Anaplan Estate - turning model exports into practical improvements

Question

Greetings, fellow Anaplanners.

I’ve always wanted something that could take thousands of line items and help me work out where to start improving a model. So I built it.

It’s called Anaplan Estate. Give it your Line Items and Actions exports, and it brings optimisation findings into an action plan, with evidence and suggested next steps.

Things like:

* Long IF formulas that could become mapping tables.
* Calculations repeated across models.
* Potentially redundant modules worth reviewing.
* Stale imports and actions sitting outside a process.

You can also explore what depends on a line item before changing it, following the chain beyond its immediate readers to help scope your testing.

It’s free, open source and MIT licensed. You can use it online or run it locally if your exports need to stay on your machine.

There’s a fictional example to explore without uploading anything:

👉 Try Anaplan Estate

👉 GitHub — source code and local setup

I’ll be actively maintaining it, and I’d love feedback from people doing this work. What would make this useful in your next model review?

klameer · Answer

Hi Kartik,

Short version: performance and correctness come first, maintainability never makes the plan, and dependency risk doesn't move an action up or down. It changes what you check before you act.

Every test that finds something suggests an action. I then ask three questions about each one, in this order.

* Can I trust the finding? Some tests have to guess, for example working out what a module does from its name. Those go to the bottom, however big they look.
* Is it one thing or a whole area? "Fix this line item" ranks above "review this module", because someone can actually go and do it.
* How much is at stake? I look at the share of the model's calculation time the line item takes and how many cells it holds. A few tests skip this step because they matter at any size. A total that might be wrong goes on the plan even if it's tiny.

Naming, long formulas and hard-coded numbers never compete for the five places. They're listed under the plan, but they can't push a performance or correctness fix out.

Dependency risk is shown rather than ranked. Every action links to the change impact view, so you see every line item that reads it before you touch anything. And if a check is needed first (is anyone still using this module?), it goes into the plan ahead of the change.

Kartik · Answer

This is interesting, especially the dependency-chain view. For larger models, I think understanding the potential impact of a change can be just as important as identifying the optimization itself.

How are you thinking about prioritizing findings in the action plan? For example, would performance impact or dependency risk take priority over things like formula complexity or maintainability?

klameer · Answer

Hi @GingerAnderson, thanks for your response. I will also add this to the repo and the site.