Model Template Structure

This article explains how to create and use model templates in Model Reef so you can reuse proven structures across multiple clients, entities or projects.

You will learn:

  • What a template model is.

  • Which parts of a model should be templated.

  • How to turn a working model into a template.

  • How to keep templates clean and up to date.

Templates save time and keep your modelling standards consistent across your workspace.


1. What is a model template

A model template is simply a model whose structure you intend to reuse repeatedly.

A good template usually contains:

  • A well designed branch tree for the use case (for example multi entity group, SaaS business, retail network, project finance).

  • A consistent category and subcategory scheme.

  • A set of variables and drivers that define standard logic and behaviour.

  • Prebuilt dashboards and reports for common outputs.

  • Clear naming conventions, notes and tags.

When you duplicate the template, you get a fully structured model that you can then populate with data for a particular business or scenario.


2. Deciding what to put in a template

Include in your template:

  • Structural elements that are the same every time, such as branch layout and categories.

  • Generic variables and drivers that represent common behaviours (for example subscription revenue patterns, standard cost structures).

  • Standard dashboards (for example P&L overview, Cash Waterfall, KPI pack).

  • Notes that explain how the template is intended to be used.

Avoid hard coding client specific details into the template, such as:

  • Client names in branch or variable names.

  • One off adjustments or local quirks.

  • Historical data imports for a specific entity.

Think of the template as the skeleton and logic, not the data.


3. Creating a template from a live model

A common workflow is:

1

Build the model

Build a model for a real use case.

2

Refine structure

Refine its structure, naming and dashboards until you are happy with it.

3

Copy the model

Make a copy of the model.

4

Clean client data

Strip out client specific data and rename branches and variables to generic labels if necessary.

5

Save as template

Save this cleaned version as your template.

From then on, you duplicate the template instead of starting from scratch.


4. Using templates for new models

When you create a new model from a template, you typically:

1

Duplicate the template

Duplicate the template model.

2

Rename model & branches

Rename the model and branches to match the new client, entity or project.

3

Connect data sources

Connect data sources (Xero, QuickBooks, PDF imports, ticker fundamentals) to populate variables.

4

Adjust drivers and timing

Adjust drivers and timing where the new case differs from the template assumptions.

5

Review dashboards

Review dashboards and reports, tweaking only where needed.

This approach keeps logic consistent while allowing for case by case tailoring.


5. Maintaining templates over time

Templates should evolve as your best practice evolves. To keep them healthy:

  • Periodically review templates to incorporate structural improvements from recent live models.

  • Update naming conventions, categories and dashboards to reflect current standards.

  • Remove any one off hacks that have crept in.

  • Version your templates if necessary, for example SaaS Template v2.

When you significantly change a template, document what has changed so teams know which version to use.


6. Template governance

If you have a larger team:

  • Designate an owner for each core template type (for example SaaS, Retail, Project Finance).

  • Encourage modellers to propose improvements to templates rather than diverging in their own models.

  • Use templates as part of onboarding for new team members, so they learn the standard structure.

This keeps your workspace coherent and reduces the risk of inconsistent modelling approaches.


Last updated