Sales-to-CS handoff briefing

A sales-to-CS handoff briefing is one packet the onboarding owner can open the morning of kickoff: who bought, what is on the order, what discovery actually confirmed, and which promises the AE made on the record. It is not a recap of every sales call. It is not a story a model infers because a name appeared once on a recording.

The quality bar is citation. Every stakeholder, SKU, timeline, and commitment in the brief should point back to a source CS can check: the opportunity and order in Salesforce, the signed proposal or contract extract, or a timestamped moment in a conversation intelligence recap from a Gong or Read.ai class tool. If CS cannot show where a line came from, it does not belong in the packet.

Kickoff goes wrong when CS walks in with a story sales told itself. The buyer who "will be the champion" never joined a call. The analytics module everyone discussed is not on the order. The aggressive go-live date was an AE hoping out loud. A cited briefing prevents that by making the Closed Won record the only input that survives into CS. Companion extracts help: promised-commitments extractor for verbal and written promises, and signed-contract data extraction for names, terms, and line items copied rather than paraphrased.

Pull four classes of record, and stop there. The rest of the trail stays in CRM and conversation tools as reference, not as the brief.

Confirmed discovery. Keep the discovery notes and call moments that still describe the deal that closed: problem, current stack, constraints, decision process, and named buyers who actually spoke. Conversation intelligence tools in the Gong or Read.ai class are useful as indexes into those moments. They are not the briefing. A recap is a search aid. The brief is a curated extract with citations.

Proposal. Take the proposal the customer accepted, not the earlier decks that still list options. Scope, implementation shape, and commercial framing belong here only as they appear in the version that won.

SKUs on the order. Line items, quantities, terms, and start dates come from the order object or signed schedule, typically sitting next to the opportunity in Salesforce. Anything discussed on a call and then dropped from the quote is out of the SKU list.

Cited promises. Verbal and written commitments that survived Closed Won: timelines, integrations, training, executive access, custom work. Each needs a citation (proposal clause, email, or call timestamp). Unsourced "we said we would figure it out" does not ship.

Do not invent context around those four. If discovery never named a security reviewer, the brief does not add one because "enterprise deals always have legal." If the transcript is messy, quote the confirmed fragment and leave the rest out. Dumping the full call trail into CS is how kickoff becomes a second discovery instead of a start of work.

Treat Closed Won as the freeze point. After the opportunity is Closed Won, the AE and the CS onboarding lead assemble one document CS will own from that moment. Sales can comment. CS edits and ships it.

Work in this order so the order constrains the story, not the other way around.

Freeze the commercial record. Pull the Closed Won opportunity, the accepted proposal, and the order lines. Copy legal names, term dates, products, and special terms. Do not rewrite them into friendlier language that quietly changes scope.

Attach discovery that still holds. From conversation intelligence and CRM notes, keep only facts that still match the signed scope: who attended, what problem they bought to solve, constraints that affect implementation. Timestamp each claim. Drop the rest of the trail, including objections that were already resolved and SKUs that were declined.

List promises as citations, not slogans. For each commitment, write the promise in the customer's words or the proposal's words, then the source. If you cannot cite it, delete it. This is also where you notice delivery tension without turning the brief into a risk register. When a promise conflicts with the SKU list or the written timeline, park that conflict for implementation risk flagger rather than smoothing it out of the packet.

Name people who appeared. Stakeholders are people who joined a call, signed, or are named as buyer, user, or approver in CRM with a role. Do not promote a forwarded email or a "we should bring Priya in later" into an owner.

Hand ownership to CS. Put the packet where onboarding already works, often a Gainsight-class CS workspace linked from the Salesforce account, not in the AE's private notes. From here the briefing feeds the first success plan. It does not replace auto-drafted success plan work. It is the cited input that plan should not contradict.

Illustrative path through one deal. An AE closes a platform subscription plus professional services. Discovery calls, indexed in a Gong or Read.ai class recap, mention an analytics add-on and a "VP of Ops, maybe joining next week." The accepted order has platform and services only. The briefing CS owns lists those two SKUs, the economic buyer who signed, the users who spoke on the calls, and the implementation window written in the proposal. The analytics add-on sits under a labeled "discussed, not sold" note. The VP of Ops is omitted until they appear on a recording or in CRM. Kickoff then starts from the order, not from a hopeful recap.

If CRM, CS, and a shared workspace each keep a copy, a cross-system handoff agent should move the same packet rather than letting each tool keep a different story.

Three failure modes show up in the opening of kickoff if you skip citation.

Inventing a stakeholder. A name in a recap, a CC on an email, or a role the model infers from "enterprise" is not a kickoff attendee. CS then emails the wrong person, or walks in assuming an executive sponsor who does not exist. Rule: if they did not speak, sign, or hold a CRM role on the opportunity, they are not in the people section.

Dumping the Gong or Read.ai class recap as the brief. Recaps are chronological and complete. Briefings are scoped and cited. Pasting the recap trains CS to re-listen to the deal. It also smuggles in objections that were resolved, SKUs that were declined, and offhand remarks that read like commitments. Use the recap to find timestamps. Quote the line. Leave the transcript in the tool.

Including a SKU that is not on the order. Sales conversations explore. Orders decide. If CS onboards an add-on that never billed, implementation scope, training, and success criteria all drift. The SKU section is a copy of the order lines, checked against the proposal. Discussed-not-sold items, if they must appear, sit in a clearly labeled parking lot so nobody builds a project plan from them.

A quieter fourth failure is invented context: filling gaps so the story reads smoothly. Gaps are useful. They tell CS what to confirm in kickoff instead of pretending discovery already answered it.

The briefing is a CS record with a sales provenance, not an AE leave-behind.

Store it on the account in Salesforce so anyone covering the book can find the Closed Won snapshot, and in the CS system of record (Gainsight-class) so onboarding tasks attach to the same packet. Do not keep the "real" version in a slide the AE presented internally.

Version once at Closed Won. If the customer changes scope later, that is a commercial change, not a silent edit to the original brief. CS can add kickoff notes in a separate section. The cited Closed Won facts stay frozen so you can see what was promised versus what later changed.

Access should be the onboarding lead, the AE of record, and the CSM who will run kickoff. Broad "whole company" sharing invites drive-by edits that reintroduce unsourced claims.

The packet is ready when a CSM who did not sell the deal can open it and answer, with citations: what was sold, who the customer-side owners are, which promises are in force, and which discovery facts still constrain implementation. If any of those four require a guess, the briefing is not done.

The AE and the CS onboarding lead should both sign off. The AE confirms nothing material was dropped from the commercial record. CS confirms they can run kickoff from this document alone. That is the quality outcome: a cited packet CS can walk into kickoff with, not a transcript dump, and not a story nobody can source.