You have a discovery call in 30 minutes. You know the prospect’s name and company. What you don’t know: what’s actually going on in their business, what pressures their industry is facing, and what angle will open a real conversation instead of triggering polite deflection.
Manual research burns 45 minutes minimum, LinkedIn, news search, company website, industry overview, back to LinkedIn to check their recent posts. By the time you dial, you’re either under-prepared or you’ve spent half your morning on a call that might not convert.
The AI workflow below compresses that into 5 minutes without sacrificing quality. It doesn’t replace your expertise. It feeds you the raw material so you spend the prep time thinking, not searching.
Step 1: Perplexity for Current Intel (90 seconds)
Open Perplexity and paste this prompt, filling in the prospect’s name and company:
Prompt: “Research [Name] at [Company]. Tell me: their current role and how long they’ve been in it, any public activity from the last 90 days (posts, interviews, mentions), recent company news (funding, product launches, layoffs, expansions, leadership changes), and the top 2-3 challenges facing their industry right now. Cite sources.”
What you get back: a structured briefing with citations. Read the citations for anything you plan to reference on the call, Perplexity is highly accurate but not infallible, and a wrong detail destroys credibility faster than showing up unprepared.
What you’re looking for:
- A trigger event (company raised a round, launched a product, hired a new CMO, entered a new market)
- A pressure point (industry headwinds, regulatory change, competitive threat)
- Something the prospect has said publicly (a post they wrote, a quote in an article)
If you find one trigger event and one pressure point, you have enough context to open a real conversation instead of a generic one.
Step 2: Claude for Strategic Analysis (90 seconds)
Now paste the company’s homepage URL and core service pages into Claude with this prompt:
Prompt: “Review this company’s website: [paste URL and key text]. Based on what they sell, their apparent customer base, and their positioning, what business challenges are they most likely facing that a [your service] provider would be able to address? Give me 3 specific challenges with a one-paragraph rationale for each.”
Claude is better than Perplexity for this task because it synthesizes rather than searches. You’re not asking it for facts, you’re asking it to reason from what the company shows publicly to what they likely need privately.
The output gives you 3 framing hypotheses. These become your call structure. You’re not going in to pitch, you’re going in to test whether one of these three hypotheses is true.
Example output for a SaaS company that sells project management tools to agencies:
- They’re growing faster than their support and onboarding infrastructure can handle, creating churn risk with new clients.
- Their sales team is closing SMBs but struggling to convert enterprise accounts who need custom integrations.
- Marketing content is thin, they have a strong product but are invisible in search, making inbound leads scarce.
Any one of those is a legitimate entry point for a consultant who does onboarding systems, enterprise sales strategy, or content marketing.
The goal of pre-call research isn’t to impress the prospect with what you know about them. It’s to walk into the call with 3 hypotheses about their real problem and the right questions to find out which one is accurate.
Step 3: Generate Tailored Discovery Questions (90 seconds)
Back to Claude with this prompt:
Prompt: “Based on this company research and the 3 challenges identified, generate 3 discovery questions for a [your role] to ask in an initial call. Each question should: probe a specific pain point, be open-ended, and invite the prospect to tell a story rather than give a yes/no answer. Avoid generic questions like ‘What keeps you up at night?’”
What good AI-generated discovery questions look like versus bad ones:
Generic (useless): “What are your biggest challenges right now?”
Tailored (good): “You’ve expanded into enterprise accounts over the last year, what’s been the biggest friction point in making that transition stick?”
The difference is specificity. The tailored question signals you did your homework and invites a real answer. The generic question signals you didn’t.
Pick the 2-3 questions that feel most relevant, add one of your own based on your actual experience in the space, and you’re ready.
The Full 5-Minute Run Sheet
Here’s the sequence timed out:
- 0:00, Open Perplexity. Paste prospect research prompt. Wait for results (45 seconds). Read. Spot-check one citation if you plan to reference it.
- 1:30, Open Claude. Paste company website prompt. Wait (45 seconds). Read the 3 challenges.
- 3:00, Paste discovery question prompt in Claude. Wait (30 seconds). Review 3 questions.
- 4:00, Write down your 3 questions on paper or in a note. Add one of your own.
- 5:00, You’re ready.
This replaces a workflow that most solos spend 45-75 minutes on, often producing less actionable prep because the research time goes toward reading rather than thinking.
The Quality-Control Rule
Before every call, verify one category of information from the original source: whichever fact you’re most likely to reference out loud.
If you found a funding announcement, click through to the original press release. If you found a leadership change, confirm it on LinkedIn. If you found a product launch, check the company blog.
AI tools aggregate information and occasionally hallucinate dates, amounts, or names. The risk isn’t that you’ll sound uninformed, it’s that you’ll sound confidently wrong, which is worse than uninformed.
One 60-second source check per call eliminates this risk entirely.
Building the System Over Time
After 20 calls using this workflow, you’ll start recognizing patterns. Certain buyer types have predictable concerns. Certain industries have structural problems that recur across companies.
Log the trigger events and pressure points you find in a simple spreadsheet:
- Prospect name, company, date
- Trigger event found (or “none”)
- Hypothesis that proved accurate
- Hypothesis that was wrong
After 3 months, you’ll have a pattern library. Your AI prompts get sharper because you know what to look for. Your discovery questions get better because you’ve seen what opens people up.
The 5-minute workflow isn’t just a time-saver. It’s a compounding system.
Most freelancers prepare for sales calls by reading about the company. The better preparation is thinking about the company, forming hypotheses, not collecting facts. AI does the collecting so you can do the thinking.
What to Do When There’s No Research
Sometimes the prospect has zero online presence. LinkedIn is sparse, the company website is thin, no news in any search.
This is not a reason to skip prep. It’s a reason to change your prep prompt.
Prompt for low-info prospects: “I’m meeting with a [buyer title] at a [company type in their industry]. I have limited information about the specific company. What are the top 3 challenges typically facing [buyer title] roles in [industry] in 2026, and what questions would surface which of those challenges applies to this person?”
You’re now prepping with archetype knowledge instead of company-specific knowledge. It’s less precise but far better than winging it. And when you ask these archetype-informed questions on the call, the prospect will still feel heard, because the questions are relevant even if they’re not company-specific.
The Compound Effect on Win Rate
Pre-call research directly improves close rates through one mechanism: it shifts the conversation from selling to diagnosing.
When you walk in with 3 specific hypotheses about their problem, you’re not selling a service, you’re helping them understand their own situation. Buyers close on consultants who make them feel understood, not impressed.
The math is simple. If this workflow improves your close rate from 20% to 28%, a modest 8-point lift, and you run 4 discovery calls per week, that’s 1.6 additional clients per 10 calls. At even a $3,000 average project, that’s $4,800 per 10 calls from 5 minutes of additional preparation per call.
Spend the time.
Ready to send stronger proposals?
Build, send, and track proposals in one place so follow-up is easier.
Start your free trial →
