This guide is for business owners who receive export files from an ERP such as Accurate or Jurnal, or from a POS system, and want to read them correctly themselves or have their team do so.

01What question do you want to answer?

Start from the decision, not the file. Good example questions: which customers' revenue is down compared with last year, which products sell a lot but at thin margins, and which branch is falling behind. One analysis should answer one to three questions.

Try it yourself

Quick check: does the ERP export total match the official report?

Enter the two totals for the same period. Any difference, however small, needs an explanation before the analysis starts.

ERP export
Official report
Enter both numbers to see the difference.

The calculation runs in your browser. When WhatsApp opens, a summary of the result is added to the message.

Ask for help tracing the difference

02What data should you export from the ERP?

Export data at the transaction line level, not just summaries. The columns you usually need:

  • transaction date and invoice number
  • customer name and branch or salesperson
  • product code and name
  • quantity with its unit (pcs, kg, ton, carton)
  • price, discount and sales value
  • returns, if recorded separately
  • cost per product, if available and if you want to discuss margin

Take a period long enough to compare, for example the last twelve months plus the same months of the previous year.

03How do you clean a messy export file?

ERP export files almost always need tidying before analysis. Check for:

  1. Duplicate rows. An invoice that appears twice because the export was run twice.
  2. Different names for the same customer. "PT Maju Jaya" and "Maju Jaya, PT" are counted as two customers, so the revenue is split.
  3. Dates in mixed formats. Some are text, some are real dates, so the month order gets scrambled.
  4. Returns and cancelled invoices. Negative values need to be included correctly, not thrown away.
  5. Mixed units. Do not add pcs to cartons or kg to tons unless the company has an official conversion.
  6. Total rows in the middle of the data. Subtotal rows must be removed so they are not counted twice.

This step often feels tedious, but it is where most number errors begin. Save the cleaning steps, for example with Power Query in Excel. Next month, you only swap in the new file and press refresh.

04How do you make sure the numbers are right before you read them?

Match the total sales in your file with the official profit and loss report or sales report for the same period. If there is a difference, find the cause before going further: a mismatched period, tax included or excluded, or invoices not yet entered. Your analysis is easier to defend in a meeting once the numbers match the official report.

05What should you read from the sales data?

Five views of the data that most often help you make decisions:

  1. Revenue per month, compared with the same month last year, to see the trend and seasonal pattern.
  2. Revenue per customer. See how many top customers account for most of the revenue. The more concentrated it is, the greater the risk if one customer stops.
  3. Revenue per product, with quantity sold and its unit, to separate growth from price from growth from volume.
  4. Customers who stopped or slowed down, meaning those who used to order every month but have not ordered in the last three months.
  5. Discounts and returns per customer or salesperson, to see who receives the largest reductions.

If cost is available, add margin per product. A product with large revenue does not necessarily produce the largest profit in rupiah.

06How do you turn the analysis into decisions?

Write each finding in one sentence that holds the number, the comparison and the action. Example form: "Customer X's revenue fell from Rp 1.2 billion last semester to Rp 800 million this semester; call this week and ask why." A finding without an action stays a report. A dashboard becomes useful only after the team agrees on the process and the definition of each number.

FAQ

Do I need a dashboard to analyze sales data?

Not at the start. A table in Excel with clean data is enough to answer the first questions. A dashboard is worth building once the questions you repeat every month are clear.

How long does the first analysis usually take?

It depends on the condition of the file. Tidying the data usually takes the longest. Once the steps are saved, you can update the monthly report much faster.

Can AI help analyze ERP files?

Yes, especially for tidying, summarizing and answering simple questions. The results still need to be matched to the official report and checked by someone who understands the source data. Also consider data confidentiality before uploading a file to any service.

Why does the total in my file differ from the accounting report?

Common causes are a different period, tax treatment, returns, or invoices entered after the export was made. Trace the difference month by month to find the source.

DataERPSales