How to learn Excel
This path assumes you already have Excel and need to get useful quickly, not become a modeller. It gets you from raw data in tables to lookups and pivot tables, then to the harder judgement: when a database or a script should replace the workbook.
Hours
50-70 hours
Calendar
about 10 weeks at 5-7 hours a week
Stages
6
Start from
Nothing beyond a working copy of Excel and basic computer use. You do not need advanced maths, programming, or macro experience; the main false prerequisite is believing you need all the menus memorised before you can analyse data.
Why the usual order fails
Excel is not one giant application to learn in order. The parts that matter for most jobs are a small set: clean data in tables, a handful of lookup and summing formulas, and pivot tables that turn rows into totals. The common order fails because it starts with the grid, formatting and long formula lists. People memorise functions they will use once, while data lives in merged cells and text-like numbers. After weeks, they can make a sheet look busy but cannot answer a question from a client extract. The real shape is smaller and less glamorous: structure data, ask questions of it, then present the answer.
This path puts data structure first. In week one you convert a plain range to an Excel Table, sort and filter it, and name cells before doing any serious formatting. Most plans have you formatting cells or typing sample budgets. Then you learn the formulas that respect that structure: XLOOKUP, SUMIFS, COUNTIFS, IFERROR and the odd text function. Only after lookups and pivot tables are working do you add charts and validation. The final stage is deliberate: a set of failing cases where the workbook is the wrong tool, so you learn to recognise them instead of adding another fragile sheet. That split is what turns spreadsheet users into analysts.
The path, in the order it has to be walked.
Each stage assumes the one above it and nothing below it. The check at the foot of each is the point — you are meant to fail it, go back, and try again before moving on.
Tables and data structure
weeks 1-2Stop treating sheets as visual layouts and start treating rows as records. Excel Tables turn a range into a named, filterable object with structured references. This sits here because every later formula and pivot table assumes clean rectangular data. Without it, lookups break and pivots need manual rebuilding.
- Excel Tables (Ctrl+T)
- structured references
- sorting and filtering
- freeze panes and column widths
- CSV import
- number vs text distinctions
- Flash Fill
You can now: Convert a messy 20-row range into a Table, rename it, and filter it to answer a question in under five minutes.
Core formulas and references
weeks 2-4Move from putting values in cells to writing expressions. Relative and absolute references, named ranges, arithmetic and basic functions are practised on real sales or expense sheets. It sits after table structure because formulas need stable references and correctly typed data. Without this, lookup formulas later feel like magic.
- SUM, AVERAGE, MIN, MAX
- relative and absolute references ($A$1)
- named ranges
- F2 editing and formula auditing
- percentage change
- LARGE and SMALL
- order of operations in formulas
You can now: Recreate a cost table with mixed relative and absolute references and prove the formula copies across and down correctly.
Lookups and data cleaning
weeks 4-6This is the heart of most office Excel work: getting a value from one table into another by a matching ID. XLOOKUP, INDEX/MATCH, and cleaning functions are drilled until you can write them without the wizard. It sits after references because lookups combine exact ranges and locking. People stall here, so it gets two weeks.
- XLOOKUP
- INDEX and MATCH
- VLOOKUP for legacy files
- TRIM, CLEAN, LEFT, RIGHT
- TEXTSPLIT and TEXTJOIN
- IFERROR
- duplicate removal
You can now: Given two client export sheets, return the current price to the first sheet for every row without manually copying a single cell.
Pivot tables
weeks 6-8Pivot tables are where ad hoc questions get answered: split by month, region, product. You build them from clean tables, control them with slicers, group dates, and create calculated fields. This sits after lookups because you need matching and cleaning done first; dirty data makes pivot rows multiply.
- PivotTable
- row, column and value fields
- sum vs count
- slicers and timelines
- grouping dates and numbers
- calculated fields
- pivot refresh and source ranges
You can now: Take 5,000 rows of order data and produce a month-by-category sales pivot with a slicer in under ten minutes.
Logical and text formulas
weeks 8-9Most spreadsheet errors come from edge cases: blanks, mismatched text, two conditions interacting. IF, IFS, AND, OR, SUMIFS, COUNTIFS and text normalisation make sheets decision-ready. This comes after pivots because these formulas usually prepare and check data rather than replace summary tools.
- IF and IFS
- AND and OR
- SUMIFS, AVERAGEIFS and COUNTIFS
- EXACT for text comparison
- conditional formatting with formulas
- data validation lists
- ISNUMBER and ISBLANK
You can now: Add a validation rule and a conditional format that highlights only rows where two conditions fail at once.
When not to use Excel
weeks 9-10Cover output and the uncomfortable boundary. Charts, print layout, data validation and simple what-if analysis give enough to report. Then work through jobs where a database, Power Query or script is a better tool: multi-table merges, repeatable ETL, concurrent use.
