Social Purpose Report FY 2026
Proportional Data is a Washington Social Purpose Corporation. We believe accountability should be observable. So this report, like our platform, shows its work: what we set out to do, what we're doing now, how we measure it, and where we're falling short.
The purposes this report is accountable to.
Our articles of incorporation commit the company to three social purposes. Every objective, action, and measure in this report maps back to one of them.
Advancing transparency in consumer-web privacy
We measure how the consumer web tracks people and publish it as a public good, on observable evidence anyone can check. Observations of industry practice, not legal conclusions about any organization.
Human-centered adoption of AI
We help organizations adopt AI that pays off by keeping human judgment in the lead, with privacy and autonomy built in, and their workforces taking part.
Shaping AI and data policy
We advocate for policy that keeps AI human-led and reinforces transparency and consumer privacy.
Short- and long-term objectives.
For each social purpose, we state a near-term objective for this fiscal year and the long-term outcome it builds toward.
Instrument the baseline.
A cross-cutting objective spanning all three purposes.
Cross-cutting objective for FY2026: stand up the platform and set the first full-year baselines, grade distribution, consent adoption, and cross-context reach, that future reports measure against.
Advancing transparency in consumer-web privacy
Measuring the consumer web as a public good.
Operate the weekly Consumer Web Index: A to D grades on consent effort and tracking restraint across about 5,800+ consumer sites, on a fully public method.
Turn findings into plain-language guidance consumers can use to see and reduce their own tracking, and sustain a reproducible dataset researchers and regulators can rely on.
Human-centered adoption of AI
Tying AI's payoff to human judgment.
Show through advisory work that AI pays off when human judgment leads, and help teams adopt it that way, with privacy and autonomy built in.
Make human-led, privacy-respecting AI adoption a demonstrated standard, with adopting workforces taking an active part.
Shaping AI and data policy
Turning evidence into better public policy.
Publish evidence-based positions for AI guided by human judgment and progressive data privacy, and take it into policy and industry forums.
Help set norms and policy where AI amplifies human judgment and protects consumer privacy as it scales.
Material actions in progress.
In our first months, most of our effort goes to building and operating the public measurement platform, the evidence engine behind all three purposes. The figures below are drawn live from platform operations, updated as we collect and score 2026 data.
Keeping the methodology fully public
The complete scoring methodology is released so any grade can be independently scrutinized and reproduced. It is the foundation of our claim to independence.
Weekly measurement and public grades
We crawl every site in the index and publish A to D grades on two axes, consent effort and tracking restraint, showing how much a site holds back by default.
Advancing our AI position
Early advisory engagements show AI pays off when human judgment guides it. We help teams work that way, and fold what we learn into our public policy view.
Open, free access to the data
The dashboard is published free of charge, so consumers and researchers can see how the sites they use handle their data.
Measures, and where we are steering them.
We hold ourselves to a small set of honest metrics. Read this as our instrument panel: where each measure stands this first year. FY2026 is the baseline year, so there is no prior period to compare against yet, and FY2027 targets will be set once this year closes.
| Metric | FY2026 | Status | FY2027 target |
|---|---|---|---|
| Sites measured weekly | 5,520 | establishing | TBD |
| Tracking vendors classified | 157 | establishing | TBD |
| Proper consent infrastructure | 27% | establishing | TBD |
| Cross-context restraint | 24% | establishing | TBD |
| Methodology transparency | 100% | at target | 100% |
| Advisory engagements | 2 | establishing | TBD |
Cross-context restraint
No cross-context tracking observed at all this year. This is the most positive reading in the product, and a baseline we intend to move.
Grade distribution observed
Most of the web is "Minimal": little consent effort, but little default tracking either. The sharper concern is the smaller "Tracks anyway" share, sites that run consent tooling yet still track heavily by default.
Challenges & limitations.
We pursue our purposes with real constraints. Naming them plainly is part of the accountability this report is meant to provide, and a guard against overstating what our data can prove. Think of them as the instrument's calibration limits: the places our reading is uncertain.
Bot detection limits crawl coverage
Some sites block or fingerprint automated crawlers, or serve them different content than they serve people. Where we cannot get a representative load, we exclude the site rather than publish a misleading grade, which narrows coverage.
We observe client-side behavior only
Our crawl sees what happens in the browser. Server-side consent enforcement, back-end data sharing, and contractual data flows are not directly observable, so a site's real-world handling may be better or worse than its observable posture.
Observable posture is not a legal judgment
Our grades describe measurable tracking and consent behavior. They are not determinations of legal compliance: a site may be lawful in ways our crawl can't see, or non-compliant in ways it never surfaces.
This is our baseline year
FY2026 sets the zero point. With no prior year to compare against, this report states levels, not trends; the movement it is built to show begins with next year's reading.
Consent regimes vary by jurisdiction
Consent requirements differ across regions, and a single crawl vantage point cannot capture every regulatory context. We are explicit about where our observations apply and where they do not.
On the record.
The findings, grades, and figures in this report reflect observable tracking and consent behavior, the behavior we can measure by crawling each site, and are not legal determinations about any organization. Platform figures update as we collect and score 2026 data. Our complete scoring methodology is published and open to scrutiny.
This is our working framework for that first annual report, following the structure of Washington's social-purpose standard, RCW 23B.25.150.