Practice Management · Brief · Intro level
Client screening for incentive work: who should not claim
Not every client who could claim a credit should. The screening patterns — documentation-averse clients, aggressive-pressure clients, statute-expired opportunities — and how to decline gracefully.
The most valuable screening decision in incentive work is the claim not filed. A credit study can be technically competent and still be a bad idea for a particular client, because the claim's survival depends on things the client controls — records, candor, patience through an exam — and some clients reliably will not supply them. Firm-side screening is the complement to vetting the provider: the boutique screens for whether a credit exists; the firm must screen for whether this client can defend it.
The documentation-averse client
Research credit claims are documentation cases. Section 41 and Treas. Reg. §1.41-4 frame the tests, but exams are decided by whether contemporaneous records connect people, projects, and experimentation — see the substantiation expectations threaded through the IRS research credit pages and the business-component detail now demanded by Form 6765 Section G. A client who tracks no time, keeps no project records, and answers "we'll estimate it" to every request is proposing a claim built entirely on reconstruction. Estimation has a legal pedigree — Cohan v. Commissioner permits reasonable approximation on a credible base, and we map its limits in the R&D credit case law — but cases like Siemer Milling show what happens when the process of experimentation itself cannot be evidenced. Screen with one question: will this client adopt even minimal tracking going forward? If yes, the honest answer is often "document this year, claim next year — and consider the payroll offset then." If no, decline.
The aggressive-pressure client
Some clients arrive pre-sold — usually by a percentage-fee marketer — on a number, and experience every professional question as fee-padding obstruction. The tells are consistent: "the other firm said everyone qualifies," resistance to interviews, shopping the claim after hearing a caution, and demands to sign by Friday. This client is dangerous out of proportion to the engagement, because the pressure lands exactly where Section 6694 puts the firm's exposure: the signature on a position the firm was not permitted to test. Pressure this year becomes blame in the exam year. The acceptance decision is also the cheapest exit the firm will ever get — a declination costs a fee; a withdrawal mid-exam costs a client, a carrier notice, and sometimes a subpoena.
The statute-impaired opportunity
Look-back claims die on dates. The Section 6511 refund window — generally three years from filing or two from payment — closes quietly, and a claim filed against a closed statute is pure cost (see statute of limitations on refunds). Screen for the near-expired pattern too: a claim that must be researched, studied, and filed in three weeks to beat the statute will be a rushed claim, and rushed claims are how estimation becomes fabrication. Related patterns worth the same skepticism: closed years "reopened" through creative mechanisms a marketer proposed, and amended claims whose only support would be a study reverse-engineered years after the activity.
Declining gracefully
The declination is a deliverable. In writing: the benefit as estimated, the documentation gap or statute problem stated plainly, the realistic exam cost, and the conditions under which the firm would support the claim in a future year. Most clients hear a plan; the ones who hear an insult were the ones the screen existed to catch. Fold the screen into the firm's standing acceptance-and-continuance process — the annual mechanism described in quality control for the small firm — so the decision is systematic rather than heroic.
Frequently asked questions
- Which clients should a CPA steer away from claiming R&D credits or similar incentives?
- Three patterns predict trouble: clients who will not or cannot document their activities, since credit claims live or die on contemporaneous records; clients who pressure for the biggest possible number and treat professional caution as obstruction; and clients whose opportunity is already statute-impaired, where the refund window under Section 6511 has closed or nearly closed. Each pattern converts a tax benefit into an exam and liability problem.
- How should a firm decline a client's credit claim without losing the relationship?
- Frame it as a benefit-and-risk calculation, in writing: what the documentation would need to show, why the current records cannot show it, what an exam would cost in fees and penalties, and what would have to change for the claim to work next year. Offering the forward path — better time tracking, a documentation process — turns a refusal into planning, and it filters out only the clients who wanted a number rather than a position.
- Can a client with poor documentation ever support a research credit?
- Sometimes. Courts have allowed reasonable estimation where a credible evidentiary basis exists — the Cohan doctrine — but estimation is a fallback with limits, not a substitute for records, and cases like Siemer Milling show claims failing where documentation of the required experimentation simply was not there. A client unwilling to build records going forward is betting the claim on judicial charity.