This week: a 5-prompt deal autopsy that tells you exactly why you lost (and how to stop it), Anthropic ships Claude for Small Business with 15 pre-built sales workflows, and an invisible AI meeting assistant that costs less than your morning coffee.
THE PLAY
The Deal Autopsy: 5 Prompts That Tell You Exactly Why You Lost
Every sales team has a graveyard of lost deals. Most of them get a one-line note in the CRM ("went with competitor," "budget cut," "timing") and never get looked at again. That is a goldmine of intelligence you are ignoring.
According to HubSpot's 2026 State of Sales survey, the average B2B sales team wins just 21% of its deals. Nearly 4 out of 5 opportunities end in closed-lost. Most teams never debrief more than a handful of those losses. The data is sitting in your CRM, your call recordings, and your email threads. You just need to pull it out and read it.
Here is a 5-prompt system that does a full autopsy on any lost deal in under 10 minutes. Run it in ChatGPT or Claude. You will walk away knowing exactly what broke and what to do differently next time.
We are going to show you the first two prompts below. They are the data gathering steps that build your deal profile. The final three prompts are the ones that actually diagnose what went wrong and give you the fix. Those are on our prompt library page because they are longer and include example outputs you will want to reference.
Prompt 1: The Debrief
This prompt builds a clean, objective timeline of the deal so the AI has full context before it starts analyzing.
I lost a B2B deal and I want to understand what went wrong. Here are the details:
Company: [prospect company name]
Deal size: [dollar amount]
Sales cycle length: [weeks/months from first contact to loss]
Decision makers involved: [titles and roles]
Competitor we lost to: [name, or "unknown"]
Our proposed solution: [one sentence]
Key objections raised: [list the main pushback you heard]
How the deal ended: [went dark, chose competitor, budget cut, etc.]
Build me a clean deal timeline with:
1. Every major milestone (first call, demo, proposal, etc.)
2. Where momentum shifted (positive or negative)
3. The last meaningful interaction before we lost
4. One sentence summary of the deal arc
Keep it factual. No spin. I need to see this deal as it actually happened.Why it works: You are forcing the AI to organize messy deal data into a structured narrative. Most reps cannot do this objectively because they are too close to the deal. The AI has no ego. It just maps what happened.
Prompt 2: The Pattern Finder
Run this after you have done Prompt 1 for at least 3 lost deals. This is where it gets powerful.
Here are the deal timelines for my last [3-5] lost deals:
[Paste the timelines from Prompt 1 for each deal]
Analyze these deals together and tell me:
1. What patterns appear across multiple losses? (timing, objections, stakeholders, deal stage)
2. At what stage do I consistently lose momentum?
3. Which objections keep showing up that I am not handling well?
4. Is there a common competitor or alternative winning these deals?
5. What is the single biggest factor in my losses?
Be specific. Use the actual data from my deals. Do not give me generic sales advice.Why it works: One lost deal is an anecdote. Three lost deals analyzed together is a pattern. This prompt forces the AI to cross-reference your losses and surface the systemic issues you cannot see when you are looking at each deal in isolation.
The next three prompts are where the real value lives. Prompt 3 ("The Gap Report") identifies exactly where you lost ground to the competitor. Prompt 4 ("The Fix") generates specific talk tracks and rebuttals for the objections that keep killing your deals. Prompt 5 ("The Never Again Checklist") builds a pre-deal qualification checklist based on your actual loss patterns so you stop repeating the same mistakes.
IN PARTNERSHIP WITH
Your next great hire lives in Slack.
Viktor is an AI coworker that connects to your tools and ships real work. Ask Viktor to pull a report, build a client dashboard, or source 200 leads matching your ICP. Most teams hand over half their ops within a week.
THE SIGNAL
1. Anthropic launches Claude for Small Business: 15 pre-built agentic workflows for sales, finance, and ops.
On May 13, Anthropic launched Claude for Small Business, a toggle install that connects Claude to the tools small businesses already use: HubSpot, QuickBooks, PayPal, Canva, DocuSign, Google Workspace, and Microsoft 365. It ships with 15 ready to run agentic workflows. The sales relevant ones: lead triaging (auto-qualify inbound and route to the right rep), pipeline pulse (surface your most important pipeline movement on a schedule), and campaign launcher (analyze HubSpot performance, draft promo strategy, generate Canva assets).
