Revit Template Setup: What a Structured Template Contains
Raul Alvarez · Founder, Datum AEC · July 4, 2026
Open a Revit model that has been through a real deadline and the recurring errors usually trace back to the file the project started from. Not the software version. The template. It is not the whole story: a foundation still has to be maintained through every decision that follows, and a strong start can be eroded mid-project. But it is where the story starts, and a project that starts on the defaults is compensating from day one.
Revit template setup is the set of structural decisions a project inherits before the first wall is drawn. Made deliberately, those decisions give the model its best chance of absorbing deadline pressure without breaking. Left at the defaults, they are why the same errors return project after project, regardless of how good the team is.
The problem with evaluating a template, whether you are building one or buying one, is that "structured" gets claimed by every option on the market. The word has stopped carrying information. This guide makes it concrete: the five configuration layers a structured template actually contains, the class of error each one prevents, and a test you can run on any template in about 15 minutes, including the one your firm uses now.
A Template Is a Set of Decisions, Not a File
The RVT/RTE file is a container. What you are actually evaluating is the configuration inside it: decisions about how views behave, how the model is organised, how elements are named, and what data travels with the geometry.
That reframe matters because the default template is not neutral. It is a complete set of decisions too; they were just made by the software's factory settings instead of by your practice. Start a project on it and the project inherits every one of those choices. The cost doesn't appear in week one. It appears at the first deadline, when the drawing set has to be consistent and the model has to answer questions it was never configured to answer.
The difference between a structured template and an improvised one is not effort. It is timing. Structure is a set of decisions made once, before the first project. Improvisation makes the same decisions during the project, one error at a time, at the most expensive possible moment.
The Five Layers of Revit Template Setup
Every structured template, whatever its origin, addresses the same five layers. Together, they are what "structured" means.
1. Naming Conventions
Views, sheets, families, types, and parameters named to one consistent system. This is the layer with the longest industry lineage: naming standards for design documents have existed since the AIA published its first CAD Layer Guidelines in 1990, and the US National CAD Standard absorbed and extended them. The transition from CAD to BIM didn't retire them. Naming is not personal preference; it is structured communication with every person who will ever touch the file.
The error class this prevents: ambiguity. The wrong view placed on a sheet. A consultant misreading a deliverable. A new team member with no way to infer where anything belongs. The National CAD Standard states the payoff plainly: when information appears in the same place in every drawing set, discrepancies fall, and with them errors and change orders.
2. Graphic Controls
View templates are the visible part of a larger system. Graphic control in Revit is a hierarchy, and a structured template configures every level of it:
- Object styles: the model-wide baseline. Line weights, line patterns, materials, and fill patterns, set per category.
- View templates: visibility settings and graphic overrides, configured per view purpose (floor plans, elevations, sections, details) and assigned so views of the same type read the same way.
- Phase filters: visibility and overrides controlling how existing, demolished, and new elements display.
- Filters: rule-based visibility and overrides for the conditions categories alone can't isolate.
- Annotation styles: how text reads, dimensions terminate, and tags draw: the type settings (fonts, arrowheads, leader behaviour, line weights) that keep annotation consistent across the set.
- Model objects: material assignments that enrich graphic controls.
The error class this prevents: inconsistent drawing output. Every level of this hierarchy is a place where graphics can drift, and an override at one level can quietly contradict another. When the levels are configured to work together, the printed set matches itself. When they are left at the defaults, sections read differently from plans, and the review cycle fills up with graphic corrections instead of design decisions.
3. Browser Organisation
The browser is the model's table of contents. What matters is not one prescribed hierarchy; it is that the organisation is decided before the project and applied consistently. Discipline, family, then type is one workable system. In the Datum AEC template, views are organised by the US National CAD Standard discipline designators and numbering, driven through Revit view types. The right system can vary with the practice and the scenario. Having one cannot: it is what lets the hierarchy scale as the project grows instead of collapsing into a single scrolling list.
The error class this prevents: navigation cost and duplication. When views take minutes to find, teams create duplicates; duplicated working views end up on sheets; and the model becomes readable through habit rather than through structure.
4. Model Organisation
Beneath the geometry, a model runs on systems: the phase system, the workset system, the revision system, project and shared parameters, and area and volume computation. A structured template configures each of them at the template level, because most resist clean correction once the model is populated. Phase settings misconfigured in week one produce demolition that displays as new work and existing walls that appear in new-construction schedules; worksets and revisions bolted on mid-project inherit whatever state the model is already in. This is the layer where "fix it later" genuinely fails.
