Engagement
Historical cohort reconstruction
For messy archives: joining older exports so last year’s joiners can be compared with this year’s without pretending the files were always tidy.
Who it is for
Organizations that changed systems mid-year or kept paper-adjacent spreadsheets alongside later exports.
What you leave with
A reconstructed identifier map, a gap log, and cohort tables that refuse to hide missing months.
Scope
Up to two source eras and one overlapping key (email, membership number, or similar).
Included
- Identifier matching notes
- Months marked unusable rather than interpolated
- A reconstructed table for the periods that survive
Outside this engagement
- Inventing activity for months with no source
- Forensic recovery of deleted databases
Who does the work
Bundlelogicbase staff experienced with imperfect operational files.
How the work proceeds
- Inventory of files and eras
- Matching rules written down before any join
- Tables with a visible gap log
Duration
Four to eight weeks depending on how many eras and how inconsistent the keys are.
Where the work happens
Written matching rules plus tables.
What to prepare
Every export you still have, including ones you think are obsolete.
Constraints
If no overlapping key exists, we stop at a gap log rather than a false join.
How the fee is set
Quote after the file inventory; this work is rarely a fixed card price.
Next step
List the file names and date ranges you still hold. Contact details · Rates