Data analyst applications: build one small project that answers a real question.
Connect Excel, SQL and Power BI to a reproducible decision, rather than a list of tool names.
Choose a question before choosing a dashboard.
An employer needs an analyst to make information usable, not simply to produce charts. Read whether the role emphasizes operations reporting, product analysis, finance or another function. Then choose a small project with an explicit question. For a fictional store dataset, ask which products generated the most recorded returns during a stated period. Define the period, unit and exclusions before calculating anything. Use public data with a suitable license or clearly synthetic records, never a former employer’s confidential export.
Make each tool earn its place.
A spreadsheet can show initial checks and a simple reconciliation. SQL can demonstrate how you join records, group results and handle duplicates. Power BI can communicate the result and document the model. You do not need all three tools in every sample. Microsoft’s Power BI certification description covers preparing and modeling data, visualization and analysis, and managing and securing the environment. That is a useful map of one product role, not proof that every analyst vacancy requires that certification.
Keep a short audit trail alongside the result.
For the fictional returns example, record the source of the data, date range, meaning of a return and treatment of missing product identifiers. Show a row-count check before and after a join. Explain whether you count return events or units, because those can produce different rankings. If an output looks surprising, trace one record rather than hiding the discrepancy. A reviewer should be able to follow the reasoning without opening ten undocumented notebooks or guessing which file is the final version.
Make the portfolio claim narrower than the evidence.
You could describe a synthetic project as “Built a returns dashboard from invented sales records, documented duplicate handling and reconciled totals against the input file.” Do not describe it as a commercial result or invent a percentage revenue gain. If you have genuine workplace evidence you may discuss, distinguish your contribution from team work. Explain one limitation, such as missing reasons for returns, and the next data you would request before making a business recommendation.
Check pay and location against the actual vacancy.
There is no single analyst salary covering the USA, UK and Canada. Compare current postings with the same city, seniority, contract type, currency and pay period; do not substitute search-volume figures for salary evidence. Ask whether a listed amount is base pay or a wider package. Remote roles still have hiring-location constraints. Use the related listings and official portals below, then tailor the project explanation to the central task in the employer’s advertisement.
Your application preparation sheet
Use this as a working note. Fill in your own facts; the examples in this guide are fictional.
- Target role and central business question:
- Dataset origin, license or synthetic-data label:
- Excel / SQL / BI tasks I actually performed:
- Validation check and result I can reproduce:
- One limitation and next question:
- City, level, currency and pay basis when comparing salaries:
Source and further reading
Matching live opportunities
Up to five matching, source-checked listings from our permitted feeds. Official careers searches below are separate resources, not confirmed vacancies.
Checking matching listings…
Official careers and application searches
These are official portals or employer careers pages. Their inclusion does not imply current vacancies, sponsorship or endorsement.
Original AI-generated editorial image; people and scenarios are illustrative. How we prepare our guides · Report an error
Practical editorial guidance, not an employment, legal, tax or financial guarantee. For advice specific to your circumstances, consult an appropriate professional.