An adoption curve only shows where it has already been

The signal

This method entry, part of Signal Atlas's cross-sector desk, looks at how an adoption curve is actually measured. Pew Research Center's own tracking of US mobile device ownership shows the shape often called an S-curve in real numbers: smartphone ownership among US adults rose from 35% in "the Center's first survey of smartphone ownership conducted in 2011" to 91% today, while ownership of a cellphone of any kind now stands at 98%. A companion fact sheet on internet and broadband use reports that "about half of all adults were already online" when the Center began tracking in early 2000, against 96% today, with home broadband reaching 78% after what the page calls rapid growth between 2000 and 2010.

The evidence

Both figures come from Pew's National Public Opinion Reference Survey; the most recent smartphone reading was collected from 5,022 adults surveyed 5 February to 18 June 2025 using address-based sampling. The Census Bureau separately tracks similar ground through its Computer and Internet Use programme, which has "been collecting data on computers and Internet use sporadically since 1984" through the Current Population Survey supplement and the American Community Survey, an independent, larger-sample government series a reader can weigh against Pew's smaller, more frequent tracking. Neither series states in advance where the curve bends next; both report where past survey waves landed.

Timeframe and confidence

Editorially, the shape connecting these dated points, a slow start, a fast middle, a flattening top, is a genuine, repeated pattern across the technologies Pew and the Census Bureau have tracked for two decades or more, and the percentages above are well-documented, survey-based readings for named years. What is not established by either source is which part of that shape a newer, still-rising technology occupies, since the flattening top of a curve is only visible after the fact.

What would change the reading

Confidence in placing a current technology on the curve would improve with two or three further years of the same survey question, tracked consistently, since one high growth rate cannot by itself distinguish an S-curve's steep middle from a curve that never flattens at a high level.

• How many waves of the same survey question exist, and over how many years?
• Has the year-on-year growth rate started to slow, or is the series still in its steepest section?
• Is the current reading from the same survey method as the earlier points, given Pew's own account of a mode shift after 2023?

An S-curve is a shape fitted after the fact to a series of dated survey readings. The readings above are real and specific; the location of any single new technology on that shape, while it is still rising, is not something the same data can yet show.

Source trail

1. Demographics of Mobile Device Ownership and Adoption in the United Stateswww.pewresearch.org · Source publication: not established · Retrieved 2026-09-16

   Gives the 2011-to-present cellphone and smartphone ownership percentages and the survey's sample and field dates.

2. Demographics of Internet and Home Broadband Usage in the United Stateswww.pewresearch.org · Source publication: not established · Retrieved 2026-09-16

   Gives the early-2000-to-present internet use and home broadband percentages.

3. Computer and Internet Usewww.census.gov · Source publication: not established · Retrieved 2026-09-16

   Confirms an independent, larger-sample government tracking series running since 1984 for comparison with Pew's survey.

Read across the evidence

• A hype cycle is a placement, not a measurement
• Smartphone ownership crossed a majority of US adults in 2013
• A first jobs print is revised twice before it is final

Adoption & operating reality · All 100 historical records