Cut Lead Review From Hours To Minutes
Sign up for a free trial of Attio, the agentic CRM.
Ask Attio to build a daily workflow that surfaces the deals that need your attention today, like anything with a stage change, a recent reply, or a new signal in the last 24 hours.
Review your pipeline in Claude, synced live from Attio via MCP.
That's it.

Let's begin
There’s a stretch of about 90 seconds after a call ends where I feel like I have a pretty good grip on the call… where I have a “feeling” about how the call went. The good stuff, the bad stuff, the raw stuff. Like when you notice that you just lost the person you’re talking to. Or the moment their eyes light up when you’re explaining something to them.
Three issues ago, I told you I have ADHD and that every deal I won was won by a system. That issue built the deal record out of what the buyer actually said, under one rule I still stand behind: my words are not evidence. Quote the buyer. Leave my opinion out of it.
The rule is still right, but this issue is going to cover what takes that system to another level… voice memos.
That issue also received more replies than any other issue I’ve ever written. One reader told me they felt “seen.” Another suggested we start a community of ADHD sales reps. One of the replies even came from a friend I had not talked to in months. I think it connected because it was not really about becoming more organized. It was about building a system around a brain that can be very good at one part of the job and completely unreliable at another.
Whether its my ADHD, or just my personal ability to “read” the person I am talking to... I know I’m not the only person with ADHD who experiences this as both a blessing and a curse. Some of us are extremely sensitive and responsive to nonverbal cues and even the smallest changes in emotion. I can be halfway through a sentence and notice somebody’s face change, their eyes move, their shoulders tighten, or the energy in the room drop, and I will start adjusting what I am saying before I have even consciously figured out what I noticed. In sales, that can feel like a superpower. But it also means I am taking in a ridiculous amount of information all at once, and once the call ends, the feeling stays while the small moments that created it start disappearing almost immediately.
The blessing is the part I’ve always loved. I can read the room. I can adjust in real time. I can tell when something I said landed, and I can usually feel when I need to stop talking and go in another direction. Those things have absolutely helped me sell.
But the curse is that almost none of it survives the call. (There are FAR more curses… but this issue isn’t about that… haha)
What survives is the conclusion. “That went well.” “He isnt the real decision maker.” “She liked the product but got nervous when I asked about budget.” I remember the feeling, but I lose the 6 or 7 small moments that created it. Then I put the conclusion into the CRM and move on to the next thing.
Thats a terrible trade.
It means the most valuable part of my ability to read somebody only exists while I am sitting in front of them. Once the meeting ends, the information starts disappearing, and the only thing left is my opinion about what happened.
I didnt need another system to help me pay attention. I needed one that could remember what I was already paying attention to.
The Feeling Is Not the Evidence
I want to make an important distinction here, because being highly responsive to nonverbal cues does not mean I am always right about what they mean. I’m often not correct, which is why this is so important.
If somebody’s face changes when I mention the price, the change is real. I saw it. But while I’m assuming they just had sticker shock… maybe they were surprised because the number was lower than they expected. Maybe a notification appeared on their screen at the exact same moment.
The observation and the conclusion both matter, but they are 100% not the same thing.
The observation is something I can go back and look at later. The conclusion is something I should be willing to question. When I only save the conclusion, I remove my ability to do that.
This is where the ADHD part gets especially difficult for me. My brain is very good at connecting things quickly. Sometimes that means I see the pattern before everybody else. Sometimes it means I connect 2 things that had absolutely nothing to do with each other.
In the moment, both can feel equally true.
The goal of this system is not to prove that my read was right. Its to use the CRM as a place for not only feelings, but also for observations. This information needs to exist after the feeling has passed and I am no longer sitting inside the moment.
Why Your Notes Lose the Best Part
Think about what most of us do after a call. We open the CRM and try to summarize 30 or 45 minutes of conversation into a few fields. We write down what the buyer wants, who is involved, the next step, and whether we think the deal is moving.
This sounds like the responsible thing to do. It is also the exact moment when the useful details get compressed out of the record.
“Champion is engaged” might have started as 5 different observations.

Those 5 observations became “champion is engaged,” and then the original observations disappear.
Now imagine reading that note 3 weeks later. If you’re like me, you have no idea why you believed she was engaged. Technically, you could go back and watch the recording, and so could your manager. But neither of you knows which 3-second moment to look for, where it happened, or why it stood out to you while you were on the call. Unless you capture the read while it is still fresh, that moment stays buried inside a 45-minute recording nobody has a reason to watch again.
And for those of us with ADHD, we are asked to do this at the worst possible time. Within a few minutes of the call ending, my ADHD brain just overloads and every other thought I’ve been holding back shows up all at once. We remember the email or Slack message that arrived during the call. We see the next meeting notification. We realize we have not eaten anything.
Then the CRM asks us to stop, organize the last hour, decide what mattered, and type it into boxes.
Of course the feeling is all that survives.
Step 1: Record the Mess Before Your Brain Moves On
The first step is the only part of this system that has to happen immediately.

