Your company needs a new capability. Should you build it, buy it, or partner for it?

Years ago, I made the case to the leadership of a global consumer electronics company that it should acquire a relatively young streaming-device company. The TV screen was becoming a dumb pipe, and the value was moving to the software that decided what viewers watched. Owning the startup would let the company keep making world-class TVs while extending its reach onto screens it didn’t make.

The internal teams disagreed. The company had deep hardware expertise, strong engineers, and its own smart TV software. Why buy what we could build? A few years later, the company shipped its own streaming device. It wasn’t as good as the one that the startup created, it never gained traction, and the project was shut down.

That young company went on to go public and become a leader in streaming.

The story pains me to share even now, but it’s one that gets told over and over by other businesses as they wrestle with the decision to build, buy, or partner. Artificial intelligence has accelerated the pace at which businesses can develop new products, enter new markets, and reach new customers. AI also exposes gaps in a company’s capabilities, so leaders are facing the build/buy/partner question more often. Many are defaulting to the wrong decision: build.

Why does a bias for building in-house persist? The reasons include overconfidence (“I just need four people and six months”), not-invented-here syndrome, and what I refer to as launch bias, or estimating what it takes to launch, not what it takes to run the capability.

Now businesses are hampered by AI overconfidence, or a belief that AI can reduce the costs and effort of building in-house. In fact, AI is changing the math, but not evenly. It has made prototypes dramatically cheaper, and the savings are real. But they’re front-loaded. Running a capability, which means maintaining, securing and improving it for years, still takes resources, and AI shrinks those costs far less than it shrinks the cost of a demo.

And unfortunately, all the other assumptions about building in-house seldom hold up. The people you thought you had get reassigned as other priorities take hold. And businesses confuse faster launch with effective ownership. By the time you realize you need to reconsider your approach, someone else has beaten you to market, or the project has lost internal mindshare.

Instead, you might be better off thinking like Ryan Gosling’s character Ryland Grace in Project Hail Mary, who realizes that saving humanity requires more than his own ingenuity. He needs a partner with capabilities he doesn’t possess. You may not need an alien ally, but you do need a map to guide you.

Before Choosing: Check Your Biases

Before deciding what to do, challenge your assumptions. Look three years out, accounting for the costs of building, maintaining, and improving the capability, not just launching it. Compare your last three internal builds against their original estimates for time, cost, and staffing. Establish a threshold for abandoning the effort before you start spending, so you don’t keep funding a bad decision. And ask someone without a stake in the outcome to challenge your thinking.

Build, Buy, or Partner?

The more strategic a capability seems to be, the more tempted you will be to own it. Importance is necessary, but not sufficient. If strategic importance were the be-all and end-all, every business would build its own oil refinery. A capability needs to be difficult for the competition to replicate and easy to integrate into your operations. In all, you should test the desired capability against four criteria: defensibility, time to value, integration risk, and cultural fit.

1. Is the capability defensible?

Does the capability create a durable moat and advantage that your customers notice? If the capability is easily commoditized and creates only a marginal improvement, building may not justify the investment.

2. How quickly do you need the value?

If you need to move quickly, do a reality check. Building might take too long, but acquiring a capability could be time consuming too, if doing so requires an acquisition. If your window is closing soon, you may need to partner with someone.

3. What is the integration risk?

Every option carries integration risk. When you build, the risk is ownership itself: compliance, maintenance, staffing, and a long tail of fixes. When you partner, you’re at the mercy of your partner’s capability development road map and factors out of your control, like the partner being acquired. With buying, you risk hidden technical debt, platform fragmentation, and culture clashes, which have harpooned many an acquisition.

4. What does the cultural fit look like?

This is an obvious question for a buy decision. Can you onboard and keep the people you need for the long run? But it’s also a potential challenge for a partner, with the additional wrinkle of managing potential competitive risks and building trust if your partnership is not exclusive.

At this point, clarity will emerge, and with clarity comes sensible decision making along these lines:

Build If: You are confident that the required capability is not only important but also creates a competitive wedge and customer loyalty for the long term, and you are willing to fund it for years.

Buy if: You need speed but don’t need to own the capability.

Partner if: You need speed and ownership of capability.

These choices aren’t necessarily permanent or mutually exclusive. You might partner first to enter the market and learn, then build or acquire once you better understand where the value lies. You’ll also notice I said nothing about AI in the framework. Defensibility, time to value, integration risk, and cultural fit address the same risks whether your engineers write every line by hand or generate a working demo over the weekend. AI changes the cost of starting. It doesn’t yet change the cost of owning.

The team I argued with could build a streaming device, and eventually did. What it couldn’t build was the position the startup already had. The next time someone tells you, “We can build that,” thank them for their confidence, and let them apply all the AI they want to build a prototype. Then ask to see year three.

内容来自 Fast Company,版权归原作者所有。

阅读原文 ↗

用 HomepageTab 新标签页关注 RSS,打开浏览器就能看到更新。

安装 HomepageTab