How to Run an AI Sales Cycle, From Discovery to Handoff

AI deals rarely die in the first meeting. They die in a pilot that never ends, a usage forecast nobody wrote down, or a security review that starts two weeks before the quarter closes.

Each of those problems starts earlier, when the account executive skipped a question the buyer would eventually need answered.

This plan runs a deal in seven stages. Each stage has a job, something you write down, and an exit gate: what must be true before the next stage starts. It was put together from public vendor documentation and practitioner write-ups, checked on September 26, 2026. It's a process to adapt, not a record of deals anyone ran with it.

The seven stages at a glance

Stages 3, 4 and 5 should overlap. Start security review as soon as the buyer agrees to test the product. Waiting until the test passes is the most common way to lose a quarter.

A worked example

This is a made-up deal, used through the rest of the guide to show what each stage looks like.

Alex sells an AI support agent. A mid-sized online retailer, call it Cedar Goods, says it wants "to use AI in support." Alex doesn't have a deal yet. By the end of discovery, Alex has one: the billing team spends a full day answering each tier-one refund question, and every month-end leaves a backlog. The support director owns that queue and wants it fixed before the holidays.

Alex keeps coming back to Cedar Goods below.

Stage 1: Discovery that finds the workflow

The goal is one specific workflow, not general interest in AI. "We want to use AI in support" isn't a deal. "Tier-one billing tickets take our team a day to answer, and we have a backlog every month-end" is the start of one.

A common checklist for qualifying a deal is MEDDICC: metrics, economic buyer (the person who controls the budget), decision criteria, decision process, identified pain, champion (the person inside who wants this to happen) and competition. For AI deals, add three questions:

• What have you tried already? Many buyers have built a prototype in-house or tried another tool. What broke tells you what your test has to prove.
• Who judges whether an answer is right? Someone has to. Find that person now. It's often not the buyer.
• What happens when the model is wrong? The answer sets how much human review the workflow needs, which shapes both the test and the price.

At Cedar Goods, the person who judges answers is a senior billing agent, not the director. Alex gets her on the next call.

Output: discovery notes that separate what the buyer said from what you guessed. The prepare-discovery and qualify-opportunity skills in the site's open-source GTM skills are built around that separation.

Stage 2: Technical qualification

This stage answers one question: can this customer actually use the product? Bring a sales engineer or forward deployed engineer to this call if your company has one.

Check four things.

1. Data. Where it lives, what format it's in, and whether the customer can share a sample for the test. Regulated data will change the timeline.
2. Integration. Which systems the product reads from and writes to, and who on the customer's side owns them.
3. Deployment. Whether the customer needs a specific cloud, region, a private setup just for them, or their own model keys.
4. Volume. Rough counts of requests, documents or conversations. You need this for Stage 4.

Output: a one-page fit summary that lists blockers plainly. A blocker you find now costs a call. One you find after the test costs the deal.

Stage 3: Design the test with the customer

This is the stage where AI selling differs most from other software selling. The buyer can't judge an AI product from a demo, because a demo shows the vendor's examples. They need to see it handle their own work.

Job posts now make running that test part of the account executive's job. Perplexity's Commercial AE runs the cycle "from initial presentation, technical evaluation, negotiation, to closing and onboarding," and Fireworks' AE works with internal teams "to shape tailored solutions, POCs, and fine-tuning approaches" (both opened September 26, 2026).

Anthropic's guide to defining success criteria gives a good test for each thing you'll measure: it should be specific, measurable, achievable and relevant. It lists common ones such as task fidelity, consistency, latency and price. Use those words with the buyer. They turn "does it work?" into a list both sides can check.

Write a test plan of one or two pages that says:

• Scope. One workflow, not five.
• Test set. Which real examples, how many, who picks them, and who approves sharing them.
• Criteria. Three to five, each with a pass mark agreed in advance.
• Judge. Who scores the outputs, and how disagreements get settled.
• Dates. Start, end and the decision meeting, with the economic buyer invited.
• Next step if it passes. The commercial terms and the first real use.

The last line matters most. A test with no agreed next step is free consulting. If the buyer won't commit to a decision meeting, the deal isn't qualified yet. For more on building test sets and scoring, see how to evaluate AI agent performance.

