How to Assess Excel Skills in Technical Hiring

The assessment dilemma: Real Excel skill is invisible in resumes

A resume says "Advanced Excel." That could mean:

• They built dashboards in power-user mode (useful).
• They took an online course and remember VLOOKUP (marginal).
• They drag-and-drop data and use Find & Replace (not useful).
• They've never used Excel professionally but watched YouTube tutorials (not useful).

Resumes don't differentiate. Assessments do. But most Excel assessments are badly designed — they're either too easy (shortcuts, trivia) or too artificial (disconnected formula chains with no context).

Tier one: Quick filtering with realistic scenarios

Use brief scenario-based questions that take 10–15 minutes per candidate. These separate "has used Excel meaningfully" from "hasn't."

The data-join test

"You have two spreadsheets:

• Inventory.xlsx: Product ID, Product Name, Stock Quantity
• Sales.xlsx: Product ID, Units Sold (last month)

Merge this data to show Product Name, Stock Quantity, and Units Sold side by side. Show your work."

This tests:

• Ability to think about data structure (matching keys, multiple tables)
• Knowledge of VLOOKUP or INDEX/MATCH
• Practical problem-solving (not hypotheticals)

Scoring: Can they connect data? Yes (hire them for interviews) or no (save time, skip).

The conditional-math test

"A retailer offers volume discounts:

• 1–10 units: full price
• 11–25 units: 10% off
• 26+ units: 20% off

Given a list of order quantities (column A) and unit prices (column B), write a formula that calculates the final price for each order including the appropriate discount."

This tests:

• Nested IF logic or INDEX/MATCH for range lookups
• Mathematical accuracy (applying percentages)
• Attention to boundary conditions

Scoring: Does the formula work correctly for all ranges? Grade as pass/fail.

The data-summarization test

"Given raw sales data with columns: Date, Salesperson, Region, Amount. Summarize total sales by region and salesperson. Any method is acceptable (formulas, pivot table, or manual grouping)."

This tests:

• Problem decomposition (what does "summarize" mean? totals, not individual rows)
• Knowledge of one or more summarization methods
• Practical awareness (most experienced Excel users jump to pivot tables)

Scoring: Is the summary accurate and organized? Yes/no.

Tier two: Depth assessment through take-home problems

Candidates who pass tier one move to a small take-home (30–60 minutes) that mirrors actual work.

Example: Sales performance analysis

"Download this sales export (CSV with 500 rows: Date, Salesperson, Territory, Product, Amount, Quantity).

Required:

1. Calculate total sales and average deal size per salesperson.
2. Identify which territory performed best by total sales.
3. Find products that sold fewer than 10 units — these need attention.
4. Create a summary that a sales manager could use for weekly reviews.

Submit your spreadsheet with formulas visible. In a text comment, explain any tricky formulas and which tool (formulas vs. pivot table) you chose and why."

This assesses:

• Multi-step problem-solving
• Formula correctness under realistic constraints
• Judgment about which tool to use (pivot tables, formulas, or both)
• Communication about technical choices
• Practical awareness of business needs (what would a manager actually use?)

Scoring: Use a rubric anchored to the role:

• Accuracy of calculations (40%)
• Clarity and usability of output (30%)
• Explanation of approach (20%)
• Formula efficiency (10%)

Tier three: Conversation to assess judgment

After the take-home, a 30-minute conversation. Bring the spreadsheet and ask:

1. "Walk me through your approach. What did you try first? What changed?" (Assesses problem-solving process, adaptability.)
2. "Why did you choose pivot tables instead of formulas [or vice versa]?" (Assesses judgment about tools, not just execution.)
3. "What would you do if the dataset grew to 100,000 rows?" (Assesses awareness of performance limits, scalability.)
4. "Have you used [feature X in their spreadsheet] before, or did you figure it out during the test?" (Assesses honest self-assessment, willingness to learn.)

The conversation reveals whether the spreadsheet is their work and their thinking, or whether they copied something and don't understand it. A candidate who built the solution confidently will explain it fluidly. A candidate who Googled a solution will stumble on edge cases or assumptions.

Why this beats automated scoring

Automated Excel tests claim to be objective. Usually, they're neither:

• They test memorization of shortcuts ("Press Alt+E to clear contents").
• They test trivia ("Which function counts text values?" COUNTIF, not COUNTA).
• They lack context (isolated formulas in artificial scenarios).

A realistic scenario + take-home + conversation actually measures:

• Competence: Can they solve real problems?
• Judgment: Do they choose appropriate tools?
• Communication: Can they explain their work?
• Honesty: Can they own what they built and what they don't know?

Those are the things that matter on the job.

Scope by role

For data analysts

Use tier 2 and 3. Include a scenario involving aggregation, filtering by multiple criteria, and creating a usable output. Add a question about handling missing or inconsistent data.

For finance / FP&A roles

Emphasize the conditional-math test heavily. Include scenarios with percentages, weighted averages, and sensitivity analysis. Ask about version control and documentation of assumptions.

For operations / coordinator roles

Tier 1 screening is usually sufficient. Focus on data-joining and conditional-logic tests. If they pass, brief tier 2 is optional—a 20-minute problem that matches their job (e.g., scheduling, inventory tracking).

For executive roles

Assess the output, not the mechanics. Can they consume a dashboard and make decisions? Do they understand the difference between a pivot table and a formula-based summary? This is less about Excel skill and more about data literacy.

Common mistakes to avoid

Mixing tools in a single test. "Use Excel AND Power BI AND SQL." Candidates won't know which is being tested. Single tools per assessment.

Asking about obscure features. "Which is faster, SUMIFS or array formulas?" Only matters if the role involves optimization. Skip it.

Grading based on formatting. If the output is useful and correct, formatting doesn't matter. Save time.

No partial credit for approach. If a formula is syntactically wrong but the logic is sound, score it 60–70%. You want to see their thinking.

Integration with your assessment process

Place Excel assessment early — second stage if they make the phone screen. It's a 20-minute filter that saves everyone time if they don't have the skill. Candidates who pass move to:

• Take-home work-sample test (if the role requires it)
• Technical interview with a senior person
• Behavioral round to assess past experience with data work

This sequence separates candidates by actual ability and saves weeks of interviewing bad fits.

Good Excel assessment is boring and feels obvious. That's the point. It filters effectively because it's not clever — it's just realistic.