For Business Owners · Module 04/08

Claude, in Practice

When and why to reach for Claude, an honest, specific read for business owners.

There's a tendency in "AI tools" content to rank the models like a consumer review: Claude is 9.1, ChatGPT is 8.9, Gemini is 8.6. That framing is useless for operators.

The right question isn't which model is better. It's which model is better for this specific task.

This module is about Claude specifically: when it earns its spot, where it has real limits, and how a small business owner can put it to practical use starting today.

Full transparency: innmotion builds its infrastructure on Claude. This is not a neutral review. But the practical guidance below reflects honest experience, including the weaknesses, because a course that only tells you what Claude does well isn't teaching you anything useful.

---

What Claude Is Good At

Thinking through a problem with you. This is where Claude genuinely outperforms the other models in consistent practical experience. If you have a complex situation, a difficult client email, a pricing decision, a business problem you're not sure how to frame, Claude will reason through it with you in a way that often surfaces angles you hadn't considered.

The character of Claude's reasoning is more like a thoughtful advisor than an answer-dispenser. It'll point out what you might be missing, name the tension in the situation, and ask clarifying questions if you give it space to. For problems where the answer isn't obvious, this is extremely valuable.

Long document analysis. If you need to understand a contract, a supplier proposal, a set of council regulations, or a lease agreement, Claude handles lengthy documents better than most models. You can paste in 20 to 30 pages of text, ask it to summarise the key terms, flag any clauses you should be concerned about, or compare two versions. It'll do this without losing context mid-document.

Real use case: a hospitality business owner was reviewing a new supplier agreement before renewing. She pasted the full contract (47 pages) into Claude. She asked it to flag any clauses that limited her ability to switch suppliers, any automatic renewal terms, and any clauses that looked unusually one-sided. It found all three categories, quoted the specific clause text, and summarised what each one meant in plain English. She went back to the supplier with two specific requests. She got one of them. Without the Claude analysis, she'd have signed it as-is.

Writing in a specific voice. When you give Claude strong examples of a tone or voice and ask it to match them, it does so with more consistency than most models. For business owners who need consistent brand voice across emails, website copy, and social posts, this matters.

Code and technical work. If any part of your business involves a website, a spreadsheet formula, or any technical document, Claude's reasoning on technical problems is strong. This is more relevant for larger businesses or those building digital tools, but worth knowing.

---

The Free Tier and What It Covers

Claude.ai has a free tier that's generous enough for moderate daily use, a few conversations per day for drafting and analysis. Anthropic publishes the current model and limits on the site itself, and they update both regularly. Check there if you want the specifics.

The paid tier (Claude Pro, around $27 to $30 AUD/month) gets you access to Claude Opus (the larger, more capable model), more messages per hour, and Projects (where you can save context about your business so you don't have to re-explain it every conversation).

For most small business owners starting out, the free tier is enough to build the habit and assess whether it's worth upgrading. Use it for a week before deciding.

---

A Real Business Use Case, Step by Step

Drafting a difficult client email.

Scenario: a client is unhappy with a project deliverable. They've sent an aggressive email. You need to respond professionally without capitulating on what was agreed in the brief.

Here's the actual prompt structure for this:

Context: I run a small web design agency. A client has sent an angry email 
claiming we didn't deliver what was agreed. The original brief said we would 
build a 5-page website with a contact form and basic SEO setup. We delivered 
all of that. Their email claims we "didn't include social media integration" 
which was not in the brief.

Here is their email: [paste client email here]

Task: Help me draft a response that:
1. Acknowledges their frustration without apologising for something I didn't 
   do wrong
2. Clearly but professionally clarifies what was and wasn't in scope
3. Offers a constructive next step (we could add social integration for an 
   additional fee)
4. Does not burn the relationship

Constraints: Under 200 words. Professional but warm. Not defensive, not 
aggressive. Don't use legal language.

Claude will produce a draft. It will usually also flag where it made assumptions about tone that you might want to adjust. That's the kind of unsolicited useful commentary that characterises its reasoning style.

---

Where Claude Falls Short

No web browsing by default. Claude on its free tier cannot search the internet. If you ask it "what are the top three competitor businesses to mine in Sydney's eastern suburbs," it cannot look them up. It can help you structure a competitive analysis, but you need to bring the data yourself. (Claude Pro with Projects and some tools enabled may have different capabilities, but the base free tier does not browse.)

Not great for "what's happening today" questions. For current events, recent pricing, news about your industry, or anything time-sensitive, Claude faces the same training-cutoff limitations described in Module 3. Use it for analysis, drafting, and reasoning. Use Google for current facts.

Can be overly cautious on sensitive topics. Sometimes Claude will add more caveats and qualifications than you need. If you're asking about a business decision with some risk attached, it may flag more risks than are practically relevant. This is a feature in some contexts. In others, it's friction. You can tell it "I understand the risks, please give me the direct answer without caveats" and it'll generally comply.

---