How to prepare a data analysis or dashboard project before hiring

1. Start with the questions

Write down three to five specific questions you want answered, and what decision each answer will support. For example: which marketing channels bring customers who buy again, or which products are losing money after returns. Clear questions keep the project focused and help the analyst choose the right methods. If you want an ongoing dashboard, list the decisions people will make with it and how often they will look at it.

2. List your data sources

Note where the data lives: spreadsheets, an online store, accounting software, a CRM, website analytics or a database. For each source, record roughly how much data there is, how far back it goes and how often it changes. Analysts quote very differently for one clean spreadsheet than for five systems that need to be combined.

3. Be honest about data quality

Most business data is messy: duplicate customers, inconsistent product names, missing values and different date formats. Tell the analyst about known problems and share a small sample. Cleaning can take as much time as the analysis itself, and a data cleaning specialist may be worth hiring first for large datasets.

4. Agree on definitions

Teams often disagree on what a number means. Is revenue before or after refunds? Is an active customer someone who bought in the last month or the last year? Write down how you define your key measures. Without this, two people can look at the same dashboard and draw different conclusions, and an analyst may build something technically correct that nobody trusts.

5. Plan access and privacy

Decide how the analyst will receive data. For one-off analysis, an export with personal details removed is often enough. For dashboards that update automatically, the analyst may need read-only access to systems. Create separate accounts with limited permissions rather than sharing your own login, and remove access when the project ends. If the data includes personal information, agree in writing on how it will be stored and deleted.

6. Describe the outputs

Say what you want to receive: a short report with charts and recommendations, a spreadsheet model, or an interactive dashboard. Mention who will read it and their comfort with data. Executives may want three headline numbers; an operations team may want detailed filters. Ask for a short explanation of methods and any limits in the data, so you know how far to rely on the results.

7. Think about after delivery

A dashboard needs care when data sources change or new questions appear. Ask for documentation of data connections and calculations, and agree on how future changes will be handled. Make sure files and dashboards live in accounts you own.

Start with a small first step

If the project is large or you are working with a new analyst, begin with a small, paid first step, such as cleaning one data source or answering a single question. It shows how the analyst communicates, how they handle messy data and whether their explanations make sense to your team. It also reveals problems in the data early, before you commit to a full dashboard. Once the first step goes well, the larger project can be planned with real knowledge of the data rather than guesses, which usually makes the final quote more accurate and the timeline more realistic.

Pre-hire checklist

• Three to five business questions and the decisions they support.
• Data sources, size and history listed.
• Known data quality problems and a sample.
• Definitions of key measures written down.
• Access method and privacy rules agreed.
• Output format and audience described.
• Ownership, documentation and future changes discussed.

Our data analytics guide explains how to compare analysts, and our data visualization guide covers charts and reports in more detail.