Dashboard

Live data

What 74,062 tasks
reveal about automation

Every assessment decomposes a job into individual tasks, then scores each one for automation likelihood. This page surfaces patterns across the full dataset — what's going first, what's staying human, and how it varies by sector.

74,062
tasks analyzed
10,223
assessments completed
418
professions covered
29%
avg automation exposure

What's going first

10 most automatable task types

Ranked by average effective automation score — tasks that are already digital, routine, and require little human judgement or social presence.

1 Updating and maintaining HR policies, procedures, and documentation in compliance with labor laws 91%
2 Analyzing design metrics, user behavior data, and A/B test results to inform design decisions 90%
3 Entering supplier shipment data into ERP/inventory management systems; reconciling purchase orders with receipts and invoices 90%
4 Gather financial data from multiple sources (ERP systems, databases, spreadsheets) and consolidate into analysis models 89%
5 Recording trip data, mileage logs, fuel purchases, or incident reports 89%
6 Write SQL queries, Python scripts, or oversee complex analytical work on ad-hoc business problems 89%
7 Email, calendar management, and administrative coordination 89%
8 Recording and documenting test results in digital laboratory information systems (LIS) for physician review 89%
9 Processing purchase orders, invoices, and maintaining procurement database records 89%
10 Maintain and update translation memory databases or term bases to improve efficiency and consistency for future projects 89%

What's staying human

10 least automatable task types

Tasks scoring near zero: mentoring, negotiation, cross-functional collaboration, and anything where being physically present and socially accountable is the whole point.

1 Mentoring junior developers through pair programming, 1-on-1s, and guidance 1%
2 Mentoring and coaching employees on innovation methodologies, such as design thinking, lean startup, or agile, to foster a culture of innovation within the organization 1%
3 Participate in design critiques, give and receive feedback from peers, and iterate on designs based on team input 1%
4 Stakeholder and client relationship management (meetings, presentations, negotiations, feedback loops) 1%
5 Collaborate with IT architects, developers, and business teams to implement knowledge management systems and ensure adoption 1%
6 Collaborate with product managers and engineers in meetings to define requirements, scope, and technical feasibility of designs 1%
7 Cross-functional collaboration: Meet with business unit leaders, understand operational pain points, scope new system implementations, and align IT delivery with business priorities 1%
8 Collaborate with department heads to identify high-impact data use cases and prioritize projects 1%
9 Attend and lead cross-functional meetings (supply chain, finance, HR, production); make decisions on resource allocation and priorities 1%
10 Conducting one-on-one meetings with managers to discuss employee performance, engagement issues, and coaching needs 1%

Task composition

How tasks break down

Every task is classified across two axes: whether inputs and outputs are already digital, and whether human social presence is integral to the work.

Input × Output modality
DD — digital in, digital out
41%
Highest automation exposure
AD — analog in, digital out
26%
Moderate exposure
DA — digital in, analog out
8%
Moderate–low exposure
AA — analog in, analog out
25%
Lowest automation exposure
Social dependency

Tasks where human presence is integral to the outcome — trust, accountability, emotional attunement — resist automation regardless of technical capability.

By sector

Average automation exposure by sector

Effective automation score averaged across all assessed tasks within each sector. Sectors with fewer than 20 sessions are excluded.

Technology
36%
Administration
36%
Finance
34%
Marketing
33%
Media
33%
Government
31%
Creative
30%
Legal
29%
Other
29%
Logistics
28%
Science
27%
Healthcare
23%
Education
22%
Trades
20%
Hospitality
20%
Social Services
18%
About this data. Scores are derived from a structured classification model applied to each assessed task. The effective automation score accounts for substrate thickness (how much tacit knowledge a task requires) and social dependency (whether human presence is integral). Synthetic sessions — used to reach display thresholds for new professions — are included in aggregate counts but have no effect on task-level rankings, which require a minimum of 5 independent assessments. Dataset last refreshed daily.