The short answer
Use the export in your own combined analytics, keep the original XLSX private and save its account, selected dates, observation time and metric scope. Inspect the actual sheets and headers before transforming anything. An analytics workbook, an account archive and a public-profile recap contain different evidence; none should silently replace another.
What you’ll take away
- Preserve the untouched source separately from your working copy.
- A spreadsheet row may be a day, a post, a ranking or an aggregate.
- Save useful data while available instead of relying on one universal retention number.
Choose the export for the question
LinkedIn documents an XLSX export in combined post analytics. Open the owner account’s analytics, choose the relevant range and metric, then use the available export control. This is different from requesting a general account archive through privacy settings. The post records needed to locate old content are not automatically a history of private analytics.
Before exporting, state the task: preserving a period report, reviewing specific posts or building a same-age baseline. One exported aggregate may answer the first task but not the third. This guide did not download or test a private account workbook; the checklist below tells you what to verify in your actual file.
Keep an untouched source and a small context note
Save the received file privately, without replacing its original values. Use a separate working copy for filtering, calculated columns and formatting. Record a descriptive local filename and export time. Avoid putting a public profile name or sensitive audience details into a shared filename if they are unnecessary.
The following source note is an original template. Fill it from the actual interface and file, not from a guessed schema. If a field is unavailable, write that explicitly. A reproducible note explains how the result was obtained without requiring you to publish a private workbook.
| Source-note field | What to record |
|---|---|
| Origin | Owner account or Page, native view, authorized access |
| Time | Chosen dates, export time and visible time-zone context |
| Grain | Daily event counts, lifetime post totals, ranking or aggregate |
| Definitions | Exact metric labels and any stated exclusions |
| Coverage | Missing fields, unavailable content types, limits or partial periods |
| Working change | Filters, excluded rows, formulas and saved derivative filename |
Inspect grain before joining or adding rows
Open the actual workbook and inspect sheet names, headers, units and date cells. Do not assume every export has the same tabs or a complete row for every historical post. A daily portfolio total is not a post-level observation. A top-post ranking is a selection, not necessarily the full eligible history.
A constructed example illustrates the problem: the daily series contains 500 September impressions, while a selected post row has 800 lifetime impressions when exported later. Subtracting one from the other does not estimate missing posts. Check row meaning first, then use the appropriate view. Retain blank and unavailable values; filling them with zero changes the evidence.
Handle conflicting retention guidance honestly
The official help checked on October 31, 2026 is not internally consistent for video retention. The general metric-definition page lists 365 days, the post-navigation page 180, and the older dedicated video page up to 400. Do not present any one of those as a universal guarantee for every current interface and metric.
Preserve the information you actually need while it is available, note the source and date, and check the current account view or support guidance if an older value matters. A stored export is evidence of what that export contained, not proof that the same field will remain available indefinitely. Do not buy a tool solely on a promise to recover a metric that its actual source never retained.
Record transformations so a later reader can check them
In the working copy, identify the exact rows used, the rule for excluding them and the formula applied. Keep duplicates, absent values and invalid rows visible until resolved. If a comparison requires 24-hour observations but your file only has current lifetime totals, the file cannot supply those earlier measurements by division or interpolation.
Share the minimal result or approved excerpt instead of the whole private workbook. Strip unnecessary identities and hidden working material from the derivative, and verify the saved output. Deleting a local working copy does not revoke a file already sent to someone else. Finish with source note, unmodified original, transformation record and a bounded conclusion.
Sources & product references
Product details reflect the publication date. External features and interfaces may change.