As soon as the call ends, open a voice memo and talk for 60 to 90 seconds. Do it before you stand up, check Slack, open your inbox. Do it immediately. Make it a habit.
Do not try to recap the call. The recording is not meant to be a shorter version of the transcript. Focus only on the things you noticed and the feeling they left you with.
Start wherever your brain starts.
“I dont know why, but something changed when I asked about their procurement process.”
“She seemed excited, but the person next to her looked like he wanted the meeting to end.”
“The person I thought would push the most back barely spoke.”
“The call felt good until pricing came up.”
You are probably going to ramble. You might repeat yourself or change your mind halfway through a sentence. Good. The entire point is to capture this before you have cleaned it up enough to make it sound like a CRM note.
I use speech because typing makes me edit. If I type “he looked uncomfortable when I showed the implementation timeline,” I have already decided that the moment was important enough to keep. If I talk for 60 seconds, the things I have not decided about yet still have a chance to come out… and sometimes those are the most important things.
Remember, you actually have to think more about what you are typing… because you cannot type as fast as you talk. So when you’re talking, you are going to get the raw, unedited, unfiltered version… for better or worse.
Once you have the voice memo transcript, run this:
You are turning a voice memo recorded immediately after a sales call into a record of what I noticed.
Do not summarize the call. Do not improve my writing. Do not add anything I did not say.
Here is the voice memo transcript:
[PASTE VOICE MEMO TRANSCRIPT]
Return every distinct thing I noticed, one per line.
For each line, include:
1. What I said, using my words
2. Who I was talking about, if I named someone
3. When it happened in the call, if I mentioned it
4. How certain I sounded
Rules:
- Keep words such as "maybe," "I think," "something changed," and "I dont know why." They show my actual level of certainty.
- Do not combine several observations into one summary.
- If I contradicted myself, keep both statements and mark them as a pair.
- If I started a thought and did not finish it, keep the fragment and mark it [UNFINISHED].
- Do not complete a thought for me, even if the ending seems obvious.Those unfinished thoughts are more valuable than you might think. “I dont know why that bothered me” might be the closest you got to naming the thing that changed. I’d rather keep that sentence exactly as it is than let an AI turn it into a confident explanation I never gave.
Step 2: Separate What Happened From What You Think It Meant
Now we need to separate the read into 2 parts.
The first part is what you observed. A pause. A change in tone. Someone stopped taking notes. Someone who had been speaking in the first person started referring to “the team.” The buyer leaned forward when you showed one part of the product and started checking their phone during another.

The second part is what you concluded. They are worried about price. She is not a real champion. The CFO is going to kill the deal. They love this feature. Implementation is the reason they will not move forward.
Again, the conclusion might be right. I am not asking you to throw it away. I am asking you to stop storing it beside the observation as if they are the same type of information.
Run this next:
Here is my observation record from a sales call:
[PASTE THE OUTPUT FROM STEP 1]
Sort every line into exactly one of two buckets:
OBSERVED: Something that happened. This includes a statement, pause, change in tone, change in language, who spoke, who did not speak, or something that was
skipped.
CONCLUDED: An interpretation, judgment, prediction, or statement about someone's motive, belief, or intent.
Then, for every CONCLUDED line, list the numbered OBSERVED lines it rests on.
Rules:
- Words such as "seemed," "felt," and "clearly" usually indicate a conclusion.
- Any claim about what someone thinks, wants, believes, or intends is CONCLUDED.
- Tone can be observed. Motive cannot.
- If a conclusion rests on zero observations, flag it in caps as UNSUPPORTED.
- Do not search for an observation just to support a conclusion.
- Do not tell me whether a conclusion is right or wrong. Show me what it stands on.The UNSUPPORTED conclusions are usually the most interesting part of the output.
They’re not necessarily wrong, they’re just beliefs you’re carrying without anything from the call fully supporting them. Sometimes you’ve been carrying them for so long that you stopped realizing they were assumptions.
“The champion is handling this internally.”
“We have executive support.”
“Legal is the only thing holding this up.”
Those sentences can survive for weeks, simply because they were entered into the CRM. Repetition makes them feel more certain, even when no new evidence has been added.
Step 3: Give the Read Somewhere to Live
In S01E06, I published a Deal Risk Scanner that scores buyer confidence, internal alignment, process clarity, and decision ownership from 1 to 5.
I still stand behind the scanner. But the scanner can only work with what you give it.
If you feed the CRM a note that says “champion is engaged,” it has to treat that sentence as evidence. It cant see the observations your brain used to reach the conclusion because none of those observations made it into the note.
This is the part I had backward. I thought the framework would catch what I missed. But I did not miss the signal. I caught it, formed an opinion about it, and then threw away the signal.
So run the scorer using only the OBSERVED section:
Score this deal from 1 to 5 across four dimensions, where 1 represents high risk:
- Buyer confidence
- Internal alignment
- Process clarity
- Decision ownership
Use ONLY the observation lines below. Nothing else is admissible.
[PASTE THE OBSERVED BLOCK FROM STEP 2]
For each dimension, return:
1. The score
2. The observation supporting that score, quoted exactly
3. One question I could ask on the next call that would help me score the dimension with stronger evidence
Rules:
- If a dimension has no supporting observation, do not return a 3 and do not
guess. Return NOT OBSERVED.
- For every NOT OBSERVED result, state what I would have needed to notice to score it.
- Do not use conclusions. They are missing from this input on purpose.
- Do not average the four scores.
- Do not produce an overall deal-health score.Then run the same scorer on the notes you would normally have written and put the 2 outputs next to each other.
I dont know what your result will be. Maybe the scores are almost identical, which means your notes are already keeping the right things. But I think a lot of us are going to find dimensions that looked healthy in the CRM and come back as NOT OBSERVED once our opinions are removed.

