From Sales Sketch to Permit-Ready CAD: Fixing the Design Handoff

Why Engineering Redraws Every Sales Layout

Picture the closing call. The rep pulls up a satellite image of the roof, drops modules onto the two south-facing planes, lands on a 9.2 kW system, and the homeowner signs. The next morning that project hits the engineering queue. A designer opens AutoCAD, pulls the same address, and starts drawing roof edges from zero.

Nothing from the sketch came along. Not the roof outline, not the module count, not the spot where the rep already worked around the plumbing vent. The sales tool and the CAD tool never shared a design, so the sketch was really just a picture of one. That is the handoff gap, and on most teams it repeats on every single job.

The reason is usually mechanical. Sales layout tools were built to sell, so they export a proposal PDF, not drawing geometry. CAD expects real coordinates, layers, and dimensioned planes. When the two do not speak the same format, a human bridges the gap by re-measuring and re-drawing. The Department of Energy points out that accurate roof geometry and layout drive both the production estimate and whether the system can actually be built, which means the redraw is not a formality, it is a second chance to get those numbers wrong.

What a Sales Sketch Actually Holds

A decent sales sketch is more loaded than it looks. To produce a believable number, the rep already decided where the array sits, how many modules fit, which roof faces get used, and roughly how the panels are oriented. The tool used that to model production and price the job. Federal guidance for homeowners frames the same variables, panel count, orientation, and shading, as the inputs that set how much a rooftop system will generate over its life, per the DOE Homeowner's Guide to Going Solar.

So the sketch holds design decisions, not just a sales pitch. The problem is that those decisions live inside the sketch tool as a proposal, and a proposal does not open in CAD. When engineering cannot read the decisions, it remakes them. Sometimes the remade version disagrees with what the customer was sold, and now you have a scope conversation on top of a schedule slip.

What Should Survive the Sketch-to-CAD Handoff

A clean handoff means the CAD file inherits the sales design and the designer edits it instead of rebuilding it. Four categories of data have to make the trip intact for that to work.

• Roof geometry: plane outlines with real dimensions, plus azimuth and tilt for each face, so CAD is not re-tracing the roof.
• Module layout: exact module positions, orientation, row spacing, and the count the customer was quoted.
• Obstructions and setbacks: vents, chimneys, skylights, and the keep-out zones the rep already drew around.
• String plan: how modules group into strings and which inverter or optimizer channel they feed.
• Design assumptions: module and inverter models, the production estimate, and the site address the numbers were built on.

If those survive, the CAD designer opens a layout that already matches the sale and spends time on engineering, not archaeology. If they do not, every item on the list gets reconstructed from a flat image and a memory of the sales call.

Roof Geometry Is the First Thing to Get Lost

Roof geometry carries more weight than any other piece, and it is the piece most often thrown away. Redraw the planes from a fresh satellite trace and the array area shifts by a few square feet, the tilt guess moves a couple of degrees, and the production model quietly drifts from what the customer signed.

Geometry is also where code lives. Fire setbacks, pathway widths, and access clearances are dimensioned off the roof edges and ridgelines, and a permit set has to show them correctly. Those requirements trace back to the National Electrical Code and related fire provisions that the NFPA maintains as NFPA 70. When the roof is measured once, in the sketch, and that same geometry flows into CAD, the setbacks are drawn against dimensions the whole team agreed on. Measure it twice and you get two roofs, one sold and one built.

The amount of power a rooftop system produces depends on how well the design matches the actual roof, including panel orientation, tilt, and shading from nearby obstructions, which is why the DOE treats accurate layout data as the basis for a sound production estimate.

DOE — Solar Performance and Efficiency

Module Layout and the String Plan

Module layout is the visible part of the design, and it feels like the easy part to recreate. It is not. Row spacing, orientation, and the exact count around obstructions all affect area and production. A designer who nudges the layout to make the CAD look tidy can change the kilowatt total the customer was quoted.

