CAD trials are 15-30 days, and most teams spend them watching demo videos. Then they buy, and month three reveals what the trial should have caught. Here's the evaluation protocol we recommend — built from the comparison data we maintain and the migration stories teams tell us.
Week 1: Your files, not their files
The single most important rule: the trial starts with your messiest production drawing, not the vendor's showcase model.
- Open your worst file: largest drawing, most complex blocks, most external references. Time the open. Orbit it. Save it. Reopen it.
- Round-trip test: save in the trial tool, open in your current tool, verify geometry, blocks, dimensions, and text survive
- Check the version mechanics: what DWG/DXF versions does it read and write? (A free 6-byte header checker tells you what your files actually are before you start: cadguide.tools — runs locally, nothing uploads.)
Week 1, continued: the automation audit
If your office has LISP routines, macros, or templates:
- Run your top-ten routines. Most port with minor edits on LISP-compatible platforms; some need line-by-line fixes; a few die. The ratio you measure now is the migration budget later.
- Load your template: title blocks, layers, dimension styles, plot styles. What breaks on import?
Week 2: Real work, side by side
Run a live micro-project twice — once in the current tool, once in the trial:
- Same deliverable, same deadline pressure. Note the friction honestly: command lookups, workflow dead-ends, "how do I even..." moments
- Measure the productivity valley: how much slower are you day one vs day five? The slope matters more than the depth — a tool you learn fast keeps paying
- Test the boring essentials: plotting to PDF at correct weights, batch operations, file open/save speed over your actual network paths
Week 3: The ecosystem check
- Integrations: PDM/PLM connector? ERP BOM export? Your plot queue? File version compatibility with your consultants' tools?
- Team reality: have a second person (ideally your most change-averse drafter) run their daily work in it for three days. Their friction list is the real one.
- Support probe: file an actual support ticket during the trial. Response time and quality during the sales cycle is the best-case scenario you'll ever see.
Week 4: The commercial x-ray
- Price the tier you actually need (the one with your must-have features — not the headline tier)
- Model three years: seats × price × term, with renewal uplift on subscriptions; compare against perpetual alternatives if they exist for this tool class
- Verify: licensing type (named vs floating), offline provisions, and what happens to your files if you leave
The scoring sheet
Rate 1-5 each, weight to your reality:
| Criterion | Weight |
|---|---|
| File fidelity on YOUR drawings | high |
| Automation survival | high if LISP-heavy |
| Learning curve slope | medium |
| Performance on worst-case files | medium |
| Ecosystem/integrations | per office |
| 3-year total cost | high |
| Support quality | medium |
A tool that wins on demo polish but scores under 3 on file fidelity is a future migration story — in the wrong direction.
Compare pricing, licensing, platforms, and feature data for 170+ CAD tools side by side at CADGuide.tools — free, no signup, no paid rankings.