That does not automatically mean the deal is bad. It means you have been scoring something you did not actually capture.
And that is fixable.
Why this works for my brain (and maybe yours too)
This works because it was designed around the brain I actually have, not the perfectly organized version of me who always updates the CRM immediately after a call.
It doesnt ask me to slow down and organize everything during the 90 seconds when my attention is already trying to move somewhere else. It lets me talk. I can be messy, uncertain, and halfway through a thought. The organization happens after the information is safely outside my head.
It also preserves the useful part of being highly responsive to the person in front of me without pretending that every feeling I have is correct. I get to keep the change in tone, the shift in body language, and the moment the energy left the room. Then I can look at the conclusion separately and decide whether the observation actually supports it.
Most importantly, it means the ability does not disappear when the call ends.
For a long time, I treated my ability to read people as something that only helped me in the room. I could adjust while the conversation was happening, but I could not use the same information to prepare for the next call, explain the deal to somebody else, or see that I had noticed the same thing 3 meetings in a row.
The feeling existed. The record did not.
Now it does.
There are still parts I have not solved. Some voice memos turn into a recap of the call instead of a record of what I noticed. I’ve improved this by tightening the prompt, but it still happens. When it happens, I go back and delete those parts myself.
I also cant tell you that every read you capture will be correct. It wont be. Being sensitive to a change does not make you psychic, and a facial expression is not proof of what somebody is thinking.
But now the read is visible. You can inspect it. You can compare it with what happens next. You can find out whether the things you notice are actually signals or whether some of them are just stories your brain is very good at telling quickly.
I’d rather have a read I can question than a feeling I can’t explain.
Rep Action this week
Pick 3 calls already on your calendar. Not your entire week. Just 3.
When each call ends, record a 60 to 90-second voice memo before you do anything else. Talk about what you noticed, what changed, and what gave you the feeling you had about the call.
Run steps 1 and 2 on all 3 memos. Then choose the call you feel best about and run the Deal Risk Scanner twice: once using the notes you would normally have written, and once using only the OBSERVED section.
Put the 2 scorecards next to each other. By Friday, you will be able to see the distance between what you noticed and what you actually kept.
Then tell me something: what do you notice on calls that you’ve never once written into your notes?
Not what a sales trainer told you to watch for, rather the thing you pick up on every time but have never had anywhere to put. One line is a complete answer, and I have a feeling the people with ADHD reading this are going to give me some of the best ones. I might even include a few replies in next weeks issue.
~ Jay
🛑🤚Wait! There’s so much more!!

The 3 steps above will show you how much of your read is currently disappearing between the call and the CRM. Doing it after every call, comparing those observations across an entire deal, and finding the assumptions already hiding in your pipeline are different problems.
This week in the Vault:
The 90 Second Close-Out. Paste one voice memo and get the observation record, the OBSERVED and CONCLUDED split, and the deal score in one pass.
The Pattern Check. Compares what you noticed across every call on one deal and finds the changes or repeated signals your brain experienced separately.
The Unsupported Sweep. Finds the conclusions sitting inside your current forecast that do not have a single recorded observation underneath them.
Members get every Vault drop plus the full back library. $15/mo, or $100/yr and save $80.
Click here for the stack I’d build today: https://www.sellingwithai.vip/stack
AI can build faster. Can your team decide better?
AI can draft the PRD and prototype the idea. Jira Product Discovery helps teams decide whether it belongs on the roadmap. Bring feedback and ideas together, prioritize as a team, and keep your roadmap connected to delivery in Jira.
You just read the motion. Now run it.
The prompts, checklists, and templates that turn this into a 10-minute execution are in the Vault.
Get Vault AccessVault access includes::
- Copy and paste execution prompt packs
- Deal, outbound, and follow-up playbooks
- Operating checklists for every motion
- Members Vault access



