Attention Before Action
Understanding technology adoption requires two independent signals — one that measures when a phenomenon enters public awareness, and one that measures when it enters operational behaviour. They are not the same signal. They are not interchangeable. Together, they tell a complete story.
The three cases analysed in this report have now reached the phase beyond what was documented here. Stablecoins shows the Infrastructure Paradox — attention down 42%, while card top-ups cross $1B/month and on-ramp share hits 60%. The full cycle is now visible: attention → adoption → invisibility. Buy Now Pay Later is now classified Dominant (89 consecutive months — Financial domain record). Open Banking shows dual-velocity transizionale: Wikipedia pageviews down 34% year-over-year while Google Trends holds above threshold. All three cases confirm and strengthen the original thesis: attention precedes adoption. What this update adds is the next chapter — what happens after adoption absorbs attention.
Updated reports: ZCR-001 Stablecoins (Infrastructure Paradox) · Dataset v2.2 (71 cases, operational status)
What Does It Mean to "Adopt" a Technology?
The word adoption is used loosely. It can mean that people are aware of something, that they are searching for it, that they have read about it, or that they are using it daily in their work. These are different things — and they happen at different times.
Recent data from the Gallup Workplace Panel (2023–Q1:2026) documents a sharp acceleration in AI use among finance and insurance workers — from roughly 8% frequent users in 2023 to nearly 45% by early 2026. The study identifies regulatory change as the primary driver: when compliance constraints eased, operational adoption followed.
This is one half of the story. The other half is: where was public attention before that adoption curve moved?
The Two Signals Are Not the Same
These signals measure different things. A person can be aware of a technology for years without using it. An organisation can be deploying a technology without the general public knowing. The gap between these two states — between awareness and behaviour — is where the most useful information lives.
Attention is a necessary condition for adoption, but not a sufficient one. You cannot adopt what you do not know exists. But knowing it exists does not mean you will adopt it. The triggers that convert awareness into behaviour — regulatory change, cost reduction, competitive pressure, organisational capacity — operate in the gap between the two signals.
Public Attention in Finance — Three Cases · 2019–2026
The Gallup adoption curve runs from roughly 8% in 2023 to 45% by Q1:2026 — a steep acceleration concentrated between 2024 and 2026. All three Zorthex attention curves were already moving before that acceleration began. Buy Now Pay Later had been above the structural threshold since 2019. Open Banking was present and stable from 2020. Stablecoins ignited in mid-2025, driven by the GENIUS Act — precisely the kind of regulatory trigger the Gallup study identifies as the primary adoption driver.
Case by Case
Attention was present and stable for three years before the Gallup adoption acceleration began. The public awareness infrastructure was already built when regulatory conditions allowed operational adoption to scale. The signal was upstream by years.
The clearest case of regulatory trigger alignment. Stablecoins had 128 months of institutional development before the GENIUS Act (July 2025) ignited public attention. The attention breakout and the adoption acceleration are simultaneous — driven by the same regulatory event. This is the model case for Policy-Trigger phenomena.
The longest lead time in the set. BNPL had been above the structural attention threshold for four years before the Gallup adoption curve began its steepest ascent. Public awareness was a necessary but far-from-sufficient condition — the conversion from awareness to operational behaviour required a decade of ecosystem development, regulatory clarification, and organisational capacity building.
The Gap Between Attention and Adoption Is Not Noise
In all three cases, public attention preceded operational adoption. But the lead times are very different: Open Banking had a 3-year lead, Stablecoins a near-simultaneous breakout, BNPL a 4-year lead. What explains the variation?
The Gallup study points to regulatory change as the primary driver of Finance AI adoption. The Zorthex data suggests this is consistent with a broader pattern: attention moves when a topic becomes publicly legible; adoption moves when the operational conditions — regulatory, organisational, economic — are ready.
The gap between these two events is not random. It is shaped by:
Organisational capacity: Do workers and institutions have the skills and processes to convert awareness into workflow?
Economic conditions: Is the cost of adoption below the threshold of friction?