🎯 Sales Impact: If your team runs HubSpot and you have been doing pipeline reviews manually, this changes your Monday morning. Claude can now pull your pipeline data, flag stalled deals, and draft the recap for you. The catch: Claude doesn’t act autonomously. You initiate every workflow and approve before anything sends. That is actually a feature, not a bug. You stay in control while the AI does the prep work.
2. OpenAI hits $25B ARR. GPT-5.5 Instant is now the default model for all users.
OpenAI is now generating $25 billion in annual recurring revenue, with enterprise customers driving the bulk of growth. Over 1,000 enterprise customers pay $1M+ per year. Meanwhile, GPT-5.5 Instant quietly became the default model, replacing GPT-5.3. The key upgrade: responses are up to 30% shorter and the model now runs parallel internal checks, essentially double checking its own reasoning before showing you the output.
🎯 Sales Impact: Shorter, more accurate responses mean your AI assisted call prep, email drafts, and deal analysis just got better without you changing anything. If you’ve been getting long, rambling outputs from ChatGPT, that should improve. The self correction feature matters for anything where factual accuracy is critical, like competitive research or pricing analysis.
THE STACK SWAP
❌ BEFORE: You’re on a discovery call with a VP of Sales. She mentions a budget constraint, a competitor evaluation, and a timeline shift, all in the same sentence. You are trying to take notes while also asking good follow up questions. You catch maybe 60% of what she said. After the call, you spend 15 minutes reconstructing your notes from memory, miss two of the four action items, and your CRM update is a vague three liner that helps nobody. When your manager asks what happened on the call next week, you cannot remember the details that matter.
✅ AFTER: The call happens exactly the same way, except Granola is running quietly on your laptop. No bot joins the call. No recording notification pops up. Nobody knows it is there. It captures the audio directly from your device and builds a real time transcript. Two minutes after hanging up, you have clean, structured meeting notes in a Notion style layout with the four action items pulled out, key decisions highlighted, and a summary you can push to HubSpot or Salesforce with one click. You did not take a single note during the call. You were fully present, asking better questions, catching the signals that close deals.
The tool: Granola (free tier for individuals, Business plan $14/user/month). Granola is an AI meeting notes app that runs locally on your Mac or PC. It captures audio directly from your device, not through a bot that joins your call. That means it works on Zoom, Google Meet, Teams, Webex, Slack Huddles, even browser based calls. After every meeting, it generates structured notes with action items, key decisions, and a clean summary. Push notes to HubSpot, Salesforce, Notion, Slack, or Attio with one click. Granola also supports custom note templates, so you can set up a MEDDPICC template or a discovery call format and every set of notes comes out in your preferred structure. It just raised $125M at a $1.5B valuation and counts Brex, Notion, and Y Combinator companies among its users. The free tier gives you AI meeting notes with limited history. Business ($14/month) unlocks unlimited notes, advanced AI features, CRM integrations, and an MCP integration that connects your meeting context to other AI tools. If your reps are spending 15 minutes after every call
QUICK WINS
1. ❌ "Let me check on that and get back to you." ✅ "I'll have the answer in your inbox by 3pm today with the documentation." Specificity signals competence. Vague follow ups signal you do not have the answer and might not find it.
2. Before your next discovery call, paste the prospect's LinkedIn About section into ChatGPT and ask: "What is this person's likely biggest frustration in their current role, and what question could I ask that would make them feel understood?" You will walk in sharper than 95% of reps they talked to this week.
3. ❌ Sending a 500 word recap email after a demo. ✅ Sending a 3 bullet recap with one question. Keep post-demo emails under 100 words. Lead with the single biggest thing they cared about. End with a question that advances the deal ("Would Thursday work to loop in [stakeholder name]?").
4. This week's prompt from the library: Paste your last cold email into ChatGPT with this prompt: "Rate this cold email 1-10 on specificity, curiosity, and brevity. Rewrite it to score 9+ on all three." See 200+ prompts like this →
THE CLOSER
Know a sales leader who needs this? Forward this email to one person who would benefit. That is how we grow.
Next week: The Competitor Kill File. A step by step system for building real time competitive intelligence using AI, so you never walk into a deal blind again.
What did you think of today's newsletter?
FREE RESOURCES
To start getting immediate value from AI I've built the library (and will continue to build more) for top AI sales tools, prompts, tricks, and tips.
