How to Brief a Data, Reporting or Dashboard Project So the Right Analyst Finds It

Data projects are the most easily misunderstood, because the owner sees the problem clearly (“I don’t know how much I’m selling”) while the analyst doesn’t know what the data is, where it is, or what decision you want to make with it. In the Data & Analytics category on Farasa, a good analyst applies to the project that answers these questions from the start. This guide shows how to write such a post. For the general rules that apply to any software project, read our guide to posting software projects.

Don’t start with “I want a Power BI dashboard.” Start with what you want to know or decide: “I want to know which products sell most each month and where I lose money,” or “I want a weekly report that reaches my manager automatically.” The question decides the kind of work; the tool comes after. If you know which tool you prefer (Excel, Google Sheets, Power BI, Looker Studio), name it and say why; if you don’t, ask applicants for their recommendation.

This is the part that changes the price more than any other. Say where the data lives: Excel files, an accounting system, a database, a website, or paper that needs entering. Say roughly how big it is (hundreds of rows or millions), how many sources there are, and whether it is tidy or has errors, duplicates and dates in different formats. If you can, attach a small sample with any sensitive details removed; it saves the analyst days of guessing.

Data cleaning and standardising: unifying names and dates, removing duplicates, merging several files. Reports: a monthly or weekly report with specific figures, tables and charts. Dashboards: an interactive view that is refreshed and used continuously. Spreadsheet automation: replacing repeated manual work with an automatic process. And a one-off analysis: studying a specific question and delivering findings and recommendations. Say which you want, because each is priced differently.

Say who will read it: the owner, the sales team, an investor. A dashboard for a manager differs from a detailed report for an accountant. Say whether the data will be refreshed by hand or automatically, how often, whether the dashboard must work from a phone, and who may see what. And be honest if some people are not comfortable with complex tools; a simple result that gets used beats an elegant one that gets ignored.

Write how you’ll know the work is right: “the figures match the accountant’s report within a small margin,” or “what took two days is done in ten minutes.” It also matters to agree on the definition of the key figures before starting: what does “sales” mean? Before or after discount, and does it include returns? Poor definitions are a common cause of reports that look right and mislead.

Your data may include customer details or financial figures. Don’t send sensitive data before you’ve agreed terms with the analyst, and start with an anonymised sample where possible. Say if the data must stay in a particular place, and whether you want to grant limited access rather than full access. Ask how the analyst will protect the files and what they do with them when the project ends.

A defined project (cleaning one file, producing one report) suits a fixed price. Ongoing reports and gradual development of a dashboard suit hourly or a monthly arrangement. Split the work into phases: start with cleaned data and one correct report, then add the dashboard and automation. Leave time to review the numbers together, since that review is the most important step. Our guide to budget and timeline for software projects goes into more detail.

“Monthly sales dashboard for a chain of shops in Basra. Data: three Excel files (sales, stock, staff), about 40,000 rows a year, with product names written in different ways and inconsistent dates. I want the data cleaned, then a dashboard showing sales by branch, product and month, and the slowest-moving products. The manager will use it from a phone. I prefer Google Sheets with Looker Studio but I’m open to your suggestion. Definition of sales: after discount, before returns. Budget: fixed price for the first phase (cleaning plus one report), then we agree on the dashboard. Start your proposal by naming a similar dashboard you’ve built.”

Start with the question, not the tool; describe your data as it is; say what kind of work it is and who will use the result; agree the definition of the figures and the measure of success; protect the data’s privacy; and split the project into phases. That way you reach an analyst who understands your project. Post your data project on Farasa.