Value creation, innovative business models, and non-traditional marketing.

November 11, 2026

About 60% of Lego's sales come from products that didn't exist before. Roughly 0% come from a new kind of brick.

That's how Rivkin of Harvard Business School describes Lego's innovation in a Working Knowledge interview. It's also the opposite of how Lego nearly died. In the 1990s the company stretched its brand into amusement parks, video games, children's clothing and wristwatches. It kept adding new brick shapes too. The HBS case notes say more bricks made products harder to assemble, forecasts harder to make and inventory harder to manage. Lego posted its first financial loss in 1998. A restructuring followed, with layoffs, streamlined production and design centers in London, Milan and San Francisco. Rivkin says that made things worse. By 2004 he describes the company as on the brink of bankruptcy, with a takeover by a rival or a private equity firm possible. Popular but unconfirmed figures say it was losing almost $1 million a day with nearly $800 million in debt. That year Jørgen Vig Knudstorp became CEO in his mid-30s. He fired five of seven manufacturing executives, removed many of the unique brick components added in the years before and brought brick manufacturing back in-house. Rivkin says the company's return on invested capital later rose above 100 percent. Lego didn't stop being creative. It chose where creativity was allowed. Sets, themes and characters keep changing, and the brick itself does not. Plenty of founders get this backwards. They add a new feature, a new tier and a new market, and call it growth. Each one adds complexity somewhere, usually in operations, where nobody is watching until the inventory piles up. Decide what in your business can never change, then be as inventive as you like around it. What's your brick?

Read on LinkedIn →

November 10, 2026

Sears helped launch an online service in 1984. Amazon didn't exist for another decade.

The service was Prodigy, a joint venture with IBM and CBS. It carried news, weather, email and message boards, and at one point it ranked among the biggest online services in the world. That's the part of the Sears story most people never hear. We tell it as a company that missed the internet. It didn't miss it. It owned a piece of it. This was a company that also started Allstate Insurance and the Discover card. Ideas were never the shortage. Sears also had something Amazon had to build from nothing: customers who already knew how to buy from a stranger through the mail. For nearly a century, its catalog taught millions of Americans how shipping, payment, substitutions and returns worked. It even sold 70,000 to 75,000 houses by mail between 1908 and 1940. That's most of what e-commerce asks of a shopper: trusting a stranger with a big purchase. So Sears had the habit, the brand, the warehouses and a ten-year head start on the technology. Then in 1993 it shut the catalog down. Amazon was incorporated the next year and opened for business in July 1995. Walmart had already passed Sears as the biggest retailer in America back in 1990. Prodigy was a $600 million venture. Amazon finished its first partial year with about $511,000 in sales, run out of a garage in Bellevue. One of them had every advantage and the other had a book and a website. Sears wasn't blind. It saw the future early and paid for part of it. Then it walked away from the one customer behavior that future was built on. If a shift is coming, protect the customer behavior it depends on before you chase the technology.

Read on LinkedIn →

November 9, 2026

Someone told you that you're a visual learner or an auditory learner. Teaching to that style is the part with little evidence behind it. In one U.S. survey,…

People do report preferences. The claim that matters is that matching teaching to a style improves results. In 2008, psychologists Harold Pashler, Mark McDaniel, Doug Rohrer and Robert Bjork reviewed the evidence in Psychological Science in the Public Interest. They counted 71 proposed learning style models. A fair test sorts learners into categories, randomly assigns them to different teaching methods, and gives everyone the same test at the end. If styles matter, each group should do best with the method that matches its style. Few studies used that design, and some contradicted matching. The authors concluded the evidence was insufficient to justify using learning styles in teaching. Say a manager sorts new hires into visual, auditory and hands on learners and builds a separate training track for each. The review studied education, not workplaces, but training is teaching too, and building three tracks costs time and budget. As of that 2008 review, the evidence that teaching to preferences improves learning was not there. Here are 3 things to try instead. They're suggestions, not research findings: - Pick the format by what you're teaching, not by who is learning. Use a diagram for something spatial, a worked example for a procedure, and practice for a skill. - Give everyone the same test at the end, so you can compare training methods. - Ask what people learned, not whether they liked the session. Has anyone ever told you that you're a visual or auditory learner? Did it change how you studied?

Read on LinkedIn →