No. The system prompt forbids quoting factor rates or APR equivalents because those depend on underwriting (card volume, bank balance, time in business, industry, prior MCA history). Quoting a factor rate from a chatbot is exactly the kind of advertising claim regulators have flagged in enforcement actions. The bot frames the typical range honestly (factor rates commonly fall in a broad range depending on risk) and routes every quote to a human underwriter.

Yes, at a framing level. New York, California, Utah, Virginia, and other states require commercial financing providers to surface APR-equivalent disclosures alongside factor rates on MCA offers. The bot captures the borrower's state early and frames that the offer will include the relevant state disclosure if applicable. The actual disclosure is produced by your underwriter or compliance system; the bot just sets the expectation. That is the framing regulators reward.

It says so. When a borrower has strong credit, several years in business, and a stable bank profile, MCA is rarely the right product; a line of credit or term loan usually fits better. The system prompt is configured to surface that routing honestly. Pushing MCA on every visitor is exactly the failure mode the FTC and state AGs have flagged repeatedly, and the system prompt actively avoids it.

Yes. The bot explains that holdback is the agreed percentage of daily or weekly card receipts that the funder collects until the purchased receivables are fully delivered. It frames the practical impact on cash flow (lower daily deposits) and the structural difference from a fixed-payment loan. The actual holdback percentage on a specific deal routes to the underwriter; the bot just explains the mechanic clearly.

Carefully. Stacking (taking a second MCA on top of an existing one) and renewals are loaded topics that have drawn regulator scrutiny for collection-practice and double-payment issues. The bot frames the questions honestly, surfaces the policy of your firm (some brokers stack, some refuse), and routes every actual stacking or renewal decision to a human. Stacking is one of the conversations where bot improvisation is most dangerous, so the prompt is configured to be conservative.

Yes, and you should. Every conversation stores in WordPress with timestamp, page URL, model name, and token count. CSV export is available for FTC, state AG, or state DFI inquiries. MCA is among the most regulator-watched products in small business funding, so a fully reviewable transcript log is genuinely valuable. Most brokers archive transcripts on a recurring webhook into their compliance system so the chat log lives alongside advertising and email evidence.

Claude Sonnet and GPT-4o both follow strict 'no factor rate quotes, no overselling' system prompts reliably and handle disclosure framing accurately. Smaller models occasionally drift into salesy phrasing or downplay cost, which is the specific failure mode MCA enforcement actions have flagged. Bring your own key for OpenAI, Anthropic, Google, or OpenRouter and pick based on your firm's risk tolerance.

No, and it actively prevents that confusion. MCA is a purchase of future receivables, not a loan, and the legal distinction matters for usury, disclosure, and collection rules. The bot frames MCA as 'a purchase of future receivables' every time and does not use loan terminology (interest, principal, term in the loan sense). That language discipline is exactly what compliance officers want to see and what regulators have flagged when it has been missing on other broker sites.