Save Space With Snapshots

DavidEdwards

When it comes to model size, few components add space quicker than versions. This is particularly true in use cases that rely heavily on scenario planning and what-if analysis. For example, a supply chain model may need to generate a monthly forecast with a best, worst, and average set of inputs. Especially when time ranges span multiple years, having such a high quantity of versions will often put a model up against its workspace limits, which proves costly to an Anaplan customer.

While versions are typically the simplest approach to dynamic scenario planning, another option is often overlooked as a viable alternative for storing large amounts of multi-scenario data (without drastically impacting the overall model size).

For companies with smaller workspace limits, as well as large multi-use case customers, removing versions altogether can be a viable option to workspace size constraints. In place of versions, a user-generated snapshot list can be used to store input data, which can then be referenced downstream in the calculation and output sections (DISCO) of a model.

See below for an illustrated overview of this technique:

Module Architecture

Input modules gather user inputs and feed them into the calculation module:

The user runs an action to create a new member in the Snapshots list, then runs a second action to copy data from the main calculation modules into identical modules dimensioned with the Snapshots list:

After snapshotting data into the appropriate list member, a selection is made to drive which snapshot is used in downstream calculations in the rest of the model:

While this example is fairly simple, the logic behind it can bring a tremendous amount of efficiency to any forecasting model and save precious space at the same time.

To add additional functionality, snapshots can also be combined with a model version list to allow for multiple scenario comparisons within a single version. Like the example above, this use case provides a snapshot selection line item for each model version. Users can theoretically have an unlimited number of snapshots, with each model version utilizing a lookup to a specific snapshot:

In summary, if a model is struggling with sizing constraints due to versioning, a snapshot list may be the solution. Provide users with the flexibility to create and compare many forecast scenarios and control which dataset flows through to the model’s final outputs. Give it a try and let us know what you think in the comments below!

---

Dave Edwards is a Certified Master Anaplanner and Solutions Architect with Columbus Consulting. During his career with the Anaplan platform, he has helped companies achieve a high level of Anaplan excellence and flexibility to support their unique business needs. He has over five years of architecture and implementation experience, with use cases spanning across Supply Chain, FP&A, Real Estate, Customer Service, and Sales Forecasting.

JaredDolich

@DavidEdwards As usual, another awesome suggestion and best practice. Thank you David.

I also like to provide the ability to add or remove snapshots, and by user. Easily handled in the new UX. Use the calculation module to drive the aggregate values, store only the values needed for decision making plus the input values, then have Anaplan highlight the best choice. It's an easy extension from there to use the optimizer if available.

david.savarin

The added layer for selecting a snapshot to be used as selectable for each version is a great swiss knife option around multi scenarios planning

DaanishSoomar

Great read @DavidEdwards!