The advice every founder hears at the clan meeting is "protect the business for the children." It sounds like wisdom. It is closer to taxidermy. The families that actually stay wealthy across generations are not the...
The advice every founder hears at the clan meeting is "protect the business for the children." It sounds like wisdom. It is closer to taxidermy. The families that actually stay wealthy across generations are not the ones that preserved one business in amber. They are the ones that kept starting businesses.
That claim is not a motivational poster. It is the central finding of a research program that quietly demolished the most repeated statistic in family wealth.
For decades, the field measured family success by asking one question: does the original company still exist under family control? By that measure, most families "fail" by the third generation, and an entire advisory industry grew up around the famous 70 percent failure rate. Then James Grubman went looking for the source of that number and found, in a 2022 paper worth reading in full, that the statistic traces back to a small 1980s study of Illinois manufacturers that was never designed to say what it is quoted as saying (Grubman, There Is No 70% Rule, 2022).
Inside that same paper sits the more important citation. Thomas Zellweger, Robert Nason and Mattias Nordqvist, publishing in Family Business Review in 2011, proposed changing the unit of analysis from the firm to the family. Stop asking "did the company survive." Start asking "did the family keep creating value." When they looked at enduring business families that way, the pattern flipped. Long-lived families were rarely faithful to a single company. They ran portfolios. They founded, bought, sold, spun off, shut down, and founded again, across generations. The firm was mortal. The family's entrepreneurial capability was the thing being inherited. Zellweger and colleagues gave it a name, transgenerational entrepreneurship, and the punchline is blunt: selling grandfather's company is not failure. It is often what success looks like mid-stride.
Harvard Business Review reached the same conclusion from the practitioner side. George Stalk Jr. and Deborah Wilen, reviewing the evidence in 2021, found no support for the three-generation doom story and observed that thriving old families treat businesses as vehicles, entering and exiting them as conditions change (HBR, July 2021). A family that exits a declining trade at a good price and redeploys into a growing one shows up in the doom statistics as a "failed" family business. Its bank account disagrees.
Attach the honest caveats before building on this. Zellweger, Nason and Nordqvist studied families that already endured, which invites survivorship bias: we see the serial founders who made it, not the serial founders who scattered capital across six bad ideas and died broke. And Grubman's demolition of the 70 percent rule does not prove succession is easy; it proves we never measured it properly. The research licenses a change of question, not recklessness. Hold that, because the second half of this piece is about the discipline that keeps serial from becoming scattered.
Strip away the family-office vocabulary and the transgenerational portfolio is something market families already do, usually without naming it.
The duka comes first. Years of small margins, stock bought carefully, a reputation for being open when the rains come. The duka's surplus buys the plot, because the family refuses to let trading profit evaporate into consumption. The plot carries two rentals, and the rent is boring money, which is exactly what you want financing the next move. Rent plus duka surplus opens the second shop in the next trading center, run by a sister who trained behind the first counter for three years. Then the daughter comes home with a plan for a hardware line, or an agro-inputs shop, or a delivery service tying the two shops together, and instead of being told to go find a job, she pitches the family bank: a written proposal, an amount, terms, a repayment schedule, a named mentor from inside the family. The family lends. She launches the third thing.
Count what just happened. One family, two decades, four assets: duka, plot, second shop, new venture. No single one of them is impressive. The sequence is the wealth. And notice what the sequence required at every joint: surplus deliberately captured, a successor deliberately trained, capital deliberately governed. The portfolio family is not a lucky family. It is a family with a pipeline.
Now run the monogamous version of the same family. All energy defends the original duka. The founder ages with the business, the shelves age with the founder, and the supermarket opening across the road does to the duka what supermarkets do. The children were taught to guard, not to build, so when the guarding fails there is nothing behind it. The family did everything the clan meeting advised, and the advice was the problem. A family with one business is one bad decade away from being a family with a story.
There is also a quieter benefit the research keeps surfacing: multiple ventures create multiple seats. The classic succession war is three capable children and one chair. A portfolio family can hand different children different ventures, matched to different temperaments, and the inheritance fight dissolves into a division of labor. Serial entrepreneurship is succession planning wearing work clothes.
Before this becomes a sermon for starting anything with a signboard, sit with the most sobering family business story on record.
Kongo Gumi, a Japanese temple-construction firm, was founded in 578 and run by the same family for roughly 1,400 years and 40 generations, the oldest continuously operating family company ever documented. It survived wars, fires, regime changes, and the Meiji suppression of Buddhism. What ended its independence in 2006 was none of those. In the 1980s bubble, the firm borrowed heavily to speculate in real estate, a business it did not understand, and when the bubble burst the debt did what debt does. The company that had mastered one difficult craft for fourteen centuries was liquidated as an independent firm within two decades of wandering outside that craft.
So the instruction is not "diversify." Kongo Gumi diversified; that is precisely what killed it. The instruction is serial within competence. Each new venture should sit close enough to what the family already knows that existing knowledge, reputation, and relationships transfer. The duka family opening a second duka, then a wholesale line, then agro-inputs, is climbing a ladder where every rung touches the last. The duka family putting its capital into a friend's mining concession because the returns sounded exciting is Kongo Gumi in the bubble. Zellweger and colleagues' enduring families were serial, yes, but their ventures typically radiated out from a core capability, the way a tree branches from a trunk rather than planting random seedlings across the county.
A workable test for any proposed venture: can a family member run it on day one with skills the family already has, and does our existing reputation bring it customers in week one? Two yeses, proceed to numbers. Two noes, you are not investing, you are gambling with the family's name attached.
Four disciplines convert this from theory into a family operating system.
1. Name the pipeline. At the annual family meeting, put one item permanently on the agenda: what is our next venture, who is developing it, and what has to be true before we fund it. A family that never discusses the second business will never build it. The pipeline agenda item costs nothing and changes what the children think the family is for.
2. Fund through the family bank, not through gifts. New ventures get loans with written terms, or equity with defined shares, decided by more than one person. Governance is what separates the portfolio family from the family that "helped" four relatives start four things that all died quietly. A venture that cannot survive a written repayment plan cannot survive a market.
3. Give every venture an exit condition as well as a launch condition. The serial family's edge is not only starting well. It is stopping well. Decide at funding time what evidence, and what date, would trigger sale or closure. The families in the HBR account thrived because they exited declining vehicles at prices worth having. Sentiment is for the founder's portrait on the wall, not for the balance sheet.
4. Train operators before you need them. Every family venture should carry an understudy, a younger member working inside it with real responsibility. The portfolio can only grow as fast as the family produces people who can run things. Capital is usually not the binding constraint. Trained hands are.
This week, answer two questions in writing at your family table. First: what is our next venture, the one adjacent to what we already do well, specific enough to have a cost estimate. Second: who would run it, named, with the training gap between where they are and where they need to be stated honestly. If no name survives the second question, your next venture is actually a person, and the family bank's first disbursement is their training.
One business is a job the family shares. A pipeline of businesses, each near the family's competence, each governed, each with an operator and an exit condition, is a legacy that does not depend on any single signboard surviving.
This piece did its job if your family meeting now has a standing agenda item called "the next venture," and the answer to "who runs it" is a name, not a shrug.