- charts and sparklines
- print and page layout
- data validation
- Goal Seek and scenario manager
- Power Query import
- when to use Access or SQL
- row and recalculation limits
You can now: Open a 100,000-row multiple-table task and decide in five minutes whether Excel, Power Query or a database is the right fit, then justify it.
The hours are fixed. The calendar is yours.
The same path at three speeds, and what actually goes wrong at each.
3 hours a week
50-70 hours becomes 17-23 weeks. That is long enough for Excel muscle memory to rust between sessions, so you will spend the first 20 minutes of each sitting re-orienting. It works for the patient but often breaks around lookups unless you practise twice weekly.
8 hours a week
This is the sustainable pace: 7-9 weeks to finish. You can keep one sample workbook alive and revisit it before each session. The main break risk is skipping weekends then trying to compress; pivot table groupings and formula locking need sleep between sessions.
20 hours a week
50-70 hours becomes 3-4 weeks. You can pass checks quickly, but pivot table judgement and the boundary with databases stays shallow because there is not enough elapsed time to meet messy real files. Cramming to a certificate often breaks when data arrives dirty.
Four ways people lose a year on this.
Each of these is appealing for a real reason, which is why they are common. What they cost is the part nobody mentions.
Starting with formatting, not data structure
People open Excel and start merging cells, colouring headers and adjusting fonts before they create a table. It feels productive because the sheet looks finished. The cost is that formulas, sorting and pivot tables later fight with merged ranges, blank rows and text caches. What looked like progress becomes rework.
Using VLOOKUP for every lookup problem
VLOOKUP is taught everywhere because it is familiar, so people force column counting and left-to-right workarounds even when XLOOKUP or INDEX/MATCH would be simpler. They spend more time managing the function's limitations than matching records. It costs them brittle sheets that break when a column is inserted.
Keeping all data in one wide sheet
People build a single worksheet with hundreds of columns, one for every scenario, and then manually link rows. It appeals because everything is visible on screen. But summarising it requires manual formulas and fragile references. The cost is hours of rework and errors that a structured table or separate data sheet would have prevented.
Staying in Excel once the task outgrows it
Once data spans many related tables, concurrent editors or repeatable import steps, people add another hidden sheet, macro or manual copy routine. It is appealing because everyone already has Excel open. The cost is fragile process, version conflicts and slow recalc. At that boundary, a database or script saves more time than another formula.
How to know it worked.
Not a syllabus you tick off — six things to attempt. If one of them is still hard, the stage it belongs to is the one to go back to.
- Convert five rows in a blank workbook to a Table and rename it in under a minute.
- Write a formula with mixed relative and absolute references and copy it across and down.
- Write XLOOKUP from memory to return a customer's region from another sheet.
- Build a pivot table that sums sales by month without using a helper column.
- Write SUMIFS with two criteria on separate columns and verify the count.
- Take a 5,000-row CSV, clean duplicate IDs and produce a month-by-region pivot in 15 minutes.
Questions about learning Excel.
The five we hear most. The intro film answers most of them in a minute and a half.
how long does it take to learn excel for a job?
For the lookups, pivot tables and data cleaning that cover most office jobs, expect 50-70 hours, about ten weeks at five to seven hours a week. You can produce a useful pivot table in half that time, but true working competence includes edge cases and knowing when a workbook is the wrong tool, which needs the longer end.
is excel still worth learning with ai around?
Yes, because AI can suggest a formula but it cannot know whether the source data is clean or the result is right. Roles that receive client extracts, build ad hoc reports, model budgets and check errors still run on Excel. The skill that matters is verification and structure, not typing formulas.
do i need to learn macros and vba to be good at excel?
No. For most jobs, tables, XLOOKUP, SUMIFS, IFERROR and pivot tables matter more than VBA. Macros become worth learning only when you have a repeated production task you can describe step by step and no simpler tool will do. Many people learn VBA early and neglect the data structure that would have removed the need for it.
what excel functions should i learn first?
Learn SUM, AVERAGE and COUNTA first, then IF and IFERROR. The workhorses are XLOOKUP for matching, SUMIFS and COUNTIFS for conditional sums and counts, and TEXTSPLIT or TEXTJOIN for name and address clean-up. Most people can skip VBA and most statistical functions for several years.
is excel good for customer data or project management?
It works for a lightweight customer list or small project backlog. Once several people edit the same rows, or you need permissions, change history, or many linked tables, use a database, CRM or issue tracker. Excel does not enforce row uniqueness or edits well, and shared workbooks create version conflicts.
A path is not a course.
This is the shape of Excel, not the teaching of it. Hand the same path to BiliMind and it writes the lessons onto it — in the form each one calls for, with practice that gets marked and a schedule that bends when your week does. Mapping a course is free and unlimited.
This roadmap was drafted on 18 September 2026 by the same engine that builds BiliMind courses, and reviewed before publishing. Hours are honest estimates for working competence, not for a demo — treat them as a starting point and adjust them against your own first fortnight.