lower-cost option for new-venture building 6The way to win in corporate venturing: Serial building and AI AI is helping companies—especially repeat builders—launch new ventures Many executives say their companies are already using AI for a wide range of tasks within venture-building efforts, though they expect to use AI in even more ways in the years ahead (Exhibit 4). The most commonly reported uses for AI within venture building in the past 12 months include operational aspects such as enhancing workflow efficiency and monitoring the scaling of new ventures, as well as more creative uses such as developing new-venture ideas or marketing campaigns for the ventures (see sidebar “How AI is reshaping venture building”). Increasingly, we see that AI agents can support companies across the entire venture-building process, from ideation and building the first prototypes to scaling the go-to-market plan. This enables companies to build multimillion-dollar businesses with only a handful of employees. Just recently, Base44, an AI coding start-up, was sold to Wix, a web development platform, for $80 million— and it had just eight employees. Responses suggest that the companies seeing the largest revenues from new ventures are the most advanced users of AI. The respondents who say their companies are using AI for more complex activities in their venture-building efforts—such as building new businesses with AI as their primary value proposition, validating initial business concepts, and improving go-to-market strategies by personalizing marketing campaigns or improving the customer journey—report ventures with revenues twice as large as those using AI for more basic tasks and four times as large as the small share of respondents who say their companies aren’t using 9 For respondents reporting advanced AI use, n = 250. For those reporting basic AI use, n = 100, and for those reporting no AI use in new-venture building, n = 39. Exhibit 4 Web <2025> <VentureBuilding> Exhibit <5> of <8> Use of AI by respondents’ companies while building new ventures, % of respondents (n = 453)1 1Respondents who said “not applicable” are not shown. Figures were calculated after excluding respondents who answered “other” or “don’t know.” For the past 12 months, n = 476. For the next 5 years, n = 453. Source: McKinsey Global Survey on new-venture building, 715 senior managers and C-level respondents at companies with ≥$100 million in annual revenue, May 13–29, 2025 Companies are already using AI in a range of activities while building new ventures, and further uses are expected. McKinsey & Company 0 Past 12 months Next 5 years 10 20 30 40 50 60 70 Enhancing work�ow e�ciency Improving product development Generating and prioritizing new business opportunities Improving go-to-market strategies Re�ning the value proposition of new ventures by adding AI features Monitoring the scaling of new ventures Building entirely new ventures, using AI as the core of their value proposition Validating initial business concepts 7The way to win in corporate venturing: Serial building and AI The most experienced venture builders—serial builders that have built more than three ventures in the past five years—are the most likely to be advanced users of AI. Seventy-two percent of respondents at these companies say they are using AI for more complex activities, compared with 60 percent of respondents at companies that have built fewer ventures. While the changing economic environment has tightened the funding spigot for new ventures overall, companies are still planning to invest in AI- or data-focused ventures. Fifty-six percent of respondents say their companies are planning to build data, analytics, or AI-driven businesses within the next five years, up from 49 percent in 2024. Even if they aren’t planning to make AI the focal point of a new venture, nearly all respondents who expect their companies to build new ventures in the next five years say their organizations will use AI to assist their efforts (Exhibit 5). Only 3 percent of respondents say their companies won’t use AI to assist with venture building in the years ahead. Additionally, 84 percent of respondents say their companies will need to integrate AI or automation into their new-venture- building process over the next five years. Recent advances in AI can fundamentally improve the venture-building process. A small survey of R&D leaders found that they expect dramatic improvements in product–market fit, product performance, workplace productivity, and time to market from their companies’ use of AI. What once required large teams and months of effort can now be validated and iterated upon in just weeks, with leaner resources and higher confidence in the outcomes. AI is changing the game at every step of the new-venture building. First, companies can use AI to mine customer and competitor data to reveal unmet needs. AI can then support efforts to assemble the business plan and operating model, design a road map