Contract-to-cash is a chain, not a feature list

Finance teams do not buy "AI." They buy the absence of an invoice that is wrong in front of a customer, and a close that does not require a second workbook to explain the first.

That is a chain. Customer (or shipper). Product. Price. Contract. Schedule or rated haul. Invoice. Cash. Journal. Skip a link and the next tool starts guessing.

Why stitched stacks invent a second set of numbers

Most revenue stacks skip links on purpose. A CPQ never met the meter. A billing product never met the bank. Collections lives in a shared inbox. Revenue lives in a sheet named final_v7. Each system is locally reasonable. Together they invent a second set of numbers.

The failure mode is predictable:

• The contract says one thing; the schedule says another because someone re-keyed it.
• Usage arrived late; billing added a line to a sent invoice because the meter tool could.
• Collections chased a wire that landed last Tuesday because bank rec is a Friday ritual.
• Revenue recognised on invoice date because the spreadsheet could not hold haul date and invoice date separately.

None of these are operator errors in isolation. They are what happens when each product owns a slice and nobody owns the handoff.

What Sonic spans — and what it does not

Sonic's bet is span on the chain from agreement to journal, scoped to one organisation at a time.

AI belongs on judgement steps: reading the PDF, suggesting the next collections sentence, interpreting a messy statement line. It does not belong on inventing an amount after send. The math is the schedule and the rate card. The model is not a second general ledger.

The chain in practice — subscription

A typical subscription path looks like this:

1. Upload the signed contract PDF.
2. Review the structured extraction — phases, products, one-time fees, payment gates if the clause says so.
3. Approve to materialise the customer, products, and billing schedule.
4. Activate the schedule; invoices generate on billing day.
5. Send the invoice; only then does Watchtower age it.
6. Customer pays via Stripe, Razorpay, or wire; you match the deposit in bank rec.
7. Journal entries post from the invoice and recognition rules you configured.

If step 3 is wrong, everything downstream is wrong with confidence. That is why we insist on review before approve, not blind trust.

The chain in practice — freight

Shipment billing is the same chain with different nouns:

1. Carrier contract with rated charges and file mapping.
2. Upload the carrier export; map shippers; create shipments.
3. Invoice preview groups by contract, shipper, and period.
4. Send; collections chases the shipper, not a subscription customer.
5. Recognition posts on the haul date; billing posts on invoice date.

There is no monthly schedule pretending to be freight. Mixing the two modes in one org workspace is a product choice, not a configuration toggle you can flip per invoice.

Correct, then fast

If you take one sentence from this essay, take this: correct, then fast.

Speed without a frozen issued invoice is just a faster way to be wrong. Sonic will refuse to add lines to a sent invoice. It will refuse to silently rewrite a carrier rate on an issued freight document. It will stop dunning when cash matches, not when someone remembers to check.

That feels slower than "just fix it in the system." It is how you remain able to answer, without sweating, what the customer was asked to pay — and what your books say you earned.