Skip to content

RapidDraft Refined Search Loop

Purpose: Stable playbook for how RapidDraft narrows the broad TextCAD ICP into a pilot-friendly, release-review wedge. Last updated: April 2026

Why this loop exists

The shared ICP is directionally correct, but the first round of company research was still too broad. It mixed true design-owning machine builders with firms that were better matches for a narrow DFM pitch, too small to matter, or too large and tool-mature to move quickly.

The refined loop sharpens the wedge around one specific problem:

Structured mechanical teams still spend too much time manually reviewing, documenting, and releasing design changes.

RapidDraft is therefore positioned first as the missing review layer around existing CAD and PDM or PLM workflows, not as a generic AI design tool and not as a cheap replacement for mature automation stacks.

Refined ICP

Company profile

The best RapidDraft targets now look like this:

Dimension Refined target Why it matters
Business model Design-owning mechanical OEM or machine builder They own the design decisions and release their own CAD and drawings
Core sectors Sondermaschinenbau, packaging machinery, factory automation, pumps/valves/HVAC, test and measurement, lab equipment These sectors combine variant pressure with real review and release discipline
Workflow shape Cross-functional review across design, manufacturing, quality, and sometimes suppliers This is where comments, revisions, and release friction pile up
CAD/PLM maturity NX, SolidWorks, Teamcenter, PDM, or another structured release backbone Base CAD/PLM maturity is a positive, not a disqualifier
Size Usually 50-500 employees, sometimes up to ~1000 if still engineering-led Large enough to feel the pain, small enough to pilot
Geography Germany first, DACH next Highest founder reachability and strongest fit with the current wedge

Person profile

The engineer feels the pain, but the practical buyer is usually one of these roles:

  • Leiter Konstruktion / Head of Mechanical Design
  • Engineering Manager / Head of Development
  • Manufacturing Engineering or Industrial Engineering lead
  • Quality or compliance lead when review traceability matters

Likely day-to-day champions are senior designers, lead mechanical engineers, CAD or PDM admins, and NPI leads.

Negative Filters

The refined loop rejects companies when one of these signals is true:

  1. They are the wrong business model: pure CNC shops, pure contract manufacturers, pure molders, or firms mainly building from customer drawings.
  2. They already signal a solved adjacent stack: DFM platforms, drawing automation, configurators, or a dedicated design-review layer.
  3. They are too large or too corporate for a fast first pilot.
  4. They are too small or too unstructured to have repeatable release pain.
  5. They look like lighthouse customers, resellers, or unusually advanced automation showcases for the exact category RapidDraft is trying to wedge into.

The important nuance from the refinement pass is that CAD/PDM/PLM maturity is good. The negative signal is not NX, Teamcenter, or SolidWorks. The negative signal is evidence that review collaboration, drawing automation, or manufacturability tooling is already solved elsewhere.

Scoring Model

The April 2026 research pass used the following weighted scoring model:

Criterion Weight What it tests
Design ownership 25% Does the company own product definition and release its own designs?
Cross-functional review pain 25% Are design, manufacturing, quality, and suppliers likely to review changes together?
Absence of competitor / automation signal 20% Is there no strong public signal that the review problem is already solved?
CAD/PLM fit 15% Is there evidence of NX, SolidWorks, Teamcenter, PDM, or disciplined release workflow?
Pilotability / speed 10% Can an engineering-led buyer likely move on a pilot?
International / English friendliness 5% Is founder-led outreach easier because the company exports or operates in English?

This pushes the loop toward mature-but-unsolved teams: enough structure to feel release pain, not so much tooling maturity that RapidDraft becomes a replacement pitch instead of a missing-layer pitch.

Closed Loop

The loop is intentionally simple:

  1. Refine the ICP. Tighten the company and buyer profile around one pain pattern instead of chasing every manufacturing company that looks plausible.
  2. Find companies. Search by sector, geography, CAD/PLM signals, and engineering-ownership signals.
  3. Rescore and reject aggressively. Treat explicit rejections as useful output, not wasted work.
  4. Run outreach against the pain angle. Lead with revision comparison, release traceability, drawing checks, and review collaboration.
  5. Capture what we learn. Which titles reply, which wedges resonate, and which sectors feel too mature or too weak.
  6. Update the loop. Feed the learning back into the ICP, filters, and target lists before the next search pass.

The result is a tighter search and outreach cycle instead of a one-time company dump.

Claudio Validation

Meeting_with_Claudio.md is the canonical meeting summary, and meeting_with_claudio.md is the underlying source note used for this wedge.

Claudio works in an NX plus Teamcenter environment and describes a change process where the actual geometry change is often quick, but the release workflow around the change is slow. The real pain is not "how do I model the part?" It is "how do I create the revision, document what changed, route the DCR, and keep the review chain traceable without doing repetitive manual work?"

Three takeaways matter for the loop:

  • The problem is strongest in structured environments, not immature ones.
  • Teamcenter and NX do not remove the review burden; they often make the workflow more formal and more click-heavy.
  • "Auto-generate the change report" is a concrete wedge that maps directly to the broader RapidDraft thesis around revision comparison, release checks, and traceable collaboration.

That is why this loop prioritizes companies with real CAD or PLM maturity and visible release pressure rather than searching for the least advanced teams.

Practical Output

The current output of this loop lives in RapidDraft_Refined_Targets.md, and the canonical click-through directory lives in Target_Companies.md.

Only meeting_with_claudio.md is used from the wiki inbox for this improvement loop. The two inbox PDFs are intentionally excluded because they do not add customer-targeting evidence.

Open Questions

  • Which borderline companies should move up next after direct founder outreach and website revalidation?
  • Should the next iteration widen more aggressively into pump, valve, and HVAC accounts or stay tighter around machinery first?
  • Which specific outreach messages best translate "auto-generate the change report" into a credible first pilot ask?

Sources