The attention signal tells you the technology is ready in the public mind. The adoption signal tells you the operational conditions have caught up. Neither alone is sufficient to understand where a technology actually is in its diffusion cycle.
Two Keys, Not One
Across the three finance cases analysed, attention consistently preceded adoption. The attention signal is a leading indicator — not of how much adoption will occur, but of when the conditions for adoption have become publicly legible.
Neither signal is sufficient alone. Attention without adoption data tells you that a technology is known but not whether it is being used. Adoption data without attention data tells you behaviour has changed but not how long the awareness infrastructure was already in place.
Together, they map the full diffusion arc — from the moment a technology becomes thinkable to the moment it becomes operational.
What This Analysis Does Not Claim
Why Not Social Media?
A frequent question: why doesn't Zorthex use social media data — Twitter/X impressions, LinkedIn engagement, TikTok views, Instagram reach — as an attention source? The answer is architectural, not practical.
Zorthex measures intentional attention — a person decides to search for something (Google Trends) or decides to read about it (Wikipedia). Both are acts of cognitive demand: the person initiates the query. This is fundamentally different from passive exposure, where an algorithm decides what appears in your feed and you may scroll past it in 0.3 seconds.
The distinction matters for what the framework claims to measure. When Google Trends shows "stablecoins" rising from 5/100 to 100/100, that means more people are choosing to type the word and look for information. It's a revealed preference. When a LinkedIn post about stablecoins gets 50,000 impressions, that means an algorithm decided 50,000 people should see it. The first is demand. The second is supply.
Google Trends — person types a query. Cognitive demand.
Wikipedia — person navigates to a page. Informational demand.
Reddit — person enters a community discussion. Declared qualitative.
Reproducible. CSV locked, timestamped, archived. Anyone can re-run the same query and verify.
Resistant to manipulation. No paid promotion on Google Trends. No bots on Wikipedia pageview counts.
Measures demand. What the public wants to know.
Twitter/X impressions — algorithm shows content. Algorithmic supply.
LinkedIn engagement — feed placement, not search. Platform-mediated.
TikTok/Instagram — passive scroll. Sub-second exposure ≠ attention.
Non-reproducible. Feeds are personalised. No two users see the same content. No locked CSV possible.
Manipulable. Bots, paid amplification, engagement farming, viral mechanics all inflate metrics.
Measures supply. What algorithms decide you should see.
This report demonstrates why the distinction is operationally critical. The Gallup adoption data shows when people use a technology. The Zorthex attention data shows when people choose to learn about a technology. Social media data would show when people were shown content about a technology. Only the middle signal — intentional attention — has a documented empirical relationship with the adoption signal. Passive exposure correlates with virality, not with the sustained cognitive engagement that precedes adoption.
There is a second, practical reason. Google Trends data is normalised, timestamped, and freely available. Wikipedia pageviews are published via an open API with daily granularity. Both can be downloaded, archived, and independently verified by anyone. Social media metrics are platform-proprietary, API-gated, subject to retroactive changes, and increasingly paywalled. A framework that depends on social data inherits the platform's instability. A framework built on search and encyclopaedic data inherits the open web's durability.
The design choice is deliberate: Zorthex measures what people seek, not what they are shown. The difference between these two signals is the difference between a thermometer and a heater. One measures temperature; the other changes it. Social media is a heater. Zorthex reads the thermometer.
Data Sources
Zorthex attention data: Google Trends CSVs locked 23 June 2026; Wikipedia pageviews via Wikimedia REST API; methodology at zorthex.com/methodology; dataset at zorthex.com/dataset.
Adoption reference: Gallup Workplace Panel 2023–Q1:2026, published via NBER-SAIF Conference on AI and Financial Markets, Spring 2026. The panel tracks AI use frequency among US workers aged 20–65 across sectors, quarterly.
Regulatory trigger reference: Decline in finance regulatory restriction density 2024–2025 as identified in the Gallup study — associated with approximately 4.7 percentage-point increase in frequent AI use in Finance.
All Zorthex CSVs available for independent verification at github.com/zorthex2026/zorthex-diffusion-lag.