The string plan is the part that almost never survives, and it is the part CAD needs most. Strings set the electrical design. String length drives voltage, and voltage has to stay inside code limits across the temperature range for the site, a calculation rooted in the NEC. When the string plan carries over from the sketch, the electrical engineer starts from real groupings. When it does not, strings get assigned late, and a layout that looked fine can turn out to be electrically awkward once the wiring is worked out. That is a redraw hiding inside a redraw.

Where the Handoff Breaks

The gap almost always shows up in the same few places. These are the failure modes worth naming so you can spot them on your own team.

• PDF-only handoff. The sketch tool exports a proposal PDF and nothing else. CAD cannot read geometry out of a PDF, so the designer traces over it by eye.
• Screenshot layouts. A rep pastes a satellite screenshot into the project notes. There is no scale, no dimensions, and no way to pull real measurements from it.
• Re-measuring the roof. The designer distrusts the sales measurement, so they re-trace the roof from scratch and now hold a slightly different array area than the one that was sold.
• Dropped obstructions. The vent the rep worked around never gets flagged in the handoff, so the CAD layout puts a module where a plumbing stack actually sits.
• Module-model drift. Sales quoted one module, engineering draws another because the exact part number never came across, and the wattage no longer matches the contract.
• Verbal string plans. The stringing lives in someone's head or a Slack message instead of the design file, so it gets rebuilt from guesswork.

Every one of these turns a fast handoff into a slow reconstruction. Most of them also create a version of the design that no longer matches what the customer signed.

What a Connected Sketch-to-CAD Workflow Looks Like

A connected workflow treats the sketch and the CAD set as two views of one design. The rep builds the sales layout, and that same layout opens in CAD as an editable drawing. The designer refines it toward permit quality instead of rebuilding it. This is the model behind moving a PVSketch layout directly into PVCAD, where the AutoCAD drawing inherits the sales design rather than starting from a blank sheet.

Here is how the two approaches compare across the data that matters.

From Carried-Over Layout to Permit-Ready Plan Set

Carrying the data over does not lower the bar for the permit set. It raises the odds of clearing it. A permit package still has to be NEC-compliant and pass the local authority having jurisdiction plan check, and those reviews look hard at setbacks, string voltages, and equipment ratings. The upside of an inherited layout is that the reviewer is checking one consistent design instead of a CAD set that quietly disagrees with the proposal.

Some jurisdictions have compressed the review itself. The DOE-backed SolarAPP+ tool runs an automated code check on standard residential systems so approvals come back in a fraction of the usual time, described by the DOE's SolarAPP+ program. Faster permitting only helps if the plan set feeding it is accurate, which loops back to the handoff. PVCAD produces permit-ready plan sets and NEC documents for systems up to about 5 MW, so the same inherited design that started in sales becomes the drawing the AHJ signs off on.

The Real Cost of Redrawing

The obvious cost is time. A redraw adds hours to days per project, and on a busy pipeline that is the difference between a permit set going out this week or next. Treat any specific hours figure as a directional estimate that depends on your own team and job mix, not a benchmark.

The quieter cost is error, and error is expensive because solar is expensive. Residential systems run roughly about $3 per watt before incentives by EnergySage's tracking, so a layout that has to be reworked after a failed plan check burns real money and schedule. The system also has to actually produce, because the customer is offsetting a utility bill, and U.S. residential electricity averaged around 16 to 17 cents per kilowatt-hour in recent EIA data. A production estimate that drifts during the redraw undercuts the savings the whole sale was built on.

Setting the Handoff Up on Your Team

Start by writing down what your handoff is supposed to carry, using the checklist above, and then check whether your current tools actually move those five things or just a PDF. Most teams find the answer is the PDF. From there, the change is to close the format gap so the sales layout and the CAD drawing are one design in two views. Reps keep selling in a fast web layout, engineers keep working in AutoCAD, and the roof, modules, obstructions, and strings ride along instead of getting rebuilt. That is the whole fix, and it removes an entire category of rework from the pipeline.

One more habit helps. Make the sales layout the design record of truth, so when a customer asks for a change after signing, the edit happens in the shared design and flows to CAD, rather than living as a note that engineering has to remember. When the sketch and the plan set can never drift apart, the question of which drawing is correct stops coming up, and the designer trusts what lands in the queue. That trust is what actually kills the reflex to start over.