Excel and Power BI or Python? What Kenyan Employers Ask For

McTaba runs coding and data bootcamps in Kenya. This guide is written to be useful whether or not you ever join one.

Picture a graduate in Roysambu with a laptop and a free evening. YouTube offers an Excel course, a Power BI course and a Python course. All three are free. Which one gets her an interview?

Opinions on that are cheap. Job ads are evidence. So instead of repeating what a trainer told us, we opened 62 Kenyan data job ads and counted which tools they name.

The count is small, dated and imperfect. Below we say exactly how we did it, so you can repeat it next month and see whether the answer has moved. For the wider career map, start with the data science career guide for Kenya.

We counted on 3 November 2026. We searched MyJobMag Kenya for five terms: "data analyst", "data scientist", "data engineer", "business intelligence" and "machine learning". We opened every ad those searches returned and read the full description.

That gave 73 distinct postings (the same role posted twice by one employer counted once). We then removed 11 that were not data jobs despite the keyword: two university lecturer posts, a data centre civil works post, two data protection posts, two software engineer posts, two engineer posts at an outsourcing firm, a data capture clerk role and a consultancy expert post. That left 62 ads, posted between August and 3 November 2026 (25 in August, 15 in September, 20 in October and 2 in November, by the month on each ad).

A tool counts as "mentioned" if its name appears anywhere in the ad, as a requirement or as "an added advantage". We did not weigh how important each tool was. Our search used the text of the ad, so a tool name inside a long employer blurb would also count.

The other boards, and why they are not in the tally:

• BrighterMonday: its search pages for "data analyst" and "data engineer" returned one ad each when we looked (data analyst search, data engineer search, checked on 3 November 2026). One of them was an old Co-operative Bank ad with a 2015 deadline. Too thin to count.
• LinkedIn Jobs Kenya: the public page showed a heading of 230 data analyst jobs but listed only 60, many of them research, monitoring and finance roles rather than data roles (LinkedIn data analyst jobs in Kenya, checked on 3 November 2026). Without logging in we could not open each ad, so we used it only to see which employers hire. That cut is in our look at where data jobs are posted.
• Fuzu: no public search page we could read without an account, so it is not in the sample.

MyJobMag search pages used: data analyst, data scientist, data engineer, business intelligence and machine learning, all checked on 3 November 2026.

The limits are real. This is one board, 62 ads, three months. Some August ads may have closed. A different week or a different board could shift every percentage by several points. Treat the numbers as a snapshot, not a law.

Tool	Ads naming it (of 62)	Share
SQL (including MySQL and PostgreSQL)	45	73%
Python	41	66%
Power BI	24	39%
Excel	18	29%
Tableau	17	27%
R	8	13%
Looker, Qlik or Looker Studio	4	6%

Sources: MyJobMag data analyst search, data scientist search, data engineer search, checked on 3 November 2026. Counts are ours, from reading each ad.

Seven of the 62 ads named none of these tools. Four named Excel and nothing else.

Two things surprised us. Excel is lower than most trainers claim. And SQL, the unfashionable one, beats everything. If you learn one thing from this page, learn SQL. We wrote a SQL for data analysis guide for exactly this reason.

Want a structured path through SQL, Python and Power BI instead of three YouTube playlists? Look at the Data Science to AI Engineering programme.

One list hides the split. We grouped the 62 ads by title: 26 analyst ads, 15 data scientist ads, 19 data engineer ads and 2 business intelligence ads.

Tool	Analyst ads (26)	Scientist ads (15)	Engineer ads (19)
SQL	14	13	17
Python	10	14	16
Power BI	15	6	3
Excel	14	2	1
Tableau	12	2	2
R	2	5	1

Sources: MyJobMag data analyst, data scientist and data engineer searches, checked on 3 November 2026. The two business intelligence ads are in the totals above but not in this table.

Read it row by row.

• Analyst ads split across Power BI (15 of 26), SQL and Excel (14 each) and Tableau (12). Python appeared in 10. An analyst ad is the one place Excel still matters.
• Scientist ads are Python and SQL almost every time (14 and 13 of 15). Power BI shows up in 6, usually as "reporting to stakeholders".
• Engineer ads are SQL and Python (17 and 16 of 19). Excel barely appears. Cloud names (Azure Databricks, Snowflake) take its place.

So the honest answer to "Power BI or Python" is: which seat do you want? If you want the analyst seat in a retailer, a hospital or a distributor, Power BI plus SQL plus Excel is the pattern. If you want the scientist seat at a bank or an insurer, Python plus SQL is the pattern.

A tally of keywords is not a hiring forecast. Three cautions.

• Mentioned is not required. "Knowledge of Python or R is an added advantage" and "must write production Python" both count as one mention. Open a few ads yourself and read the wording.
• Job ads are a lagging signal. Many Kenyan hires happen through referrals and are never advertised. Ads show what HR departments write, which can trail what the team actually uses.
• Entry-level ads are scarce. Only six of the 62 titles carried a word like junior, intern or assistant. Most ads ask for years of experience. 52 of the 62 ads state years of experience, and the middle lower bound is three years (MyJobMag, checked on 3 November 2026). Your portfolio has to stand in for those years.

That last point is why we link tools to projects. A Power BI dashboard built on a real Nairobi shop's sales is a better answer to "two years of experience" than a certificate. See the portfolio projects that get interviews.

The tally points to an order. It is not the only order, but it matches what we saw in the ads.

1. SQL first. It was in 73% of the ads we counted (MyJobMag, checked on 3 November 2026) and it appears in every role. A few weeks on SELECT, JOIN, GROUP BY and window functions pays off whichever way you go.
2. Then pick one of two forks. Analyst fork: Excel pivot tables, then Power BI. Scientist or engineer fork: Python with pandas, then a machine learning library.
3. Add the second fork later. Analysts who add Python become more flexible. Scientists who add Power BI can talk to finance and operations. Our guide for moving from Excel to Python covers the first step, and the Python for data science beginner guide covers the second.
4. Tableau and R last, unless an employer you want names them. Both appeared in under a third of ads.

If you are still choosing between the tools themselves, read our comparison of Power BI, Tableau and Python.

We should compare ourselves with the evidence, since we sell a course. The Data Science to AI Engineering programme starts on 11 January 2027 and costs KES 140,000. It runs 5 months full time or 8 months part time. Its tool list includes Python, SQL and PostgreSQL, Pandas, Power BI and Streamlit.

That matches the top three tools in the tally. It does not list Tableau, and Excel is not a module of its own. If your target employer is an Excel and Tableau shop, a shorter self-study route may serve you better, and we say so plainly.

No school can promise a job, including us. The programme's first cohort has not started, so we have no graduate outcomes to show you. What we can show you is the syllabus. Compare it with the ads above, then decide.

If the syllabus fits the jobs you are aiming at, read about the Data Science to AI Engineering programme and book a call.