Coined by Morton Grodzins (Metropolitan Segregation, 1957), modeled by Thomas C. Schelling (Dynamic Models of Segregation, 1971) and Mark Granovetter (Threshold Models of Collective Behavior, 1978), and popularized by Malcolm Gladwell (The Tipping Point, 2000).
Spots systems where each choice depends on what others choose, so nothing seems to happen until a threshold is crossed and then change comes fast.
A hundred sales reps and a new pricing tool: the same champions, placed together or spread thin
Each rep starts using the tool once enough colleagues they work with are using it: somewhere between 6 and 18 of them, counting a teammate in full and a rep in another region as a third. The champions use it from week one. Placed in one region, they tip that region and then the firm; spread four ways, the same number may never get started. Find the smallest number of champions that tips the firm each way. Illustrative.
Reps using the tool, out of 100
Champions placed in one regionChampions spread across four regions
Week 16 of 16
- Champions placed in one region
- 100
- Champions spread across four regions
- 12
An example
A building-supplies distributor bought a pricing tool for its 100 field sales reps and rolled it out the usual way: a webinar for everyone and three reps per region named as local experts. Two months later twelve reps used it, the same twelve as in week one. The reps said what reps say: nobody else is on it, and customers ask why our quotes look different.
The sales operations lead read that as a threshold problem. In a pilot, reps had started using the tool once about half a dozen people they worked with did, and some needed many more. So the relaunch put the same budget into one region: all twelve champions in the Midwest, side by side with its other thirteen reps for two weeks, with the region's manager reviewing quotes in the tool. The Midwest tipped in three weeks. The other regions, watching a full region quote faster, followed within the month, and usage passed 90 reps by week seven.
Why it works
When each person's choice depends on how many others have already chosen, the system has a threshold. Below it, every push decays: the people at the margin see too few others and wait, so nothing seems to happen however long the effort runs. Above it, each new adopter tips the next, and change arrives quickly and mostly on its own.
That shape breaks ordinary planning, which assumes results in proportion to effort. The same effort can produce almost nothing or almost everything depending on where it lands. Spread thin, it may stay under the local threshold everywhere. Concentrated, it can push one group over, and that group becomes the visible proof that lowers the next group's threshold. The practical question stops being how much effort and becomes where to put it to cross the line.
The thinking behind it
Morton Grodzins Morton Grodzins, a University of Chicago political scientist, is credited with coining the tipping point in his studies of white flight, including Metropolitan Segregation (1957): the share of Black residents at which white residents of a neighborhood began to leave in large numbers. The term named a threshold beyond which a gradual change turns abrupt and is hard to reverse, which is the sense the word still carries.
Thomas C. Schelling Thomas C. Schelling's Dynamic Models of Segregation (1971) showed with simple models of who lives next to whom that mild individual preferences about neighbors can tip into stark collective patterns that nobody in the model wants. He developed the theme of individual choices adding up to surprising aggregates in Micromotives and Macrobehavior (1978), and shared the 2005 Nobel Memorial Prize in Economic Sciences.
Mark Granovetter Mark Granovetter's Threshold Models of Collective Behavior (1978) gave each person a threshold: the number of others who must act before that person will. He showed that the spread of thresholds across a crowd decides whether a riot, fad, or movement takes off, more than the crowd's average attitude does. Two crowds alike in every way but one person's threshold can end in opposite places.
How to run it
- Ask whether choices depend on each other. If people adopt, leave, or join partly because others have, assume a threshold exists.
- Estimate the threshold from small trials. Pilot in one team and note how many colleagues each person needed to see before joining.
- Find the group people actually watch. Thresholds count the people someone works beside, so the unit is usually a team or a site, and the whole company only for very visible choices.
- Concentrate effort to cross the line. Put the budget, the champions, and the manager's attention into one group until it tips.
- Use the tipped group as proof. Make its results visible to the next group, which lowers that group's threshold.
- Watch for signs of the tip. Track weekly adoption in each group; the curve turns from flat to steep, and that is when to move effort on.
Where it goes wrong
The common failure is reading a stall as rejection. Adoption sits flat for months, and the effort gets canceled just under the threshold. Before concluding the idea failed, check whether any one group ever reached the pilot's threshold; if none did, concentrate before quitting.
A second failure is assuming a tipping point where choices are independent. If each person's decision depends only on price or need, effort pays in proportion and concentrating it wastes the other groups. Ask people whether colleagues' use affected theirs. Third, thresholds cut both ways: a few visible defections can tip a group back. Watch for the first few people dropping out, and respond fast.
When to use something else
- Network Effects. When the product itself becomes more valuable as more people use it, and the question is pricing and growth as well as adoption.
- Crossing the Chasm. When the adoption gap lies between early enthusiasts and pragmatic buyers, and the move is picking one beachhead segment.
- Technology S-Curve. When the question is where a technology sits on its maturity curve, not how to make a group adopt it.
Where it sits in SOLVE eX
Go to the source
- Metropolitan Segregation, Scientific American, 1957
- Dynamic Models of Segregation, Journal of Mathematical Sociology 1(2), 1971
- Threshold Models of Collective Behavior, American Journal of Sociology 83(6), 1978
- The Tipping Point: How Little Things Can Make a Big Difference, Little, Brown, 2000