Bulk Importing Haulage Jobs from Excel: A Clean-Data Checklist

Bulk job import can turn a customer spreadsheet into planned work in minutes. It can also create 200 duplicated or incomplete jobs faster than a planner can detect them. The control is a clean staging process with preview, validation and a safe way to reject errors.

An import creates records; it does not prove that a route, load, vehicle or driver assignment is lawful or feasible.

Keep the original file unchanged

Save the customer’s original workbook with the received time and source. Create a working copy for cleaning. This preserves evidence of what was supplied and prevents arguments about whether the office changed a reference or address.

Use one responsible importer and a second-person review for large or high-risk uploads. Restrict access because customer files may contain personal contact details and commercially sensitive rates.

Map columns explicitly

Create a data dictionary for required and optional fields:

• unique customer/job reference;
• billing customer;
• collection and delivery locations;
• ordered stop sequence;
• planned date and time with timezone;
• commodity, quantity, weight and units;
• vehicle or trailer requirements;
• driver instructions and site restrictions;
• purchase order and proof-of-delivery needs;
• charge or rate where the import supports it.

Do not infer a postcode, time or unit silently. Reject or quarantine ambiguous rows.

Normalise before upload

Trim spaces, standardise dates, separate address fields, validate postcodes and use consistent units. Remove formula errors and hidden rows. Convert display-formatted dates to real values and check Excel has not changed long references into scientific notation.

Use stable customer and location identifiers where possible. “Tesco RDC” may refer to several sites; a saved location should be matched by a controlled code, not the first similar name.

Detect duplicates using more than one field

Check exact customer reference first, then compare date, collection, delivery and load. A revised file may contain the same work with one corrected column.

Choose an explicit rule: reject duplicates, replace only in a controlled update process, or import as a new revision after human approval. A blind upsert can overwrite a live job already assigned to a driver.

Validate operational feasibility

Before dispatch, review time windows, stop sequence, estimated duration, weight, vehicle type, site limits and required equipment. Compare assignments with driver, vehicle and trailer availability and recorded expiry warnings.

Imported work should enter an unassigned or review state unless every assignment control is intentionally satisfied. A spreadsheet cell naming a driver is not proof that the driver is licensed, available or within hours.

Preview, import and reconcile

The importer should show accepted rows, rejected rows and reasons before creating anything. Save the preview and approve the count.

After import, reconcile:

1. source row count;
2. accepted and rejected count;
3. created job count;
4. unique references;
5. total stops;
6. date range;
7. sample of first, middle and last job;
8. any assignment or price fields.

Never correct rejected rows only in the live system while leaving the reusable source broken. Update the controlled working file and preserve the error report.

Use a rollback-safe pilot

Test a small file in the trial or approved environment. Confirm cancellation or deletion rules before importing a full day. In production, use a batch identifier so created jobs can be found and reviewed.

HauliK Pro supports bulk CSV/XLSX job import. It does not turn unverified customer data into safe route decisions. Use the dispatch workflow to review and assign imported jobs.

Frequently asked questions

Should the customer’s spreadsheet become the template?

Only after controlled mapping. Customer layouts change; maintain your own import specification and transformation notes.

Can assignments be imported?

Technically some systems allow it, but every driver and asset still needs a live availability and compliance check before dispatch.

What should happen to invalid rows?

Reject them with a clear reason and no partial hidden record. Correct and re-import through the controlled process.

Is CSV safer than Excel?

It is simpler, but can still lose leading zeros, encode characters badly or use ambiguous dates. Validate the actual exported file.