Step 3: Capture Business Context Fast

Edward Gotham

20d • General discussion

You don’t need to “think” your way to better AI outputs.

You need to feed it better inputs. Context in = quality out.

Claude can gather live information on your company or competitors quickly when using the right tools.

What this unlocks - Once you can gather inputs properly, Claude can:

analyse competitors in minutes
break down any business model
extract winning messaging angles
find patterns across top-performing content
generate insights instead of generic fluff

Heres a couple of key methods to use to grab useful information for markeitng.

Method 1: Chrome MCP / Computer use

This is the simplest as we've already set it up in step 2 and you simply ask claude in the terminal.

Example prompt - Review this website like a strategist and give me an overview of what the business does. Save it in business_overview.md file.

Do this now for a business you own or one you work with or even just one of your favourite brands.

Method 2: Scraping (Firecrawl / Apify)

When you want depth and repeatability it's good to scrape content online, save it in a clean folder under /raw so you've got a database of research on a particular business readily accessible on your computer.

Two of the most well known and used scraping tools are Firecrawl and Apify. Yoou'll need to setup both of them via mcp.

I've found the easiest way to do this is just to ask Claude to take you through it but before you do that go to firecrawl.dev and create an account, you get some free credits to do some test runs.

Then in your terminal ask cluade "I want to add the firecrawl mcp to crawl websites. please help me set it up." You'll need to share your login details. Its safest not to paste api sensitive info into the terminal, claude will ask you to put it in your env.local

Firecrawl

Use for: