AI for data teams: which tasks to automate first
Updated
Across the 418,730 US jobs in data and analytics, the work scores 64 of 100 for AI exposure, weighted by how many people do each job. AI can take on about 1,016 hours a year per full-time person, 51% of a working year. The task that frees the most hours across the team is to clean and manipulate raw data, about 100 hours a year for each person who does it.
- AI exposure, weighted by employment
- 64 out of 100 High exposure
- Hours AI can take on
- 1,016 hours a year per full-time person
- Task time AI can do today
- 30% AI speeds up another 65%
Tasks to automate first
The 15 tasks where AI can take on the most hours across data and analytics: how exposed each task is, how much of the job it takes and how many people do that job. Hours are per person in the role that does the task. How we score each task.
| Task and who does it | AI exposure |
|---|---|
| Clean and manipulate raw data using statistical software. Cleaning and reshaping raw data is highly automatable with code assistants. Done by data scientists. | 80% AI can do it 100 hours a year for each person in the role |
| Write new functions or applications in programming languages to conduct analyses. Code assistants write most analysis functions; developer reviews and tests output. Done by data scientists. | 80% AI can do it 100 hours a year for each person in the role |
| Create graphs, charts, or other visualizations to convey the results of data analysis using specialized software. AI generates charts and dashboards code quickly; human checks accuracy. Done by data scientists. | 75% AI can do it 94 hours a year for each person in the role |
| Prepare management reports defining and evaluating problems and recommending solutions. AI drafts report text; analyst checks analysis and recommendations. Done by operations research analysts. | 62% AI can do it 79 hours a year for each person in the role |
| Review research literature. AI searches and summarizes papers quickly; analyst checks key sources. Done by operations research analysts. | 72% AI can do it 72 hours a year for each person in the role |
| Formulate mathematical or simulation models of problems, relating constants and variables, restrictions, alternatives, conflicting objectives, and their numerical parameters. AI helps code models; analyst defines constraints and objectives. Done by operations research analysts. | 55% AI helps 68 hours a year for each person in the role |
| Report results of statistical analyses, including information in the form of graphs, charts, and tables. Code tools generate charts, tables, and summary text quickly. Done by statisticians. | 72% AI can do it 86 hours a year for each person in the role |
| Prepare data for processing by organizing information, checking for inaccuracies, and adjusting and weighting the raw data. Code tools handle cleaning, error checks, and routine weighting well. Done by statisticians. | 68% AI can do it 80 hours a year for each person in the role |
| Process large amounts of data for statistical modeling and graphic analysis, using computers. Code tools process large datasets and generate graphics reliably. Done by statisticians. | 72% AI can do it 77 hours a year for each person in the role |
| Review, classify, and record survey data in preparation for computer analysis. Coding, cleaning, and recording responses is highly automatable routine work. Done by survey researchers. | 85% AI can do it 100 hours a year for each person in the role |
| Prepare and present summaries and analyses of survey data, including tables, graphs, and fact sheets that describe survey techniques and results. AI drafts summaries, tables, and charts quickly; humans check results. Done by survey researchers. | 70% AI can do it 96 hours a year for each person in the role |
| Conduct research to gather information about survey topics. Literature reviews and web research are well suited to AI tools. Done by survey researchers. | 75% AI can do it 94 hours a year for each person in the role |
| Enter data into computers for use in analyses or reports. Data entry from digital or scanned sources is highly automatable. Done by statistical assistants. | 88% AI can do it 158 hours a year for each person in the role |
| Code data prior to computer entry, using lists of codes. Assigning codes from a list to responses is well suited to AI. Done by statistical assistants. | 80% AI can do it 135 hours a year for each person in the role |
| Compute and analyze data, using statistical formulas and computers or calculators. Statistical software and LLMs run formulas and summarize results quickly. Done by statistical assistants. | 72% AI can do it 134 hours a year for each person in the role |
What stays with people
5% of the task time across these roles needs a person, and the output of the rest still needs someone to check it. The tasks below take the most of that time.
- 30% AI can do it
- 65% AI helps
- 5% Needs a person
- Collaborate with others in the organization to ensure successful implementation of chosen problem solutions. Operations research analysts, 15% exposed
- Observe the current system in operation, and gather and analyze information about each of the component problems, using a variety of sources. Operations research analysts, 15% exposed
- Supervise and provide instructions for workers collecting and tabulating data. Statisticians, 20% exposed
- Direct and review the work of staff members, including survey support staff and interviewers who gather survey data. Survey researchers, 15% exposed
- Confer with researchers, clinicians, or information technology staff to determine data needs and programming requirements and to provide assistance with database-related research activities. Bioinformatics technicians, 15% exposed
Adoption path
Start with the tasks where AI helps a person, then automate the routine ones, then give whole workflows to agents with a person approving the output.
Roles in data and analytics
10 roles, each linked to its task-by-task score. Of the roles with at least 2% of the jobs, data scientists score highest at 67, operations research analysts lowest at 58.
- Data scientists, 62.7% of the team, score 67 out of 100
- Operations research analysts, 25.9% of the team, score 58 out of 100
- Statisticians, 6.9% of the team, score 63 out of 100
- Survey researchers, 2% of the team, score 59 out of 100
- Statistical assistants, 1.1% of the team, score 71 out of 100
- Bioinformatics technicians, 0.9% of the team, score 66 out of 100
- Mathematicians, 0.5% of the team, score 58 out of 100
- Biostatisticians, Counted under a related job in US data, score 60 out of 100
- Clinical data managers, Counted under a related job in US data, score 64 out of 100
- Data analysts, Counted under a related job in US data, score 65 out of 100
Questions about AI for data teams
Which tasks should a data team automate first?
The first task to automate is to clean and manipulate raw data (80% exposed). Next come the tasks to write new functions or applications in programming languages to conduct analyses (80%) and to create graphs, charts (75%). Together they free the most hours across the team.
How many hours can AI save a data team?
About 1,016 hours a year per full-time person, 51% of a 2,000-hour working year. For a team of 10 that is about 10,160 hours, the time of 5.1 full-time people. The figure is the share of task time AI can remove today, weighted by how many people do each job.
Which roles in a data team are most exposed to AI?
Of the roles with at least 2% of the jobs, data scientists score highest, at 67 of 100, and operations research analysts lowest, at 58. Each role's page lists its tasks one by one.
Will AI replace jobs in data and analytics?
AI can do 30% of the task time across these roles today. Another 65% gets faster with AI while a person stays in charge. 5% needs a person. One task that stays with people is to collaborate with others in the organization.
Where should a data team start with AI?
With the quick wins: approved assistants for the tasks AI helps with, under a written AI policy. Then automate the 14 tasks on this page AI can do end to end. An AI audit maps the tasks in your team and ranks them the same way this page does.
Next steps
Find the tasks to automate first in your team
We map the tasks in your data team, score each one the way this page does and build the automations that pay back first.