for launch, validate assumptions in real time, and automate financial simulations. Next, AI copilots and generative design tools are speeding up product development and can streamline the production of go- to-market plans. Companies can also scale the venture by expanding to new segments and geographies, and AI can continuously re-segment customers, identifying the highest-value targets and orchestrating personalized outreach at scale. Finally, companies can use AI to codify the lessons learned into reusable playbooks and models. An operational-intelligence company was eager to become an AI-powered analytics leader in agriculture and fuels. But grain procurement was opaque and reliant on infrequent, high-level insights, which limited buyers’ ability to optimize pricing and supply. In the fuels sector, sales and inventory planning lacked accurate signals for forecasting demand, which caused inefficiencies and missed margin opportunities. The company developed an AI-driven analytics product for each industry, integrating predictive insights, weather-informed analytics, and automated recommendations. Each product has the potential to unlock at least $1 billion in annual value in its respective industry through improved procurement decision-making and demand forecasting. What’s more, the potential for operational improvements is nearly endless. A roadside-assistance company in Europe identified a new value pool within its operations to maintain its position in the market and cut costs. Improving the diagnostics for broken-down vehicles was identified as one of the biggest areas of opportunity, enabling the company to send the right assistance to each customer. After six weeks of building and testing it in call centers, the company implemented an AI-enabled triage system with the capability to be refined in-house as more data about breakdowns became available. As a result, the company saw 100 percent accuracy for specific types of cases and a 63 percent improvement in prediction accuracy compared with their previous diagnostics. How AI is reshaping venture building 8The way to win in corporate venturing: Serial building and AI The key fundamentals to get right: Technology and people New ventures are only as successful as the people, culture, and technology used to build them. Our survey findings suggest that companies have gotten better at the nuts and bolts of venture building over the past five years and now realize that getting the people equation right is a leading indicator of a venture’s potential success. Last year, having a C-suite sponsor to champion new-venture-building efforts, dedicated financial resources, and a systematic approach to monitoring and evaluating were the elements that stood out most among the most successful venture builders. However, the latest results suggest that these are now foundational elements and that workforce-related capabilities, such as the ability to upskill and to use the latest technologies, have become the biggest areas of distinction. Use technology as a catalyst Companies that successfully build ventures tend to have the technology and modern infrastructure necessary for launching new businesses. Having advanced technological capabilities—such as generative AI, digital platforms, and a scalable IT infrastructure—is one of the most common traits among companies with successful ventures. Investing in advanced tools and technologies, while ensuring proper alignment with the organization’s operating models and upskilling initiatives, can amplify a new venture’s competitive edge and improve its ability to scale. Build with the right culture and people When we looked at the attributes that companies have in place when building new ventures, we found that a culture that embraces risk-taking and experimentation stands out. Sixty-eight percent of business leaders who report such a culture say their ventures have been successful at meeting or exceeding expectations (Exhibit 6). Embedding a culture in which employees feel emboldened to experiment—without fear of failure—inspires innovation that can drive growth. In our experience, we have seen that the right setting allows this to happen: Every employee has clarity on what to deliver each day and how their work contributes to a broader goal, and they are not only allowed but expected to take measured risks. Exhibit 5 Web <2025> <VentureBuilding> Exhibit <4> of <8> Company will use AI in new-venture building in next 5 years, % of respondents (n = 453)¹ AI use will be required to deliver on new ventures’ value propositions in next 5 years, % of respondents (n = 536)² 1Asked of only the respondents who said they expect their companies to build ≥1 new venture in the next 5 years. Figures were recalculated after removing respondents who said “don’t know.” ²Asked of only the respondents who said they expect their companies to build ≥1 new venture in the next 5 years. Figures were recalculated after removing respondents who said “don’t know” or “other.” Source: McKinsey Global Survey on new-venture building, 715 senior managers and C-level respondents at companies with ≥$100 million in annual revenue, May 13–29, 2025 Nearly all surveyed executives expect their companies to use AI in their new ventures over the next �ve years. McKinsey & Company 97 80 9The way to win in corporate venturing: Serial building and AI The success of corporate venture building also hinges on the upskilling of people at all levels of the organization. More than half of respondents with successful new ventures say their organizations foster the necessary skills for new-venture building through upskilling programs that train employees on topics such as customer discovery methods or hypothesis testing, whereas 42 percent of leaders with unsuccessful ventures report having upskilling programs. Companies eye data-driven ventures for the years ahead Creating new ventures is a strategic imperative for business leaders. Most respondents continue to expect venture building to be one of the top strategic moves their organization will pursue within the next 12 months (Exhibit 7), as was true in 2024. They are nearly as likely to expect new- venture building as they are to expect organizational transformation and cost cutting, which are common value-creating moves during times of uncertainty. Exhibit 6 Web <2025> <VentureBuilding> Exhibit <6> of <8> Share of respondents reporting each attribute and also a successful new venture,1 % of respondents 1Successful new ventures are those that, according to respondents, continue to operate and perform in line with or exceed expectations for scale and growth. Source: McKinsey Global Survey on new-venture building, 715 senior managers and C-level respondents at companies with ≥$100 million in revenue, May 13–29, 2025 Success with a new venture is most commonly reported when an organization has a culture of experimentation. McKinsey & Company A culture that encourages risk-taking and experimentation Upskilling programs to foster skills needed for new-venture building Technology and infrastructure necessary for launching new ventures Mechanisms to incentivize the success of new ventures Established partnerships and/or alliances A dedicated team with clear responsibilities and incentives for managing new-venture creation A systematic method for monitoring and evaluating outcomes of new ventures An established process for identifying, evaluating, building, and launching new ventures ≥1 C-suite sponsor to champion new-venture-building e�orts Dedicated �nancial resources set aside speci�cally for new-venture creation A dedicated leader of new-venture building who is responsible for new-venture-creation e�orts 68 52 47 45 45 44 43 43 41 40 38 10The way to win in corporate venturing: Serial building and AI As we have seen in the previous two editions of this annual research, respondents largely expect to create new ventures focused on data, analytics, or AI in the next five years (Exhibit 8). This is especially true for companies in technology, media, and telecommunications; financial services; and advanced industries. Venture building is one of four prevalent approaches to capturing value from AI that companies take today. The first is innovating existing products and services by embedding new AI-driven capabilities into offerings to create fresh revenue streams. The second is monetizing data through products or business models, while the next is building an entirely new AI-native business unit to disrupt operations or markets. Finally, the fourth involves reshaping an organization’s entire portfolio using state-of-the-art AI tools and technologies. Respondents working in energy and materials continue to expect to build environment- or sustainability-focused new ventures, while those working for consumer goods or retail companies most often expect to see new physical products. Exhibit 7 Web <2025> <VentureBuilding> Exhibit <7> of <8> Share of respondents who expect their organizations to make given strategic move in next 12 months, % of respondents (n = 710)1 1Figures were calculated after removing the share of respondents who said “don’t know.” The share saying “none of the above” and the strategic moves expected by <10% of respondents are not shown. 