Importing rental records can multiply a small naming inconsistency across many transactions. A property marketed under one name, a street address written another way and a unit identifier with a meaningful leading zero may be treated as different records. Build a verified mapping before importing so that correct totals do not conceal records assigned to the wrong home.
This Dallas edition is an original planning worksheet for the specific decision in the title. It does not report a local survey, a property inspection or a completed software test. Use your own confirmed documents and observations for the inputs. Every amount in the worked example is hypothetical and is included to make the method reviewable, not to suggest a local price or a promised result.
Create a canonical property and unit register
List each actual property with a stable internal identifier and its verified address. Add units as separate records linked to the correct property. Preserve the spelling or formatting used in supporting documents where relevant, while choosing one consistent operational label for the destination system.
Keep marketing names separate from property identity. Two buildings may use similar descriptive names, and a property can be renamed without becoming a different asset. Do not rely on a nickname as the only key for importing financial or maintenance records. Use a documented cross reference that an authorized reviewer can understand.
Check how the destination system expects property and unit fields to be structured. A combined address field in the source may need to be separated, but do not guess the split when the original data is ambiguous. Resolve the identity from appropriate existing documents before moving the affected rows.
Preserve meaningful unit characters
Treat unit identifiers as text when that is necessary to preserve their actual form. A leading zero, letter or punctuation mark can distinguish the intended label. Test what the import process does with a fictional sample rather than assuming a spreadsheet’s displayed value will survive unchanged.
Do not remove characters merely to make the data look tidy. First establish whether two forms refer to the same unit and record the mapping. Unit 01 and Unit 1 might be aliases in one property and require clarification in another. This worksheet does not assume a universal rule for every building’s numbering system.
Separate missing values from shared areas. A blank unit field may indicate an incomplete record, while a property level maintenance task may legitimately have no unit assignment. Use a clear supported convention for the latter instead of filling every blank with the first unit in the list.
Stage the import and inspect relationships
Work from a preserved copy of the source and create a staging file containing the proposed destination identifiers. Keep a separate exception list for rows whose identity is unresolved. Do not silently discard those rows merely to achieve an import success message.
Test a small fictional or appropriately protected sample through the normal authorized process. Check the resulting property and unit links from within the destination system and, where available, from an export. A count of accepted rows does not establish that the rows went to the right records.
Before a live import, understand the provider’s supported correction or rollback process and preserve the required records. Avoid making an irreversible bulk change just to discover how naming behaves. The practical aim is to catch identity errors while the affected sample is still small enough to inspect carefully.
Use this practical worksheet
| Record | Evidence to retain | Completion question |
|---|---|---|
| Canonical register | Verified property and unit identifiers | What is the authoritative destination for each row? |
| Alias mapping | Source labels connected to canonical records | Do similar names actually identify the same place? |
| Unit format | Text values and sample import output | Were significant characters preserved? |
| Unresolved rows | Exception list with owner and next action | Which records cannot yet be assigned confidently? |
| Reconciliation output | Counts and selected amounts by property and unit | Does the import preserve both totals and assignments? |
Review canonical register
Give every canonical entry a stable reference and keep aliases in separate fields. The reference should survive a marketing name change. Limit editing of the register to an appropriate process so that two reviewers do not create competing versions. Record the evidence used to resolve an ambiguous label and keep the resolution accessible to the people preparing the import.
Review alias mapping
Search the source for repeated variations, but review proposed matches rather than merging them automatically. An abbreviation can be an alias, while a similar address can be a different property. Keep one row per source variation in the mapping file. That makes the transformation explicit and allows an exception to be corrected without rewriting the entire source dataset.
Review unit format
Include a fictional unit with a leading zero and another with a letter in the test where those forms are relevant to your records. Compare the original value with the imported result. If the system changes the display, determine whether the underlying identity remains correct. Record the observed behavior rather than assuming a visual difference is harmless or automatically wrong.
Review unresolved rows
Keep unresolved rows outside the approved mapping until the identity is supported. Record the source reference, the ambiguity and who will investigate it. A row should not be assigned to a convenient property just to complete the file. Include the unresolved count in the reconciliation so that excluded records remain visible and do not vanish from the project.
Review reconciliation output
Compare totals at several levels, starting with the whole sample and then each property and relevant unit. A correct grand total can coexist with two records swapped between units. Inspect individual references as well as sums. Keep the review record with the import batch so that a later question can be traced to the exact source and mapping used.
Work through a hypothetical example
Use a fictional sample of eight maintenance records across two properties. Property A has five records totaling $700, and Property B has three totaling $450. The expected sample contains eight records and $1,150 in selected amounts.
Suppose the imported sample still has eight records and totals $1,150, but a $100 record from Property A has been assigned to Property B. The destination totals become $600 for Property A and $550 for Property B. The grand total passes while the property comparison fails. This is why a successful import message and a balanced overall sum are insufficient checks.
After correcting the mapping through the supported process, the expected totals return to $700 and $450. Review the affected unit as well, because a property level correction does not prove that the unit is right. The figures are invented for a data exercise and do not describe Dallas maintenance prices or the results of a product test.
Test a naming change without creating another property
Add a fictional marketing rename to the staging exercise. Keep the canonical property identifier unchanged while updating the descriptive label through the supported process. Then inspect an older work record and a newly imported one. Both should still point to the intended property in your test requirements. If they do not, record the observed limitation before using the process on live records.
Next, test a source row containing only a partial address and a unit number that exists at more than one property. That row should remain unresolved until supporting evidence identifies the destination. Do not let the first matching unit in a spreadsheet become the assignment by default. The exception is valuable because it shows where a seemingly simple match would create a false record relationship.
Compare the staging file with the preserved source using stable row references. Every approved row should have a documented destination, and every excluded row should appear in the exception list. If the import process creates its own new identifiers, preserve a mapping to those results where available. That mapping supports later corrections without requiring someone to guess which imported entry came from which source row.
For software evaluation, ask the provider to show the supported handling of duplicate property labels, text unit identifiers and an unmatched source row. Record the actual behavior and any required manual step. A provider may offer several import methods with different limits, so identify the method demonstrated. Do not generalize from one screen to every bulk import option.
Review property and unit assignments with someone who knows the portfolio’s real structure before proceeding with live data. A technically valid file can still describe the wrong relationships. That reviewer should focus on identity, while the person preparing the import checks format and completeness. In a small operation one person may perform both reviews, but the two questions should remain distinct.
Keep the approved mapping with the import batch and note the correction process actually used for any test error. If a later source file introduces new aliases, review them before reusing the mapping automatically. A successful earlier import establishes a tested method for that sample, not permanent assurance that every future address variation belongs to an existing record.
Optional software evaluation resources
Affiliate disclosure: Homzora may earn a commission through the following links. Referral relationships do not establish that a product supports the workflow in this article. Ask the provider to demonstrate current capabilities, charges and export options using your own fictional exercise.
Explore Buildium and explore Rentec Direct as possible evaluation candidates. Neither product is ranked or reported as having passed a test here. Your existing process may already satisfy the requirement.
Record observed results in the Homzora software scorecard and return to the Dallas edition for related resources. Keep each open exception assigned to a person and a next action. Closing an exercise means that the evidence supports the conclusion, not merely that every field contains some text.