- Why Claude specifically, for this kind of work
- The proposal prompt that saves time
- Turning messy call notes into a usable discovery summary
- Building a framework or diagnostic without starting from zero
- The uncomfortable bit nobody tells consultants
- Client reporting without the Sunday night dread
- Handling the awkward client emails
- A quick worked example, start to finish
- Where this fits with social content too
- Claude versus other tools for this kind of work
- When it's worth getting outside help instead
- Frequently asked questions
The short version: Claude is better than ChatGPT at long, structured consulting work because it holds more context and doesn't flatten your voice into generic LinkedIn mush. The prompts below are built for the actual grind of consulting: proposals, discovery notes, frameworks, client reports and the awkward emails nobody wants to write. Copy them, change the brackets, and stop staring at a blank page.
Why Claude specifically, for this kind of work
I use Claude for anything that needs to hold a lot of information at once and still sound like a person wrote it. Consulting work is exactly that: a client brief, three stakeholder calls' worth of notes, a pricing structure, and a deadline. Ask Claude to pull all of that into one coherent document and it generally does, without the weird repetition ChatGPT sometimes slips into on longer outputs.
That said, Claude is not magic. It doesn't know your client's politics, it doesn't know which stakeholder is secretly the blocker, and it will happily write a confident-sounding recommendation based on nothing if you let it. The prompts below are built to stop that happening by forcing you to feed in real detail before it writes a single word.
The proposal prompt that saves time
Most consultants write proposals from scratch every time, which is a waste of an afternoon. Here's a working example: say you run a small marketing consultancy and a mid-sized e-commerce brand has asked you to pitch on a six-month growth retainer. You'd give Claude something like this:
- "Act as a senior consulting proposal writer. I'm pitching a six-month marketing retainer to [company], an e-commerce brand selling [product category] with roughly £2m annual revenue. Their stated problem is [stagnant organic traffic and low repeat purchase rate]. My scope will cover [SEO, email lifecycle, paid social]. My monthly fee is £4,500. Write a proposal with these sections: situation summary, proposed approach, 90-day milestones, deliverables, pricing, and next steps. Keep the tone direct and confident, no jargon, no filler sentences."
What comes back won't be perfect, but it'll be 80% there in under a minute, and you edit the 20% that needs your actual judgement. That's the real time saving: not zero effort, but skipping the structural scaffolding every single time.
Turning messy call notes into a usable discovery summary
This is the one that pays for itself fastest. After a discovery call, most consultants have a page of scrappy notes and a vague memory of what mattered. Paste the raw notes into Claude with this prompt:
- "Here are my raw notes from a discovery call with a potential client [paste notes]. Turn this into a structured discovery summary with: business context, stated problems, unstated problems you can infer from what they said, budget signals, decision-making process, and three questions I should ask in the follow-up call. Don't soften anything, flag it plainly if something sounds like a red flag."
If you're summarising a lot of calls, this pairs well with the structure in my piece on meeting summary prompts, which works the same way whichever model you're running it through.
Building a framework or diagnostic without starting from zero
Consultants sell frameworks, and clients pay partly for the feeling of a structured method, not just advice. A worked example: say you advise small accountancy firms on digitising their client onboarding. You could prompt Claude with:
- "I need a five-stage diagnostic framework for assessing how ready a small accountancy firm (5 to 20 staff) is to digitise client onboarding. Name each stage, give a one-line description, and list three diagnostic questions per stage I could ask in an audit call. Make it something I could put my own name on, not generic consulting language."
Then you stress-test it. Ask Claude to argue against its own framework: "Now tell me the three biggest weaknesses in this framework and where it would fail with a resistant client." This step matters more than people admit. A framework that's never been attacked is just a nice-looking slide, and clients can smell that.
The uncomfortable bit nobody tells consultants
Here's what most advice on this topic skips over: if your entire deliverable is a document Claude could produce in five prompts, your client will eventually work that out, and your pricing becomes very hard to defend. The AI isn't the problem, pretending it isn't involved is. The consultants doing well with this right now are the ones being upfront that AI speeds up drafting, and pricing themselves on judgement, relationships and accountability instead of on hours spent typing. If your invoice only reflects typing time, AI is a genuine threat to your business model. If it reflects the decisions you're making on the client's behalf, it's a tool that makes you faster, not cheaper to replace.