For Cedar Goods, the plan names the refund queue, a sample of real past tickets the billing agent picks, a pass mark for each criterion, and a decision meeting on a fixed date with the director's boss in the room.

Output: a test plan the buyer has approved in writing. Exit gate: that approval, plus a date for the decision.

Stage 4: Price on usage without surprising anyone

Many AI products charge by how much you use them. Model providers publish per-token prices, as the OpenAI API pricing and Claude pricing pages show. Some app vendors charge per result instead. Sierra described outcome-based pricing on December 10, 2024 as pricing tied to results "such as a resolved support conversation." As of September 26, 2026, Intercom's pricing page lists its Fin agent at "$0.99 per outcome."

Whatever the model, the buyer needs a cost forecast they believe. Build it from the test.

1. Record usage per task during the test: tokens, documents, conversations or resolutions, whichever unit you charge for.
2. Multiply by the volume from Stage 2, and show a low, expected and high case.
3. Write down every assumption, especially the share of tasks the product will handle without a human.
4. Offer whatever commitment options your company sells, such as prepaid credits or a yearly spend commitment, and explain the trade-off.

The buyer's finance team will ask what happens if usage doubles. Answer it before they ask. Agreeing on how the forecast is built is the exit gate, even if the final price is still being negotiated. The build-business-case skill shows how to model the buyer's value without making up ROI, and the prepare-negotiation skill covers the commercial conversation.

Stage 5: Security and legal review, started early

AI adds questions to a normal security review. Expect some version of these:

• Training on customer data. Buyers will ask whether their data trains the vendor's models. Know your company's written answer and where it's published. Anthropic's privacy center, for example, says "By default, we will not use your inputs or outputs from our commercial products to train our models" (Anthropic Privacy Center, checked September 26, 2026).
• Outside companies that touch the data. If your product calls another company's model, the buyer will want to know which one, where it runs, and how long data is kept.
• Security checklists. Security teams may send a questionnaire based on the Cloud Security Alliance's AI Controls Matrix, which CSA first released on July 10, 2025 and updated to version 1.1 in July 2026, alongside their usual SOC 2 and data processing requests.
• Wrong answers. Who is liable for a wrong answer, and what review steps the workflow includes.

Send the security package the same week the test plan is signed. Name someone on your side to own every open question. If a review won't finish before the target close date, move the date openly with your champion, rather than finding out on the last day of the quarter.

Alex sends Cedar Goods' security team the package the day the test plan is signed, not the day the test passes.

Output: security approval and agreed contract language. Exit gate: both.

Stage 6: Close on the date you agreed

If Stages 3 to 5 went well, the close is a meeting, not a negotiation marathon. Hold the decision meeting from the test plan. Show results against the agreed criteria, including the ones that didn't pass and what you plan to do about them. Then ask for the signature.

Use a mutual action plan, a shared list both sides can see, to keep the last steps on track: procurement, legal signatures, purchase order and the first production milestone, each with an owner and a date. The create-mutual-action-plan skill describes a version the buyer owns.

If a criterion failed, say so. A buyer who finds a hidden failure after signing is likely to leave and won't recommend you.

Stage 7: Hand off to customer success or the FDE team

On a usage-priced product, the deal isn't done at signature. Revenue arrives when the customer actually runs the work through it. Baseten's AE posting, opened September 26, 2026, describes the cycle as running "from first call to first production workload."

Write a handoff document that gives the delivery team:

• the workflow, the owner and the business reason from Stage 1
• the fit summary and blockers from Stage 2
• the test plan, results and every promise made during the deal
• the usage forecast and its assumptions from Stage 4
• open security or legal commitments from Stage 5
• the names of the champion, the economic buyer and the person who judges quality

If your company uses forward deployed engineers, the FDE engagement playbook shows what they need from the deal to plan the production work. The handoff-customer skill gives a checklist for the customer success version.

Stay on the account until the first production milestone. Your next deal with this customer depends on it.

Where to go next

• Read what the AE role owns in the AI account executive career guide, and practice with the AE interview guide.
• See the buyer's side in what is different about selling AI.
• Understand how deals turn into pay in the guide to OTE and sales pay plans.
• If you book meetings for AEs, the AI SDR outbound playbook covers the stage before discovery.