How to Understand an Inherited Excel Model
Someone leaves, and their model lands on your desk with no documentation and no handover. You now own numbers you did not build and cannot yet explain. The way out is not to read every cell but to reverse-engineer the model methodically, from structure down to the assumptions that drive it.
Map the Sheets Before Touching Formulas
Resist the urge to dive into cells. First, walk the tabs and build a mental map: which sheet holds inputs, which does the heavy calculation, which presents results.
Note the flow. Most models move from assumptions to a calculation engine to outputs, and recognizing that shape orients everything that follows.
- List every sheet and a one-line guess at its purpose
- Identify where the headline outputs are presented
- Find the sheet that holds the assumptions
- Note any sheets that look like scratch work or old versions
Trace the Headline Number Backwards
Pick the most important output, the one the model exists to produce, and walk back through its precedents. This is the fastest way to learn what actually drives the model.
Excel's Trace Precedents draws arrows, but on an unfamiliar dense sheet they overlap into noise. Stepping back one layer at a time, following each reference deliberately, builds real understanding.
- Select the model's key output cell
- Trace its direct precedents and note which sheets they live on
- Step into each precedent and trace its inputs
- Keep going until you reach the raw assumptions
Find the Assumptions, Including the Hidden Ones
Once you reach the inputs, separate the real assumptions from the buried ones. Some will sit in labeled cells. Others will be hardcoded inside formulas where the previous owner never pulled them out.
Cataloguing every hardcode is essential. Each one is an undocumented assumption you now own, and you cannot flex what you cannot find.
- Labeled input cells: the documented assumptions
- Constants typed inside formulas: the hidden ones
- External links: assumptions that live in another file
Find What Reaches Outside the Workbook
Inherited models often pull from other files: a master assumptions workbook, last quarter's actuals, a shared rates file. If those links break or go stale, your model quietly goes wrong.
Identify every cell that references another sheet or workbook so you know exactly what the model depends on beyond its own walls.
Document As You Learn
As you trace and discover, write it down. Even a short notes sheet listing the key drivers, the hidden hardcodes, and the external dependencies turns a black box into something you can defend.
The next person, possibly future you, will thank you. An inherited model becomes truly yours only once you can explain it without re-deriving it every time.
Formula Trace
Walk an inherited model's headline number back to its assumptions one layer at a time, with no documentation required.
Get ModelMint See how it worksFAQ
Where do I start with a model I did not build?
Map the sheets before touching formulas. Understand which sheet holds inputs, which calculates, and which presents results. That structural map orients everything else and stops you from getting lost in individual cells.
How do I learn what drives an unfamiliar model?
Trace the headline output backwards through its precedents, one layer at a time, until you reach the raw assumptions. Following the chain deliberately teaches you the logic far better than reading cells at random.
Why catalogue the hardcodes in an inherited model?
Because each hardcode is an undocumented assumption you now own. They hide inside formulas where no label points to them, and you cannot flex or defend an assumption you cannot find.