The Recap Email That Keeps a Deal Alive Between Discovery and the Demo

11 min read

Most recap emails summarize what you said — flip it to what the buyer said and it becomes something they actually forward.

The gap where deals go to die

You had a great discovery call. The buyer was engaged. They told you real things about real problems. You hung up feeling good.

Then nothing happened for nine days.

By the time your demo rolled around, the buyer had a vague memory of the conversation, their boss had never heard of you, and you spent the first fifteen minutes of a thirty-minute demo re-discovering what you already knew.

I have watched this happen hundreds of times. When I was selling inside AWS and Dell, I watched it happen to me. The gap between discovery and demo is not a scheduling problem. It is an information problem. The buyer goes back to their desk, gets pulled into five other things, and the urgency you built on the phone evaporates because there is nothing physical holding it in place.

The recap email is the only artifact you control in that gap. And most reps write one that actively makes things worse.

Why your current recap gets ignored

Pull up the last recap you sent. I will bet it reads like a summary of what you said on the call. "As I mentioned, our platform does X. I shared that we work with companies like Y. I explained how Z works."

Nobody forwards that. Why would they? The buyer did not get on the phone to hear about your platform. They got on the phone because they have a problem. If your recap is about you, it is useless the moment it lands.

The recap email has exactly one job: give your champion something they can forward to the person who was not on the call. That person — the VP, the CFO, the operations lead — is the one who decides whether the demo is worth attending. If they never see the recap, they show up cold or they do not show up at all.

So you need to write the recap as an internal document that happens to come from an external sender. Their words. Their priorities. Their numbers. In the order they said them.

The structure that works

Every recap I send follows the same five sections. No headers in the email itself — just natural paragraphs that flow in a sequence the buyer already recognizes because it mirrors what they told me.

Section 1: What they said the situation is

Open with a two-sentence summary of their current state, using phrases they actually used on the call. If the buyer said "we are drowning in manual RFPs," you write "drowning in manual RFPs." Not "experiencing inefficiencies in the RFP process." Their language is the signal to the person reading the forward that this is not a vendor pitch. It is a reflection of something their colleague actually said out loud.

Section 2: The three problems, with numbers

Most discovery calls surface two to four problems. Pick the three that had numbers attached. If the buyer said their team spends four hours per load on manual carrier matching, that is a number. If they said they lost two nurses last quarter because of scheduling friction, that is a number. If they said their close rate dropped since they changed CRMs, ask for the number before you hang up — or flag it as an open question in the recap.

List these as three short paragraphs, each one anchored to a specific metric the buyer gave you. Not your ROI calculator. Their words, their math.

This is why good discovery matters so much. If you did not pull real numbers on the call, you have nothing to put here, and the recap falls apart. If you are running discovery in healthcare, the 25-minute playbook for CMIO and clinical ops calls walks through exactly which questions surface the numbers you need for this section. For freight and brokerage, the discovery playbook for brokerage buyers does the same thing.

Section 3: What you agreed to show and to whom

This is the paragraph that creates accountability. "On the demo Thursday at 2 PM ET, I will walk through how we handle [specific thing they asked about] and [second specific thing]. You mentioned Sarah should see the reporting piece — I have included her on the calendar invite."

Notice what this does. It names the demo. It names the time. It names what will be covered. And it names a second person. More on that last part in a moment.

Section 4: Open questions you still need answered

Every discovery call leaves loose ends. Maybe you asked about their contract renewal timeline and they said they would check. Maybe you need to know which EHR they are running or how many loads they move per month. List two or three of these as direct questions.

This gives the buyer a reason to reply. A recap with no open questions is a dead end. A recap with two specific questions is a conversation.

Section 5: The date, restated cleanly

End with one line: the date, time, timezone, and a link to the calendar invite. Nothing else. No "looking forward to it." No "please let me know if you have any questions." Just the logistics.

Worked example 1: Clinical ops discovery

Here is a full recap email I would send after a discovery call with a Director of Clinical Operations at a mid-size health system.


Subject: Recap — staffing visibility + overtime costs

Hi Dr. Patel,

Thanks for the time today. I want to make sure I captured everything accurately so this is easy to share with Lauren before Thursday.

You mentioned your team is managing shift assignments across three facilities using a combination of the Kronos schedule and a shared spreadsheet that your charge nurses update manually. The core issue is that when someone calls out, the backfill process takes 45 minutes to an hour because there is no single view of who is available, credentialed, and not already in overtime.

Three things stood out:

  1. You are running overtime costs north of $1.2M per year across the three facilities, and most of it is avoidable — driven by last-minute fills going to whoever picks up the phone first rather than whoever is the most cost-effective option.

  2. You lost two experienced nurses in Q3 who cited scheduling unpredictability in their exit interviews. You said replacing an experienced RN takes about four months and costs roughly $56K when you factor in agency coverage during the gap.

  3. Your CNO has asked for a dashboard that shows real-time census-to-staff ratios by unit, and right now that data lives in three systems that do not talk to each other. You are producing that report manually every Monday and it is already stale by Tuesday.

On Thursday at 2 PM ET, I will walk through how we handle real-time shift optimization across multiple facilities and show the census-to-staff dashboard Lauren asked about. I have added her to the invite — if there is anyone else on her team who owns the overtime reporting, it might be worth having them see the reporting piece as well.

Two things I still need from you: the name of the EHR you are running at the Riverside facility (you mentioned it might be different from the other two), and whether your Kronos instance is cloud or on-prem.

Thursday, June 12, 2:00 PM ET — [calendar link]

Talk soon, [Name]


Let me walk through what is happening here.

The subject line uses the buyer's words, not mine. "Staffing visibility" and "overtime costs" are phrases Dr. Patel said on the call. When Lauren sees this forwarded, the subject line tells her this is about a problem she already knows about.