Client reporting without the Sunday night dread
Monthly reports are the task consultants put off longest because they're repetitive but still need to be good. Build a reusable prompt like this one:
- "I'm writing a monthly report for a client on [area, e.g. paid social performance]. Here's this month's data: [paste numbers]. Here's last month's: [paste numbers]. Write a report with: headline summary (two sentences max), what worked, what didn't, one thing I'm changing next month, and a plain-English explanation of why the numbers moved. No corporate filler, write it like you're talking to a smart business owner who doesn't have time for jargon."
Save this as a template and you can run it in roughly five minutes a client a month, instead of the forty-five minutes most consultants spend wrestling a spreadsheet into prose.
Handling the awkward client emails
Scope creep emails, late payment chasers, "we're pausing the project" responses, these take disproportionate mental energy for their length. One prompt structure covers most of them:
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
- "Write a short email to a client who has asked for [extra deliverable not in scope]. I want to say yes but only if we adjust the fee or timeline. Tone: friendly but firm, no apologising for having boundaries, three sentences max before the ask."
I'd avoid letting Claude send anything that touches money or contracts without reading it twice yourself. It writes a confident email even when the underlying logic is shaky, so you're still the one responsible for what goes out under your name.
A quick worked example, start to finish
Say you're a solo operations consultant and a logistics company wants a one-page proposal by tomorrow morning. Realistic sequence:
- Step 1: Paste your notes from the intro call into Claude and ask for a discovery summary (roughly 2 minutes).
- Step 2: Use that summary as context for the proposal prompt above, feeding in your actual day rate (roughly 3 minutes).
- Step 3: Ask Claude to cut the draft by 30% and remove any sentence that doesn't contain a fact or a number (roughly 2 minutes).
- Step 4: Read it aloud once, fix anything that doesn't sound like you, send it.
That's a proposal built in under fifteen minutes of active effort instead of an evening, and the quality holds up because you controlled the inputs at every stage.
Where this fits with social content too
Consultants who are also building a personal brand often need to turn the same insight into a LinkedIn post, a client email and a workshop slide, three different formats from one idea. The approach in my post on Claude prompts for social captions works well for the LinkedIn version, and if you're planning content around specific dates in the year, the structure in my holiday content prompts for Claude post adapts easily to a consulting audience rather than a retail one.
Claude versus other tools for this kind of work
People ask me constantly whether Claude or ChatGPT is better for consulting output. My honest take: Claude tends to win on anything long-form and structured, ChatGPT still edges it on quick back-and-forth brainstorming. If you're weighing Claude against other models specifically, I've broken that down in Grok vs Claude, which covers where each one holds up under real work pressure rather than demo conditions.
When it's worth getting outside help instead
If you're a consultant drowning in admin and prompting your way through it feels like a part-time job in itself, that's usually a sign you need a system built once rather than more prompts. That's a different conversation to this post, but it's worth having, and it's exactly what I cover with clients through an AI implementation coach setup rather than another list of prompts to copy and paste.
Frequently asked questions
Is Claude better than ChatGPT for consultants?
For long documents like proposals, reports and frameworks, Claude tends to hold context better and drift into generic phrasing less often. For fast idea generation and quick chat-style back and forth, ChatGPT still holds its own. Most consultants I'd recommend using both rather than picking one forever.
Can Claude write an entire consulting proposal for me?
It can get you 70 to 80% of the way there if you feed it real details about the client, your scope and your pricing, but you still need to add your own judgement on anything strategic, and you should never send a proposal without reading every line yourself.
Will clients notice if I use AI for reports and proposals?
Some will, and increasingly they expect it. The thing that matters is pricing yourself on the thinking and accountability behind the document, not the typing, so that AI speeds up your delivery without undermining your fee.
What's the single most useful Claude prompt for a busy consultant?
The discovery call summary prompt. Turning messy call notes into a structured summary with stated problems, unstated problems and follow-up questions saves more time weekly than any other single prompt on this list.
Run a business, a brand or a blog? I feature tools, products and expert guest posts for an audience of 15,000. See how it works