Prepare a property spreadsheet for a reliable import

A reliable property import starts with a spreadsheet in which every row represents one listing and every column has one clear meaning. The objective is not simply to make the upload succeed. It is to preserve enough context that an agent, a search filter and a property assistant interpret the resulting record in the same way. A technically valid file can still produce misleading records if rental periods, units or references are inconsistent.

LeadProps supports CSV and Excel imports as part of its property workflow. The checklist below explains how to prepare a source file, inspect a small test import and reconcile the final result. Keep an untouched original before making changes. Use the fields supported by the current import screen, and do not assume that a column will become searchable merely because it appears in the spreadsheet.

Decide what one row means

Use one row for one advertised listing. Do not put several apartments into the same row just because they share a building. Their prices, sizes, availability and photos may differ. Likewise, avoid splitting one listing across multiple rows for separate photos unless the importer explicitly requires that layout. Decide how to represent repeated units before exporting the file, because later duplicate checks depend on that decision.

Give each row a stable reference. A listing title can change, and two properties can have the same title. A reference such as an agency inventory code provides a safer way to compare the source with imported results. Preserve references as text if they contain leading zeros. Spreadsheet software may otherwise turn a code like 00184 into 184, creating a mismatch with other systems.

Keep source URLs in their own column. They help a reviewer investigate a question without searching through notes. Do not treat a URL alone as proof that the listing is current or authorised for reuse. The spreadsheet should contain only records the team is entitled to process and maintain.

Separate values that need separate interpretation

Store the amount, currency and payment period separately where the import mapping allows it. “145000 AED yearly” is understandable to a person but harder to validate consistently than three clearly named fields. Similarly, keep a property’s size and size unit distinct. A value of 950 means very different things in square feet and square metres.

Avoid decorative formatting as a substitute for data. A red cell may mean unavailable to the author, but that colour usually disappears when the file becomes CSV. Put the status in a real field. The same applies to comments, hidden columns, merged headings and formulas that depend on another workbook. Export a simple table with clear headers and the actual values needed for the import.

Check how blank values are represented. An unknown price should not quietly become zero. “Price on application” is a status, not a numeric price. An unknown bedroom count is different from a studio, and a missing bathroom value is different from no bathroom. Agree those distinctions before mapping columns so the imported records do not become misleading search results.

Normalise locations without deleting useful detail

Choose a consistent spelling for the area and building name. Differences such as an abbreviation, a trailing space or an alternate spelling can divide one location into several apparent locations. Keep the original wording in a source note if you need it for reconciliation, but use a controlled form for filtering and grouping.

Do not force an uncertain location into the nearest familiar area. If the source is vague, flag it for review. It is better to have an incomplete record than to recommend a property in the wrong place. Separate a broad community from a building or street where the source supports that distinction. A customer searching for a particular tower needs more precision than a community-level label.

Preserve factual descriptions, but remove copied interface text and repeated boilerplate that does not describe the property. A useful description explains the layout, relevant features and known conditions. It should not contain internal agent notes, private owner contacts or claims that no one has checked.

Prepare image links carefully

Use the image format expected by the import screen. Some systems accept a single URL, while others accept a defined list format or repeated image fields. Follow the actual mapping instructions instead of inventing a separator. A comma can be part of a URL, and an unescaped separator can break an otherwise valid row.

Test a few image links outside your signed-in browser. The importer may not have access to private storage, expired links or files that require cookies. Confirm that the link returns an image rather than a preview page. If images are missing, keep the listing out of customer-facing recommendations until the team decides whether the remaining information is sufficient.

Check ownership and relevance as well as accessibility. The correct image is one that belongs to the specific listing and is authorised for the intended use. A convenient image from another apartment is not a valid replacement. Label floor plans, renders and community photos distinctly in any source notes your workflow retains.

Clean the file before uploading

Remove empty rows above the header and summary totals below the table. Avoid merged cells in the data area. Give every mapped column a unique, descriptive header. If two columns are both named “Price,” it may be unclear whether one is monthly and the other annual. Rename them before import rather than depending on column position.

For CSV, use an encoding that preserves names and punctuation, typically UTF-8. Check the delimiter and quoting when a description contains commas or line breaks. Open the exported file in a plain text viewer or re-import a copy into your spreadsheet application to catch malformed rows. Do not edit the only original while testing an export.

Dates deserve particular attention. An entry such as 04/05/2026 can be read in two different ways. Prefer an unambiguous year-month-day representation when the importer supports it. Keep a date’s meaning in the header: date listed, date checked and date available are different facts. None should be substituted for another simply because it is the only date present.

Run a deliberately varied sample

Choose a small sample that includes ordinary rows and edge cases. Include a rental, a sale listing if relevant, a studio, a missing optional field, a long description and a row with several photos. Testing only the cleanest first row gives little confidence about the rest of the file.

After importing, compare each sample record with its source. Look at the displayed price, currency, rental period, bedrooms, area, location and images. Search for the record using the filters an agent would use. If the record is present but cannot be found by the expected field, investigate the mapping rather than assuming the import was successful in practical terms.

Write down the mapping that worked. For a recurring file, keep a short template showing the expected column names and examples. This prevents a colleague from unintentionally changing a field’s meaning in next month’s export. A stable template is more useful than repeated emergency cleanup after each upload.

Reconcile the full import

Compare expected rows with imported, failed and skipped rows. A successful upload message does not necessarily mean every source record became a usable listing. Inspect any warning or failure detail the system provides. Keep the original reference beside each correction so you can trace the issue back to the source.

Do not repeatedly upload the full file to fix a few rows without checking how duplicates are handled. Some workflows update existing records, while others create additional records. Verify the current behaviour with a small test or supported documentation before retrying. If duplicate handling is unclear, resolve that uncertainty before running another large import.

Finally, review the customer-facing collection. Remove obvious test records, resolve contradictory prices and check that inactive listings are not treated as available. Store the reviewed source and a short import summary in the team’s normal records. The goal is a repeatable process, not a one-time successful transfer.

Pre-import checklist

• One row represents one clearly identified listing.
• References are stable and leading zeros are preserved.
• Currency, price period and area units are explicit.
• Unknown values are not disguised as zero or a guessed default.
• Location names use consistent, reviewed spelling.
• Photo links work without a private browser session.
• Status is stored as data rather than cell colour.
• Headers are unique, and merged cells are removed.
• A varied sample has been checked in the destination workspace.
• Expected, imported, skipped and failed counts can be reconciled.

Frequently asked questions

Is CSV better than Excel for property imports?

Neither is automatically better. Use a format the current importer supports and that preserves your data accurately. CSV is simple and easy to inspect, but it does not preserve formatting or multiple sheets. Excel can retain more structure, yet merged cells and formulas can make an export harder to interpret.

Should missing prices be entered as zero?

No. Zero is a numeric value and can make an unknown-price property appear to be the cheapest option. Leave the value genuinely missing or use the supported representation for an unavailable price. Check how the destination displays that state before including the listing in recommendations.

Why did an image fail even though the link opens for me?

Your browser may have access that the importer does not. The URL could require authentication, expire after a short time or point to a preview page instead of the image itself. Test an authorised public link in a fresh browser context and use the importer’s supported image format.

Can I upload the corrected file again?

First check the importer’s current update and duplicate behaviour. Re-uploading may create another copy instead of updating the earlier record. Test with one controlled example and keep a record of references so you can reconcile the result safely.

What is the most important final check?

Compare a representative set of destination records with their source and confirm that the facts a customer sees are correct. Row counts matter, but they do not reveal a rental period mapped incorrectly or photos attached to the wrong unit.