Sea Trials, Change Orders, and Mold Revisions: The Boatyard Records That Can Support an R&D Credit

TL;DR

• The regulations don't prescribe a single record format. A claim needs records "in sufficiently usable form and detail" to show the expenses are eligible, and boatyards already create many of them.
• The most useful records show what was uncertain, which alternatives you tried, and how you evaluated them. Design revisions, layup and material trials, sea trial logs, and engineering change notices all do this.
• A few light habits can make next year's records far easier to use, like project codes on hours and a one-line "why" on every change.

A new hull rarely goes from drawing to dealer without surprises. The running surface gets reworked after the first sea trial. A core material that looked right on paper delaminates in testing. The helm station moves six inches because the sightlines were wrong. That work is often exactly what the federal research credit under IRC Section 41 was written to reward: working through technical uncertainty with a structured process of evaluation.

What many builders worry about is proof. The good news is that the regulations don't require a special R&D notebook. Treasury Regulation 1.41-4(d) asks a company claiming the credit to keep records "in sufficiently usable form and detail" to show that the expenses are eligible. The IRS's own research credit guide on substantiation lists ordinary business records as the kinds of documents that help show what was done. Examples include project summaries, meeting notes, test data, work orders, and contracts.

Most boatyards already produce those records. They just aren't organized around the credit. Here is where to look.

Start with the component, not the whole boat

The credit is evaluated one "business component" at a time: a product, process, technique, or piece of software. Treasury Regulation 1.41-4(b)(2) sets out what's called the shrinking-back rule. If a whole new model doesn't qualify as a unit, the analysis can move down to smaller elements until it finds the piece where the qualifying work happened. For a boat, that might be the hull, the deck mold, the drive and propulsion package, or the electrical system.

That matters for your records. Documentation is most useful when it ties to a specific component. "We built the new 26" is hard to support. "We redesigned the running surface of the 26 to reduce porpoising at cruise, tried three strake configurations, and tested each on the water" is much easier.

Record 1: Design files and their revision history

Your CAD models, hull lines, stability and weight calculations, and structural drawings are often the backbone of a claim. What matters here is less the final design than how you got to it.

What helps most:

• Saved versions rather than overwritten files, so earlier alternatives still exist
• Dates on revisions, so the order of changes is clear
• Notes from design reviews explaining why an option was rejected

The regulation describes a process of experimentation as three steps: identify the uncertainty, identify alternatives, and evaluate them. A revision history with even brief notes can show all three.

Record 2: Layup schedules, material trials, and failed parts

Composite work involves a lot of testing that never gets written down. A new laminate schedule, a switch in core material, or a different resin system can all involve real technical uncertainty, depending on the facts. So can an infusion process that has to be tuned before it fills cleanly.

Records that can support this work include:

• Test panels and their results
• Supplier data sheets you compared
• Cure and fill notes from trial parts
• Photos of parts that failed, with a note on why

A phone photo of a delaminated test panel with a date and a one-line caption is a surprisingly useful record.

Record 3: Sea trial and on-water test logs

For many builders, the water is the lab. Sea trials that compare prop sizes, engine placement, trim tab settings, or hull changes can be strong evidence that you evaluated alternatives.

A useful log usually captures:

• The date, hull, and configuration being tested
• What question the run was meant to answer (speed, trim, fuel burn, handling, vibration, noise)
• The measurements or observations
• What changed afterward, if anything

A routine delivery run or a customer demo is different. Those usually aren't part of the analysis. The question is whether the run was designed to test something technical.

Record 4: Engineering change notices and change orders

Change paperwork often holds the clearest record of problem solving in a boatyard, because it captures the moment something didn't work as planned. The weakness is that many change notices record what changed but not why.

Add a single line for the reason, for example: "hatch drain undersized, water pooling at 15 degrees of heel, increased to 1.5 inches and retested." That turns an administrative form into a record showing the uncertainty, the alternative, and the evaluation.

Record 5: Who did the work

Wages are typically the largest part of a claim, and they usually cover more people than the design team. Under Section 41, wages can count when they're paid for directly performing, directly supervising, or directly supporting qualified research. In a boatyard, that might include laminators building test parts, the shop lead running a trial, or the technician rigging a sea trial, depending on what they actually did.

Payroll records help most when they're combined with anything that links people to projects, such as job codes, work orders, schedules, or project emails. Formal timesheets are useful, but they aren't the only way to show this.

Record 6: Materials and outside help

Supplies used up in the research can count. That includes resin, cloth, and core used in test panels and prototype parts. Depreciable property generally doesn't count as a supply expense under Section 41(b)(2)(C), and that often includes production molds and tooling. Keeping prototype material costs separate from capital purchases makes this easier to sort out.

If you pay an outside naval architect or engineer, keep the agreement and the invoices. Qualifying contract research can generally count at 65% of the amount paid. That only applies when the agreement meets specific requirements, so the contract itself is part of the record.

Know where the line falls

Research that happens after a component is ready for commercial production generally doesn't qualify. Treasury Regulation 1.41-4(c)(2) excludes work done after a component "meets the basic functional and economic requirements" for sale or use. For a boat builder, routine production of an established model, cosmetic changes, and ordinary quality control usually fall on that side of the line.

That makes dates matter. Records that show when a hull or system moved from development into regular production help separate work that may qualify from work that doesn't. A model-year change that raises new technical questions can be a new round of development and may be judged on its own facts.

Three habits worth starting now

You don't need to overhaul how the shop runs. A few small habits can make next year's records much easier to use:

1. Give development work its own job codes. When a new hull or system is in development, a dedicated code on hours and materials keeps that cost from blending into production.
2. Write one line of "why" on every change. It takes seconds, and it's the most useful thing you can add to change paperwork.
3. Date and label test runs and trial parts. A short sea trial log and dated photos of test panels cover a lot of ground.

Next steps

If your team is developing new hulls, systems, or build processes, you may already have much of what a claim relies on. Organizing it is the hard part, and that's where a specialist can help. For one example, Solace Boats identified more than $300K in R&D credits. Individual results vary depending on the specific activities, documentation, and facts.

Strata starts with a $0 initial assessment to see whether a study makes sense. Studies typically take about 4 to 8 weeks, depending on the company and the years involved. Your team's time commitment is modest because most of the work sits with us. You receive a documentation package built to support the claim, which you pass to your CPA to file with the return.

Talk with Strata about your projects

For the form used to claim the credit, see the IRS Instructions for Form 6765.

This post is for informational purposes only and does not constitute tax or legal advice. Consult a qualified tax professional regarding your specific circumstances.