Parameters deserve particular attention: a structured parameter set, ready for project documentation and ready to extend. Parameters are what turn geometry into information: they are the fields your schedules, tags, and title blocks read from.
The error class this prevents: schedules that need manual overrides and tags that display nothing. When the parameters exist before drawing starts, the schedule pulls from the model because the model has something worth pulling. When they are added mid-project, every schedule becomes a reconciliation exercise.
5. Content Management
A template must carry the basic content required to function. The datum and documentation content: levels, grids, elevations, sections, callouts, and view title styles. The annotation content, the actual types behind the styles governed in layer 2: line styles, text styles, dimension styles, tags, symbols, detail items, and title blocks. And the model content: default materials and working wall, door, window, floor, ceiling, and roof types, configured as real assemblies rather than the generic placeholders Revit ships with.
What it must not carry is your whole library. Typology-specific families load independently, as the project requires them. What would be the benefit of preloading a bed family into a workplace project? A structured template ships lean: the content every project needs, and nothing that only some projects need.
The error class this prevents: both extremes. Placeholder content survives into construction documents; a "Generic, 200mm" wall carries no reliable data, so it schedules wrong, details wrong, and eventually gets duplicated into near-identical types with conflicting structures. Preloaded everything fails the other way: every project starts heavy, slow, and full of content that doesn't belong to it.
Why Structure Works: The Model as Managed Information
Five layers, one underlying move: each treats the Revit model as managed information rather than accumulated geometry.
This is also how the international standard frames it. ISO 19650, the standard for information management in construction, defines every model file, schedule, and sheet as an information container: a named, persistent set of information that can be retrieved. Naming conventions, browser organisation, and shared parameters are simply what keep those containers retrievable. The standard is equally clear about the alternative: it notes that considerable resources are spent correcting unstructured information and solving problems of coordination and reuse.
Two more points from the standard counter the two most common objections.
First, structure does not mean more detail. The concept many teams know as LOD appears in ISO 19650 as the level of information need (LOIN), and its stated purpose includes preventing the delivery of too much information: the framework defines the minimum required for each purpose, and anything beyond that minimum, the standard says, is waste. A structured template is not a heavier template. It is a template where the data that exists is the data that gets used.
Second, this is not enterprise machinery. The ISO 19650 standard states that its principles should be applied proportionately to the scale and complexity of the project, and that this matters particularly where small and medium-sized enterprises are involved. A delivery team, in the standard's own terms, can be one person. The five layers above are proportionate application in practice: the discipline of a large-firm BIM standard, at the scale of a starting file.
How to Evaluate Any Template in 15 Minutes
You can test all five layers on any template: one you built, one you inherited, or one you are considering buying. Start a new project from it and check:
- Names: Read ten view and family names. Could a new hire infer the system from them?
- Graphics: Open a plan and a section. Is a view template assigned to each, and do object styles carry deliberate line weights, or are the graphics ad hoc?
- Browser: Does the project browser follow a deliberate organisation system, or is it the default flat list?
- Model systems: Place a door and open the door schedule. Does data populate, or do the fields sit empty? Then check the phase filters and worksets: configured, or untouched defaults?
- Content: Open the wall types. Real assemblies with defined structures, or "Generic"? And is the loaded content the working basics, or half a library?
A template that passes all five is structured, whoever built it. A template that fails most of them is not a bad file; it is a set of decisions still waiting to be made. The only question is when they get made: before the next project, or during it.
If you are weighing build versus buy, this checklist is also the honest cost comparison. Building means making every decision behind all five layers yourself, then testing those decisions across live projects until the failures surface. Buying means evaluating someone else's decisions, with the checklist above, in an afternoon.
Start Structured, or Inherit the Defaults
Recurring Revit errors are not random, and they are not a team problem. Most trace back to a configuration layer left at its default. A template can't make the decisions that come after it; what it does is set the foundation those decisions stand on. Structure is a decision made before drawing starts; every project either inherits a set of deliberate decisions, or a set of factory settings.
The Revit Project Template applies everything in this guide: all five layers, configured decision by deliberate decision, aligned to ISO 19650, and applied across 400+ live projects. The launch list is open at datumaec.com.