For incumbents and start-ups alike, the odds of success with a new business are low. Only 20 percent of start-ups launch successfully—in other words, reach product and market fit (see sidebar, “What defines product and market fit?”) and scale up—and just 24 percent of incumbents’ new ventures become viable companies, according to a recent McKinsey survey.  Even so, start-ups continue to experience historic growth: the value of global venture deals has risen more than sixfold in the past decade, from $48 billion in 2010 to a projected $295 billion in 2019.

Meanwhile, incumbents have been slowly expanding their portfolios of corporate ventures with dedicated budgets, governance, and operating models. The shock of COVID-19 has accelerated this trend, with a 21-percentage-point increase in companies citing new-business building as one of their top strategies.  As incumbents plan their new ventures, they can boost their chances of success by avoiding common missteps that have derailed past efforts.

How to avoid common failure modes in a new-business launch

We’ve observed that when new-business launches fail, it’s not usually because of problems with technical feasibility. Rather, failure typically stems from flawed assumptions about the the new product or service’s desirability to customers. Incumbents face a particular challenge here. Their culture often biases them to seek perfection in planning and predictability in execution. Past experiments with new ventures may have left them loath to take on more risk for fear of another failure.

Drawing on experiences with a number of successful projects, we’ve identified five steps incumbents can follow to get their new venture off to the best start. (Read “The big boost: How incumbents successfully scale their new businesses” to learn more about how to scale new businesses that have successfully launched.)

[Here are the first two of five steps.]

Step 1: Talk to customers about your concept before you write a line of code

Incumbents often talk about product “features” and “solutions.” Smart teams think instead about creating an offering that doesn’t just serve a purpose but connects to customers’ emotional and social lives—and they don’t stop at understanding what a customer wants today but ask how it might change tomorrow.

Validating customer desirability can be a challenge for incumbents accustomed to a top-down approach. Typically, an executive team does some modeling, draws up a product road map, defines what the business needs, and gets IT to write the requirements into code. With this approach to problem solving, external factors don’t always get the attention they deserve. For instance, a successful launch doesn’t always have to rely on a novel concept; sometimes it can come from observing what works in another sector or country and adapting a promising idea to fit a new context.

Before writing a line of code, smart teams test their ideas with a few target customers—say, five to 15. They describe the proposed new offering, ask for feedback, and pay careful attention to how customers behave as well as to what they say, noting anything that elicits an emotional response. Afterwards, teams group their observations into customer pain points and jobs to be done and create a customer-research “backlog” list of the steps needed to validate the new concept. They then gather more signals from customers and other stakeholders, including IT, operations, and business partners.

One start-up in Singapore adopted this kind of “low-fi” approach when developing a loyalty program. It planned to launch a digital version of a stamp card to reward customers for frequenting their favorite restaurants and shops. On advice from the main investor, the start-up team went straight to the start-up’s target customers—merchants—and asked a single question: “Do you know who your top ten customers are?” This simple exercise quickly revealed that the merchants had no reliable way to identify their best customers and no digital channel for interacting with them.

The start-up team realized that the merchants’ main challenge was to identify, learn about, and grow their customer base, not just reward customers with promotions and discounts. The team went back to the merchants with a short presentation and managed to secure informal letters of intent from 50 of them within a week. The push for early positive feedback helped validate the concept enough to start building a platform with a heavy focus on each merchant’s analytics and personalization features.

Before writing a line of code, smart teams test their ideas with a few target customers.

Step 2: Observe customer preferences, don’t ask about them

Incumbents tend to rely on familiar forms of customer and market research, such as focus groups and quantitative surveys. Gut feeling among company owners or board members also plays a big part in decision making at some organizations. But methods like these don’t offer a fast or reliable way to validate the customer appeal of a new offering.

Experience has shown that a better alternative is to develop a workable minimum viable product (MVP) and use social media to observe—not ask—how customers respond to pain points, value propositions, and calls to action. Asking customers what they want seldom leads to breakthrough innovations. Who knew they wanted an iPhone until Apple launched it? Instead, smart companies observe customer behavior, imagine a new concept, and then test it by observing rather than asking. A good way to start is to take the customer-research backlog created in step 1 and identify concepts to test. For example, an Asian company tested different value propositions for a home wireless broadband business not by asking consumers which features they liked but by running short video ads on social media and tracking audience response rates. This quickly revealed that parental controls for the Wi-Fi router would be key to the success of the new offering.

A company’s first MVP can be as simple as a presentation, mock-up, or social-media ad. It should be short, glanceable, and geared to distinct customer pain points. The next step is to link the MVP to a landing page inviting customers to join a wait list, and then measure the click-through rate from the ad and the conversion rate on the landing page. Key metrics are tracked daily by platform, and the findings are combined with data from customer interviews and market research. The resulting insights and implications are shared across the organization via ambient posters, slides, or a website setting out the vision for the venture.

After a telecommunications company in the Philippines had completed initial research for its new loyalty program, it launched video story ads on Instagram to test whether consumers wanted a points or membership program and whether they preferred cash back or other kinds of rewards. A click-through rate of more than 2 percent on a display ad (when 1 percent would have counted as “good”) gave the company confidence that the concept was attractive, enabling it to move to the build stage.