You can learn the basics of data analysis with a mid-range phone, a few gigabytes a month and a notebook. You won't do it by streaming hours of video. You'll do it by downloading lessons once, practising on small datasets, and building two or three projects you can show people.
This guide gives you a plan that assumes patchy power, costly data and no laptop at the start. If you have a laptop later, everything here still applies.
What data analysis actually involves
Data analysis means taking a table of records, cleaning it, answering a question with it, and explaining the answer clearly. A shop owner asks, "Which products sell best on weekends?" You count, group, compare and report.
The core skills are the same in almost every entry-level role:
- Spreadsheets: sorting, filtering, formulas, pivot tables, simple charts.
- Cleaning data: fixing duplicates, blank cells, inconsistent spellings (for example "Lagos", "lagos " and "LAGOS").
- Basic statistics: averages, medians, percentages, and knowing when an average misleads.
- Communication: a chart and three sentences that a busy manager understands.
SQL and Python come later. Many beginners rush to them and skip the spreadsheet stage, then struggle because they can't tell whether their results make sense.
Step 1: Set up your phone for low-data learning
- Pick one spreadsheet app from the free ones available on your phone's app store. Choose one that opens common spreadsheet files (.xlsx and .csv). Check that it works offline.
- Turn off automatic app updates on mobile data. Set updates to Wi-Fi only so they don't eat your bundle quietly.
- Create a "Learning" folder for downloaded files, notes and screenshots.
- Find your cheap data window. Many networks in Nigeria, Ghana, Kenya and South Africa offer cheaper night or off-peak bundles. Check your provider's current offers, because they change often. Do large downloads then.
- Charge smart. Study when you have power, and keep a power bank for outages. Dim the screen and use battery saver while practising offline.
Step 2: Choose a few free sources, then download
Don't collect twenty courses. Choose one main course and one backup. Look for these features:
- Lessons you can download, or text-based tutorials that load light.
- Practice files you can download and work on offline.
- Exercises, not only video.
Many free learning sites, public-library style platforms and spreadsheet help pages publish written tutorials. Text uses far less data than video. If you use video, set quality to the lowest setting that's readable, and download over Wi-Fi at a friend's place, a campus, a library or a hub when you can.
Beware of "free" courses that hide a payment before you get a certificate. A certificate matters far less than your projects. If you want help judging paid programs, read How to Spot a Tech Bootcamp That Isn't Worth the Fee.
Step 3: A four-week study plan
Aim for 45 to 60 minutes a day, five days a week. If you can only manage three days, stretch the plan to eight weeks. Consistency matters more than length.
Week 1: Spreadsheet basics
- Enter data, format columns, freeze the header row.
- Sort and filter.
- Learn SUM, AVERAGE, COUNT, MIN, MAX.
- Practice: make a table of your own weekly spending and total it by category.
Week 2: Cleaning and lookups
- Remove duplicates and trim extra spaces.
- Use IF, COUNTIF and SUMIF.
- Learn a lookup function (your app's help page will name the current version).
- Practice: take a messy list of 50 names and phone numbers and tidy it.
Week 3: Pivot tables and charts
- Group data by category, month or location.
- Make a bar chart and a line chart. Label axes and give each chart a plain title.
- Practice: summarise a sales list by product and by day.
Week 4: One real project
- Pick a question, find or collect data, clean it, analyse it, and write a one-page summary.
Step 4: Find data you can practise on
Real data is better than invented numbers. Options that cost little:
- Your own life or business: airtime and data spending, a provision shop's sales, a church or association's contributions, your study hours versus test scores.
- Public datasets: national statistics offices and open data portals publish tables you can download as CSV. Search for the statistics office of your country, then look for downloads.
- Small surveys: ask 30 classmates or neighbours three questions through a free form tool and analyse the answers. Get consent, and don't collect names, phone numbers or other personal details you don't need.
Step 5: Use AI as a tutor, not a shortcut
General AI chat tools can explain a formula, suggest why your pivot table looks wrong, or create practice questions. Use them for three things:
- Explain: "Explain COUNTIF to a beginner using a shop sales example."
- Debug: paste the formula and the error message and ask what's wrong.
- Quiz: ask for five practice questions, attempt them, then ask for feedback on your answers.
Two rules. First, never paste private data such as customer phone numbers, bank details, BVN or NIN into an AI tool. Replace them with dummy values. Second, check the answers yourself. AI tools sometimes give confident but wrong formulas or statistics. If you can't explain the result, you haven't learned it yet. For better questions, see Prompting Basics for Everyday Work.
Text chats use little data, so this is one of the cheapest ways to get one-to-one help.
A worked example (hypothetical)
Say Ada runs a small phone-accessories stall and writes sales in an exercise book. Over four weeks, she types 120 rows into a spreadsheet: date, item, quantity, price, payment method.
- She fixes spellings so "charger" and "Charger " match.
- She adds a column: total = quantity × price.
- She builds a pivot table by item and sees chargers bring in 40% of revenue, while cases bring in 10% but take up most of her shelf.
- She groups by weekday and sees Fridays and Saturdays are strongest.
- She writes a summary: "Stock more chargers before the weekend. Reduce case variety."
The numbers are made up, but the process is exactly what a junior analyst does. Ada now has a project: a one-page summary, two charts and a description of what she did.
Step 6: Turn practice into proof
Employers and clients want evidence. Build two or three small projects and present each one with:
- The question you asked.
- Where the data came from.
- What you cleaned.
- Two or three charts.
- A short recommendation.
Export as a PDF, which is light to send on WhatsApp or email. If you're starting from zero, our guide to building a freelance portfolio with no experience shows how to present this kind of work.
Common mistakes
- Course hopping. Finish one course before starting another.
- Watching without doing. If you haven't opened the spreadsheet, you haven't studied.
- Skipping cleaning. Real data is messy, and employers know it.
- Jumping to Python on day one. Spend the first month in spreadsheets.
- Using private data. Anonymise anything that belongs to other people.
- Paying for a certificate you don't need. Projects count for more at entry level.
When to move beyond spreadsheets
After a month or two, add SQL, which lets you query large tables. Many free practice sites let you write queries in a mobile browser, though typing on a phone is slow, so a cheap Bluetooth keyboard or a borrowed computer helps. After SQL, you can try a dashboard tool and basic Python. Treat these as months three to six, not week one.
Frequently asked questions
Can I really learn this without a laptop?
You can learn spreadsheet basics, cleaning and charting on a phone. It's slower, and some tools work better on a computer. Use a laptop at a hub, school or café when you need it for SQL or Python.
How much data will I need?
It depends on your sources. Text lessons and spreadsheets use little. Video uses the most. Download once, study offline, and track your usage in your phone settings for a week to see what you really spend.
Do I need a statistics or maths degree?
No. You need comfort with percentages, averages and reading charts. You can learn them as you practise.
How long before I can get paid work?
It varies with your time, the market and your portfolio. A realistic aim is to complete three small projects first, then apply for internships, assistant roles or small freelance tasks such as cleaning spreadsheets for small businesses.
Your next step today
Open a blank spreadsheet and enter your last seven days of spending: date, category, amount. Total it by category. That's your first dataset, and it costs no data at all.
Results from any side hustle or online work vary widely, and nothing here is a promise of income. Be wary of anyone who asks you to pay upfront to start earning.