Soft Skills That Make Data Professionals Stand Out

Soft skills for data professionals are often treated as an extra, something to mention in a CV summary. The job data disagrees. The US Department of Labor's O*NET database rates how important 35 skills are in each occupation (O*NET 31.0 database). The code below compares data scientists, business intelligence analysts and data warehousing specialists with software developers.

import pandas as pd

# O*NET 31.0 essential and transferable skills: importance 1 (not important) to 5 (extremely important), rated by analysts
ts = pd.concat([pd.read_csv(f"onet/{f}.txt", sep="\t") for f in ("Essential Skills", "Transferable Skills")])
roles = {"15-2051.00": "data scientist", "15-2051.01": "BI analyst", "15-1243.01": "data warehousing",
         "15-1252.00": "software developer"}
imp = ts[ts["Scale ID"].eq("IM") & ts["O*NET-SOC Code"].isin(roles)].pivot_table(
    index="Element Name", columns="O*NET-SOC Code", values="Data Value").rename(columns=roles)[list(roles.values())]
print(f"S1 importance of skills (1-5), {len(imp)} skills, sorted by the data roles' average")
imp["data roles"] = imp[["data scientist", "BI analyst", "data warehousing"]].mean(axis=1)
imp["vs developer"] = imp["data roles"] - imp["software developer"]
print(f"   {'skill':<34}{'data sci':>9}{'BI':>6}{'DW':>6}{'dev':>6}{'data - dev':>12}")
for skill, r in imp.sort_values("data roles", ascending=False).head(12).iterrows():
    print(f"   {skill:<34}{r['data scientist']:>9.2f}{r['BI analyst']:>6.2f}{r['data warehousing']:>6.2f}"
          f"{r['software developer']:>6.2f}{r['vs developer']:>+12.2f}")

print("\nS2 communication and people skills: data roles vs software developers")
rank = imp["data roles"].rank(ascending=False, method="min")
for skill in ["Speaking", "Writing", "Active Listening", "Social Perceptiveness", "Persuasion", "Coordination", "Instructing", "Negotiation"]:
    r = imp.loc[skill]
    print(f"   {skill:<22} data roles {r['data roles']:.2f} (rank {rank[skill]:>2.0f} of {len(imp)}) | developer {r['software developer']:.2f} | {r['vs developer']:+.2f}")

# O*NET work styles: distinctiveness rank 1-10 marks styles that set an occupation apart (hybrid AI-expert ratings)
ws = pd.read_csv("onet/Work Styles.txt", sep="\t")
print("\nS3 work styles that most distinguish each role (distinctiveness rank 1 = most)")
for code, name in list(roles.items())[:3]:
    w = ws[ws["O*NET-SOC Code"].eq(code) & ws["Scale ID"].eq("DR") & (ws["Data Value"] > 0)].sort_values("Data Value")
    print(f"   {name}: " + ", ".join(w["Element Name"].head(5)))

S1 importance of skills (1-5), 35 skills, sorted by the data roles' average
   skill                              data sci    BI    DW   dev  data - dev
   Critical Thinking                      4.25  3.88  3.88  3.88       +0.12
   Reading Comprehension                  3.88  4.00  3.88  3.50       +0.42
   Active Listening                       3.50  3.88  3.50  3.38       +0.25
   Speaking                               3.62  3.88  3.38  3.12       +0.51
   Mathematics                            4.38  3.50  3.00  2.75       +0.88
   Complex Problem Solving                3.75  3.38  3.62  3.38       +0.20
   Active Learning                        3.62  3.88  3.00  3.50       +0.00
   Writing                                3.50  3.75  3.25  3.25       +0.25
   Judgment and Decision Making           3.38  3.62  3.38  3.62       -0.16
   Programming                            3.75  2.50  3.75  4.00       -0.67
   Systems Analysis                       2.75  3.12  3.50  3.50       -0.38
   Monitoring                             3.25  3.12  2.88  3.00       +0.08

S2 communication and people skills: data roles vs software developers
   Speaking               data roles 3.63 (rank  3 of 35) | developer 3.12 | +0.51
   Writing                data roles 3.50 (rank  7 of 35) | developer 3.25 | +0.25
   Active Listening       data roles 3.63 (rank  3 of 35) | developer 3.38 | +0.25
   Social Perceptiveness  data roles 2.92 (rank 16 of 35) | developer 2.88 | +0.04
   Persuasion             data roles 2.79 (rank 17 of 35) | developer 2.75 | +0.04
   Coordination           data roles 3.00 (rank 14 of 35) | developer 3.00 | +0.00
   Instructing            data roles 2.79 (rank 18 of 35) | developer 2.88 | -0.09
   Negotiation            data roles 2.46 (rank 21 of 35) | developer 2.62 | -0.16