The three problems all carry numbers Dr. Patel gave me. I did not calculate them. I did not round them. I wrote what she said. When Lauren reads this, she sees her own organization's data reflected back, not a vendor's estimate.

The line about "anyone else on her team who owns the overtime reporting" is the multi-threading line. I am not asking Dr. Patel to add a third attendee. I am suggesting that Lauren might want to bring someone. That is a much easier yes because it frames the additional attendee as Lauren's decision, not mine. Most of the time, this results in one or two more names on the invite without me ever having to ask directly.

Worked example 2: Brokerage discovery

Now the same structure after a discovery call with a VP of Operations at a freight brokerage.


Subject: Recap — carrier matching time + margin compression on spot loads

Hi Marcus,

Appreciate the conversation today. Wanted to get this down while it is fresh so you can share it with Deb ahead of Tuesday.

You are running about 800 loads a week with a team of 14 brokers, and the biggest bottleneck is carrier matching on spot freight. Your brokers are spending roughly 35 to 40 minutes per load checking availability across your TMS, DAT, and a handful of carrier relationships they manage in their heads. On contract freight that process is faster, but spot loads — which you said make up about 40 percent of your volume — eat most of the team's day.

Three problems that came up:

  1. Margin on spot loads has compressed from about 18 percent to 11 percent over the past year, and you think a big piece of that is speed — by the time your brokers find a carrier, the rate has already moved. Faster matching would let you lock rates closer to the initial quote.

  2. You are losing carriers to brokerages that confirm loads faster. You mentioned two specific carriers in your top 20 who told your team they are moving volume to competitors because response time is too slow. That is capacity you cannot replace easily.

  3. Your ops team spends Friday afternoons building a manual report on load-to-broker ratio and margin by lane, and Deb has said she wants that in real time. Right now it is a four-hour process pulling from three different exports.

On Tuesday at 10 AM CT, I will show how we handle automated carrier matching on spot loads and walk through the margin-by-lane reporting Deb wants to see. I have added her to the invite. If she has someone on the finance side who touches margin analysis, it might make sense for them to see the reporting walkthrough too.

Two open questions: you mentioned you are evaluating whether to renew your current TMS contract in Q1 — can you confirm the exact renewal date? And I would like to know whether your carrier onboarding data lives in the TMS or in a separate system.

Tuesday, June 17, 10:00 AM CT — [calendar link]

Talk soon, [Name]


Same structure. Same principles. The buyer's words, the buyer's numbers, the buyer's priorities. The freight discovery playbook is built to surface exactly the kind of specifics — loads per week, margin by load type, time per task — that make this recap possible. If you skip those questions on the call, you end up writing a generic summary that no one forwards.

The line that multi-threads without asking

I want to zoom in on one sentence because it is doing more work than it looks like.

"If there is anyone else on her team who owns [specific function], it might be worth having them see [specific section of the demo] as well."

This works for three reasons. First, it is framed as a suggestion about what is useful for the buyer's colleague, not a request for more access. Second, it ties the additional attendee to a specific part of the demo, which makes the ask concrete instead of vague. Third, it puts the decision in the champion's hands. You are not asking to be introduced. You are making it easy for your champion to say "hey, you should probably sit in on this."

I have used this line in every vertical I have sold into. It works in SaaS discovery conversations where you need RevOps and the CRO in the same room. It works in healthcare when you need the CMIO and the Director of Nursing. It works in freight when you need Ops and Finance. The mechanic is always the same: name the person, name the section, make it their idea.

The mistakes that kill a recap

Three things I see reps do that guarantee the email gets archived and never forwarded.

Writing about yourself. If more than one sentence in the recap is about your product, your company, or your experience, you have written a pitch, not a recap. Strip it out. The buyer knows who you are. The person they forward it to does not care who you are — they care about the problem.

Rounding or guessing at numbers. If the buyer said $1.2M in overtime costs, write $1.2M. Do not write "over a million" because it sounds more conservative. And absolutely do not invent a number because the buyer did not give you one. If you do not have a number, describe the problem in words and flag it as an open question.

Skipping the open questions. A recap with no questions is a monologue. It does not invite a reply. It does not move the deal forward. Always leave two or three threads hanging that require the buyer to come back to you. That reply is your proof of life between now and the demo.

What I would do with this

If you are sending recaps after every discovery call and still losing deals in the gap, the problem is almost certainly in the structure — you are summarizing instead of reflecting. I built DrillCall to help reps practice the actual calls that generate the material you need for a recap like this, because a good recap is impossible without a good discovery call. If your discovery is pulling real numbers and real language from the buyer, the recap practically writes itself. If it is not, no template will save you.

Send the next one using the structure above. Watch whether the reply comes back with a new name cc'd. That is how you know it is working.

Practise these calls

The playbooks behind this post — a scripted opener, the objections you will actually hear, and an AI buyer to run it against.

About the author

Timothy Yang

Founder & CEO, DrillCall

I build products by getting on the phone. Four businesses built and exited, including a micro-task marketplace with 170,000+ users, and the common thread in every one was the same: nothing moved until I picked up the phone and sold. Cold outreach, discovery calls, closing. The unglamorous work that actually creates revenue. Right now I am building DrillCall, an AI-powered voice training platform where sales reps practice live calls against realistic AI buyer personas, 310 of them across 31 industries, and get a scorecard after every call. Think flight simulator, but for cold calls. I also run Vibe Coding Club, a community of over 3,500 builders shipping products with AI, and I have spent time inside AWS and Dell, so I have seen how enterprise sales machines work from the inside as well as from the founder seat. What I care about: expected value thinking, fast iteration, and talking to customers before writing a line of code.

← All posts