Americans are posting less: What the Incogni survey means for developers who build on social data

Most social media coverage tracks engagement metrics from platforms, which have an obvious incentive to report those numbers optimistically. Survey data from users themselves tells a different story, and lately that story is about retreat.

The Incogni findings are worth paying attention to if your product touches public social content in any way: as training data, sentiment signal, user-generated content, or raw social graph. The supply of organic public posts is shrinking, and the people posting least are now the ones who were most active.

What happened

Incogni, the personal data removal service that spun off from Surfshark in 2022, surveyed 1,000 Americans across four generational cohorts: Boomers (1946-1964), Gen X (1965-1980), Gen Y (1981-1996), and Gen Z (1997-2012). The full results are on Incogni's blog.

The headline finding: more than half of respondents said they post less than five years ago and have become more selective about who can see what they post. More than half also agreed that "maintaining an online presence feels like work," with around a third of those choosing "strongly agree" and only 16% disagreeing.

[ Free report ]

The Agentic SDLC

How teams like Stripe, Ramp, and Uber move from solo coding agents to a coordinated, team-level system.

Download the guide

Key features

The survey's most useful data points for developers, by cohort and topic:

• Posting fatigue skews young: 60% of Gen Z say maintaining a social presence feels like work, versus 38% of Boomers. The group most active on social platforms is also the most exhausted by it.
• Stepping away reads mostly as relief: Positive reactions to disconnecting (peace, relaxation, relief) averaged around 21%, edging out negative reactions (anxiety, FOMO, discomfort) at 19%. Gen Z was the exception: 27% reported negative reactions versus 26% positive, the only cohort where the split flipped.
• Politics is accelerating the pullback: 44% of respondents agreed that political content is driving people off social media, while only 20% disagreed. Among Gen Z, 48% agreed, and just 13% disagreed.
• Security is the leading quit trigger: More than half could picture deleting an account over security concerns; almost half cited harassment or hate speech. One in six said nothing could make them quit entirely.
• The US is the largest data broker target: Incogni's research ranks the US first for personal data harvesting, ahead of Canada and the Netherlands, driven by a large online population and limited data protection regulation.

Why it matters

Public social posts are raw material for a lot of software: sentiment analysis, trend detection, social listening, training pipelines, and data broker profiles all assume a steady flow of public, human-authored content. That assumption is weakening on two fronts at once. People post less, and the posts they do make increasingly sit behind restricted privacy settings rather than public feeds.

For teams building social features, "posting feels like work" is not an abstract attitude shift. It is a measured complaint about a specific product failure: algorithmic feeds that bury friend content in sponsored posts and suggested videos, turning social networks into broadcast channels rather than places where people actually connect. If your product reduces that maintenance burden, you are solving a documented problem.

For teams that build on public social data, the composition of that data is changing faster than the volume numbers suggest. Organic user posts are declining, while creator and sponsored content are growing to fill the gap. A sentiment pipeline calibrated on 2020 public post data may be reading a fundamentally different mix today.

Example use case

A data engineering team runs a Python pipeline that pulls public social posts through platform APIs for brand sentiment scoring. Over several years, they have watched the ratio of ordinary user posts decline while promotional content has grown to fill the gap in raw volume.

The practical response is to instrument the pipeline for composition, not just volume. Track the ratio of organic posts to promotional content per batch. Flag when public-post density drops below a defined threshold. Weight sentiment models toward sources that still carry an authentic signal, such as review platforms or opt-in first-party feedback channels.

Treat this as a monitoring problem now. Catching it through a dashboard alert is a lot cheaper than explaining to a stakeholder why sentiment scores have been quietly drifting for a quarter.

Competitive context

Incogni operates in the personal data removal space, scanning people-search websites and data aggregator databases and automating opt-out requests on users' behalf. PCMag has named it an Editors' Choice in that category. The company's interest in this particular survey is direct: social media is one of the primary data sources for the brokers Incogni fights, so measuring retreat from public posting is a form of market research for its own business.

The survey gives you something platform data cannot: generational granularity on a US sample from a single balanced study, rather than engagement metrics that platforms control and have reason to present optimistically. The cohort-level splits on posting fatigue, quit triggers, and reactions to disconnecting are the data hardest to derive from platform-reported numbers alone.

My take

The finding I keep coming back to is the Gen Z exception on stepping away from social media. Every other cohort reports a net positive reaction to disconnecting. Gen Z is the one group in which negative reactions slightly outnumber positive ones, meaning the youngest cohort is both the most fatigued by maintaining a social presence and the most anxious when they step back from it. That is a harder problem to design around than simple disengagement.

For developers, the more immediate question is whether your product's data assumptions were calibrated to 2019 social behavior and whether anyone has checked them since then. The decline in organic public posting has been gradual enough that it may not have triggered any alerts yet, even if it is already affecting signal quality.