An R&D experiment log is a record of individual experiments — a structured entry for each test, evaluation, or iteration that documents what was tested, how, and with what result. Where project documentation captures the overall story of a project, an experiment log captures the detail of each experiment within that project. This page explains what an R&D experiment log is and what each entry should capture. It is educational and is not individualized advice. For the project-level companion, see our page on R&D project documentation.
What an R&D Experiment Log Is
An R&D experiment log is a structured record of individual experiments. Each entry in the log corresponds to one experiment, test, evaluation, or iteration, and captures the key information about that experiment: what question it was trying to answer, what approach was taken, what alternatives were evaluated, what the test showed, and what happened next. The log connects experiments to the project they belong to and to the evidence they produced.
What Each Entry Should Capture
A useful experiment log entry captures several elements:
Question
What technical question was this experiment trying to answer? What uncertainty was it intended to address? The question should be specific and technical, not a general business goal.
Hypothesis or Approach
What was the expected outcome, or what approach was being tested? What did the team expect to learn? For more on the concept, see our page on documenting technical uncertainty.
Alternatives
What alternatives were being evaluated? If the experiment compared multiple approaches, materials, or configurations, which ones, and how did they differ? For more on the evaluative-process concept, see our page on process of experimentation.
Method
How was the experiment conducted? What was the test method — modeling, simulation, physical testing, systematic trial and error, or another evaluative process? The method should be described in enough detail that someone reviewing the log later can understand what was done.
Test
What was actually tested? What were the inputs, conditions, and parameters? What was measured or observed?
Result
What did the test show? What was the outcome — did the alternative work, did it not work, or was the result inconclusive? The result should be recorded as it was observed, not as it was hoped to be.
Unexpected Outcomes
Were there any unexpected results — outcomes that differed from the hypothesis or that raised new questions? Unexpected outcomes can be important context, and recording them preserves information that might otherwise be lost.
Next Step
How did the result inform the next step? Did the team proceed with the alternative, revise it, try a different alternative, or discontinue the line of inquiry? Recording the next step helps show the progression of the evaluative process. For more on documenting iterations, see our page on documenting trial and error in R&D.
Linked Evidence
What evidence did the experiment produce — test data, measurements, photographs, design files, or other artifacts? The evidence should be linked to the experiment entry, not just stored in a separate file. For more on evidence organization, see our page on R&D evidence tracking.
Date and Participants
When was the experiment conducted, and who participated? The date helps establish the timeline, and the participants help connect the experiment to the personnel involved.
Formal Lab Logs Are Not Universally Required
A common question is whether R&D experiment logs must be formal scientific lab notebooks — the kind used in academic or industrial chemistry laboratories. They do not. There is no universal requirement that R&D documentation take the form of a formal lab notebook. What matters is whether the records, taken together, capture the technical substance of the work — the question, the approach, the alternatives, the test, the result, and the next step. A structured digital log, a spreadsheet, or a purpose-built tool can serve the same purpose as a paper lab notebook, provided it captures the same information. For more on the documentation concepts, see our page on R&D tax credit documentation.
Why an Experiment Log Matters
An experiment log helps show the evaluative process — the sequence of questions, tests, and results that constitutes R&D. Without a log, the experiments may be remembered only as "we tried some things and it worked." With a log, the sequence is preserved: what was tried, what it showed, and how it informed the next step. This level of detail can be important for understanding the technical substance of the work, whether for internal review, for professional review, or for responding to questions.
Contemporaneous Entries
Experiment log entries should be made as the experiments are conducted — contemporaneously — not reconstructed later. At the time of the experiment, the team knows what they were testing, what they expected, and what they observed. Months later, those details can be harder to recall accurately. For more on why timing matters, see our page on contemporaneous R&D documentation.
What an Experiment Log Does Not Do
An experiment log does not establish that a business's activities qualify for any tax credit. It does not determine whether costs are qualified research expenses. It does not replace the professional judgment of a CPA or qualified tax professional. It does not make records "IRS-compliant" by itself. It is a record of experiments; the substantive analysis is a separate step.
Key Takeaway
An R&D experiment log is a structured record of individual experiments, with each entry capturing the question, hypothesis or approach, alternatives, method, test, result, unexpected outcomes, next step, linked evidence, date, and participants. Formal scientific lab logs are not universally required — what matters is whether the records capture the technical substance of the work. An experiment log does not establish tax-credit eligibility or replace professional review; it helps preserve the detail of the evaluative process. For a companion on documenting the iterative process, see our page on documenting trial and error in R&D.