2This combines two related answer choices o�ered in the survey. They are “organizational transformation (eg, restructuring the shape of the organization, cost cutting)” and “signi�cant cost cutting across the organization.” Source: McKinsey Global Survey on new-venture building, 715 senior managers and C-level respondents at companies with ≥$100 million in annual revenue, May 13–29, 2025 New-venture building is one of the most commonly expected strategic moves in the year ahead. McKinsey & Company t Organizational transformation and cost cutting2 t Venture building t M&A t Scaling of existing underutilized businesses/assets t Joint venture and/or strategic alliance t Enterprise-wide digital transformation and/or transformation of core technology stack t Signi�cant reallocation of resources across the organization t Strategic diversi�cation 41 39 34 28 25 23 21 12 11The way to win in corporate venturing: Serial building and AI Exhibit 8 Web <2025> <VentureBuilding> Exhibit <8> of <8> Types of new ventures expected to be built by respondents’ companies in next 5 years, by industry, % of respondents expecting their companies to build ventures1 1Figures were calculated after removing the 2% of respondents who said “don’t know” and the 17% of respondents who said they do not expect their companies to build new businesses in the next 5 years. ²Includes banking, insurance, and private equity. Source: McKinsey Global Survey on new-venture building, 715 senior managers and C-level respondents at companies with ≥$100 million in annual revenue, May 13–29, 2025 Across most industries, survey respondents most often expect to build data- and AI-driven new ventures in the next �ve years. McKinsey & Company Advanced industries (n = 78) Financial services² (n = 79) Consumer goods and retail (n = 91) Energy and materials (n = 107) Healthcare, pharma, and medical products (n = 55) Travel, logistics, and infrastructure (n = 80) Technology, media, and telecom (n = 77) Total (n = 582) Data, analytics, and/or AI-driven businesses Digital and direct-to- consumer businesses Platforms, ecosystems, or marketplaces (connecting multiple parties) Connected or smart products (eg, Internet of Things) Physical, industrial products Physical, consumer- focused products Hardware or robotics Subscription or “everything-as- a-service” models Environmental or sustainability- focused businesses Other services (eg, remote healthcare) 5768 374454 5275 56 1946 471818 2021 27 1830 21118 2927 21 4113 81213 1829 18 233 02811 158 13 101 631432 112 19 344 245 616 10 2815 776 2042 18 2317 86017 4115 27 49 3412 67 6 Business leaders are confident that new ventures will contribute positively to their enterprise- wide revenue in the next five years, particularly when compared with their historical performance. According to our research, new ventures built over the past five years have contributed about 12 percent of total enterprise-wide revenue10 across industries. Looking ahead, respondents anticipate this figure will increase to approximately 19 percent in the next five years. 10 This weighted average was calculated by multiplying the share of respondents in each revenue category by the midpoint value of that range, then summing resulting values. This approach ensures that the average reflects both the range of revenue shares and how many organizations fall into each range, while excluding “don’t know” and “not applicable” responses. 12The way to win in corporate venturing: Serial building and AI In periods of economic uncertainty, it’s easy for business leaders to default to a focus on immediate profitability, delaying longer-term growth initiatives such as new-venture building. But our research shows this could be a costly mistake. Our survey shows that companies that invest in new-venture building—and particularly those that deploy AI capabilities to launch ventures more efficiently—can find outsize success with the practice, quickly generating new revenue that often meets or exceeds their expectations. That success is a catalyst for more success; once business leaders create one venture, they are eager to create others. Those who merely dabble in venture building don’t see the same level of success as repeat builders, who can develop ventures with bigger revenues for relatively similar amounts of investment as others. Our findings suggest that business leaders are concentrating on launching new ventures that might have higher chances of success, and they are more proficient at building them than they used to be. To make sure they’re equipped to get started, other companies can focus on training their employees with the skills they will need to help scale new ventures. AI-enabled skill assessments, structured learning journeys with continuous development opportunities, and capability-building programs can all help. Designed by McKinsey Global Publishing Copyright © 2025 McKinsey & Company. All rights reserved. Daniel Aminetzah is a senior partner in McKinsey’s New York office; Jason Bello is a senior partner in the Washington, DC, office; Jerome Königsfeld is a partner in the Cologne office; Paul Jenkins is a senior partner in the Oslo office; Corinna Leist is an associate partner in the Zurich office; and Robin Martens is an associate partner in the Amsterdam office. The authors wish to thank Cara Droege, Eléonore Duroyon, Gustav Kölby, Liam Alhomsy, Markus Berger-de León, and Sarah Stohrer for their contributions to this research. This article was edited by Heather Hanselman, a senior editor in McKinsey’s Atlanta office. Scan • Download • Personalize Find more content like this on the McKinsey Insights App 13The way to win in corporate venturing: Serial building and AI