S3 work styles that most distinguish each role (distinctiveness rank 1 = most)
   data scientist: Innovation, Achievement Orientation, Intellectual Curiosity, Integrity, Attention to Detail
   BI analyst: Achievement Orientation, Integrity, Attention to Detail, Dependability
   data warehousing: Intellectual Curiosity, Cautiousness, Integrity, Attention to Detail, Dependability

What the ratings show

Thinking comes first. Critical thinking was the most important skill for data roles on average, with data scientists rating it 4.25 out of 5. Reading comprehension came second, and mathematics was important for data scientists above all (4.38).

Communication outranks programming. Speaking and active listening tied for third place for data roles, ahead of programming. Writing came seventh. For business intelligence analysts, speaking, listening and writing all scored higher than programming.

Data roles need more communication than developers. Speaking was rated half a point higher for data roles than for software developers (+0.51), and writing and active listening a quarter of a point higher. Data work exists to inform decisions made by other people, so explaining and listening are part of the job, not decoration.

Other people skills were about equal. Social perceptiveness, persuasion and coordination were rated similarly for data roles and developers, in the middle of the 35 skills.

Distinctive work styles. O*NET also marks the work styles that distinguish each occupation from others. Integrity and attention to detail appeared for all three data roles; intellectual curiosity for data scientists and data warehousing specialists; cautiousness for data warehousing. These ratings come from a newer hybrid AI-expert method (O*NET: work styles), so read them as a well-informed description rather than a measurement.

Employers say the same thing in surveys. In the World Economic Forum's 2025 report, analytical thinking was the most sought-after core skill, considered essential by seven out of ten companies (WEF: Future of Jobs 2025).

The soft skills that matter, and how to build them

1. Asking the right question (critical thinking)

Before touching data, restate the request as a decision: "You want to know whether to extend the discount to all regions?" Many wasted analyses answer a question nobody asked. Practise by writing the decision, the metric and what would change your mind at the top of every analysis.

2. Listening to understand the problem

Active listening means asking follow-up questions until you understand what the stakeholder will do with your answer. Summarise back what you heard. In interviews, this is often tested with deliberately vague case questions.

3. Explaining results simply (speaking)

Lead with the answer, then the evidence, then the caveats. One chart that answers the question beats ten that describe the data. Presenting data to non-technical stakeholders covers this in depth.

4. Writing that others can act on

Short summaries, clear README files, decision logs and well-commented queries. Writing scales your work: people use your analysis without needing you in the room.

5. Integrity with numbers

Report what the data shows, including uncomfortable results, and state uncertainty honestly. A reputation for straight answers is the most valuable thing a data professional can have, and integrity appears among the distinctive work styles of all three data roles.

6. Attention to detail without losing the big picture

Check joins, totals and filters before you share. Then step back and ask whether the result makes sense. Most embarrassing data mistakes are caught by one of those two habits.

How the emphasis changes with seniority

• Junior: listening carefully, asking clarifying questions early, and writing clear updates. Nobody expects you to influence executives, but everyone notices when you understood the request the first time and kept people informed without being chased.
• Mid-level: explaining results to non-technical audiences, managing expectations about timelines and data quality, and pushing back politely when a request does not make sense.
• Senior and lead: persuading people to act on evidence, resolving disagreements between teams about definitions and priorities, and teaching others. At this stage your soft skills multiply the output of everyone around you.

A practical habit at every level: after each meeting where you presented, ask one trusted colleague what was unclear. Feedback on communication is rare unless you ask for it, and a mentor can speed this up considerably.

How to show soft skills to employers

• In a CV, describe outcomes that needed them: "presented churn findings to the sales leadership team; two recommendations were adopted".
• In a portfolio, write each project's summary for a non-technical reader.
• In interviews, prepare stories using the STAR structure, as covered in behavioural interview questions for data roles.
• At work, volunteer to present your team's results; it is the fastest way to practise and to be noticed.

Communicating findings is part of every project in the Data Analytics course.

Sources

• O*NET 31.0 Database (O*NET Resource Center) — O*NET Resource Center
• Work Styles file, O*NET 31.0 data dictionary (O*NET Resource Center) — O*NET Resource Center
• Future of Jobs Report 2025 (World Economic Forum) — World Economic Forum