Cost Guru, Affordable AI Implementation

R&D Audit Risk Scoring

A score that shows where a claim is exposed. It combines four-part test support, documentation gaps, wage and contractor ratios and the completeness of Section G, and it shows the reasons behind the number.

What the score is

A claim has many features that an examiner might look at, and it is hard to hold them all in mind. Risk scoring combines several of them into one figure and, more usefully, into a list of the areas that pull the score in each direction. The score is a summary. The reasons are the point. A firm that sees only a number learns little. A firm that sees why the number is what it is knows what to do.

What it looks at

The score looks at four-part test support, documentation gaps, wage and contractor ratios and the completeness of Section G. Each is checked on its own, and results are combined. Which factors count most, and how they are weighed, is confirmed by Swanson Reed. What matters for users is that each factor is visible, so a reviewer can see which one is driving the result.

The areas and what they ask

Descriptions are general. Exact tests and weighting are confirmed by Swanson Reed.

AreaTypical question
Four-part test supportDoes the evidence show each part of the test for each project?
Documentation gapsWhich records are missing or thin, and how much of the claim do they affect?
Wage ratiosDo the shares of wages claimed look supported and reasonable?
Contractor ratiosIs the proportion of contractor costs supported by the records?
Section G completenessIs the business-component detail complete and consistent with the rest of the claim?

How it works

  1. 1

    Assess

    Each area is checked against the records loaded.

  2. 2

    Score

    Results are combined into a risk score, with the contribution of each area shown.

  3. 3

    Act

    You see which areas to improve first, and what would improve them.

  4. 4

    Re-score

    After fixes, you run the score again to see what changed.

What you get

  • One score, with the reasons behind it.
  • A way to prioritize fixes, starting with those that move the score most.
  • A shared language for talking about risk with clients and colleagues.
  • A before-and-after view, so you can show what was improved.

A worked example (illustrative)

This scenario is invented, and no real scoring is shown. A claim receives a high-risk result. The breakdown shows that most of the weight comes from two areas: documentation gaps on one large project, and a high share of wages claimed with no supporting time records. Test support and Section G look reasonable. The firm gathers time records and fills the gaps on the large project. On the second run, the score falls, and the breakdown shows the same two areas improved. The firm keeps the two reports together as a record of what changed.

Using scores well

Scores can be misused. Treat a low score as a sign that fewer weaknesses were found in what was tested, not as a sign that the claim is safe. Treat a high score as a reason to look closely, not as proof that the claim is wrong. And do not use the number as a target in itself: improving a score by understanding a finding and fixing it is progress, while improving it by working around the test is not. The most useful habit is to read the breakdown, not just the headline.

Limits

The score depends on the records loaded and the way the factors are defined. It does not know facts that nobody wrote down, and it is not a prediction of examination or of the outcome of one. A mock audit does not guarantee acceptance of a claim or predict whether it will be examined. It is not tax or legal advice, and a licensed CPA or enrolled agent should review any claim before it is filed.

A mock audit does not guarantee acceptance of a claim or predict whether it will be examined. Integrations, review tiers and partner pricing are proposals until confirmed by Swanson Reed.

Questions to ask when you see a score

  • Which area contributes most, and is it something we can fix?
  • Is the weakness about records, about wording or about substance?
  • Has anything changed since the last score, and why?
  • Would a reasonable reviewer reach the same view by reading the file?
  • What would we say if an examiner asked about this area?

Scores across a portfolio

When a firm scores many claims, patterns appear. A weakness that shows up in every claim, such as round-number time allocations, points to a habit that can be fixed once, upstream. A weakness in a single claim points to something particular to that client. Looking at scores side by side is therefore useful beyond any one claim: it tells the firm what to change in its own process. The portfolio dashboard is built for this view.

Explaining a score to a client

Keep it plain. Say that the score reflects how well the claim's records support what it says, that it is a way of finding things to improve, and that it does not predict what a tax authority will do. Show the breakdown, and focus on what the client can do. Avoid presenting the number as a grade or a guarantee. Clients who understand that the aim is a well-supported claim tend to engage with the findings rather than argue with the score.

Scoring questions

What is a good score?

There is no number that makes a claim safe. Use the breakdown to decide what to fix, and let your adviser decide whether the claim is ready.

Who decides how factors are weighted?

That is confirmed by Swanson Reed. We do not state the method on this site.

Can I see why a score changed?

The intention is that you can, by comparing the breakdown before and after.

Is the score shared with clients?

That is up to the partner. Think about how you describe it, so that it is not mistaken for a promise.

Please note: Estimates and drafts are prepared by software for review by a licensed CPA or enrolled agent. This is not tax advice.

Find out what AI would cost you

Book a free call. We will map three opportunities and give you a rough idea of cost and payback.