Most tool stack content is written to be long, not useful. You get a list of thirty options, every one rated five stars, every link an affiliate link, and nobody tells you what any of it is actually for. So here's the real list. This is what the two of us open in a normal week to run Luda, grouped by the job each one does. Nothing is here to pad the list. If it's below, we use it.
Automation and workflow tooling
n8n. This is the core of the business. Client work runs through it and so does a lot of our own back office. A form gets filled, n8n catches it, routes the data, logs it, notifies the right person, and kicks off whatever follow-up is supposed to happen. Internally it's the same lead notifier we give away, plus a pile of small jobs that used to be somebody's manual afternoon. If one tool on this list had to survive, it's this one.
Thinking and research
Claude. Our default for drafting and working through problems out loud. Content drafts, client copy, cleaning up messy logic before we build a workflow, and rubber-ducking a design when we're not sure how to structure something. It's less a single feature and more the thing that's open while we work.
Perplexity. For fast research when we need an answer with sources attached, not a guess. Not a daily tool, but when we need to check a claim before it goes in a piece, or get up to speed on something fast, this is where we go.
Building the web dev side
Cursor. Where the actual code gets written. Client sites, internal tools, and our own CRM (more on that below) all get built here. Claude handles the thinking and the drafts; Cursor is where that becomes a working build.
Canva. Handles the design side of web work, quick mockups, visual direction, assets that need to look right before anything gets built out. Same tool we reach for in content, doing a second job here.
The base layer: docs, storage, and email
Google Suite (Drive, Docs, Sheets, Gmail). The boring foundation everything else sits on. Drive holds the files. Sheets doubles as a lightweight database for the automations that don't need a real one yet, which is why so many n8n jobs end in a new row somewhere. Gmail, on Google Workspace for the custom domain, is how we talk to clients, because clients don't want another app.
Notion. Docs, content planning, and the checklists we reuse for onboarding.
CRM and internal tracking
Our own CRM. Built in-house, running on Cursor, Supabase, and Vercel, the same tools already on this list. We started on a Google Sheet, the way most people do, and it worked fine until lead volume and follow-up tracking outgrew what a spreadsheet could handle cleanly. Rather than buy something built for a business that isn't ours, we built one. Leads, follow-up queues, call logs, and email outreach all live here now, and it's genuinely one of the tools we open most during the day, not a side project we built for fun.
Twilio. The dialer inside that CRM runs on Twilio. When someone on the team makes a call or the system logs one, this is the infrastructure underneath it.
Finding clients: lead sourcing and outreach
This is the pipeline that fills the top of the funnel. Each tool does one job, and they run in order.
Outscraper. Pulls business data from public sources so we can build targeted lists instead of buying stale ones. This is usually the first step: get the raw list of the kind of business we want to reach.
Apify. The other half of our scraping. When we need data a standard tool won't grab, or a more custom pull, this is what we reach for. Between this and Outscraper we can build most lists we need without paying for a prepackaged database.
Prospeo. Finds the contact email for the people on those lists. A raw list of businesses isn't useful until you can actually reach a person, and this is the step that turns one into the other.
NeverBounce. Verifies those emails before we ever send. Skip this and you send to dead addresses, your bounce rate climbs, and your domain reputation takes the hit. This is the unglamorous step that protects everything downstream.
Instantly. Where the verified list actually gets emailed. Runs the cold campaigns and the sending logic so outreach goes out steadily instead of in one risky blast. The scraping and verifying above exist to feed this cleanly.
Content
Adobe Premiere. The heavier video editing, mainly long-form. When a cut needs real control instead of a quick trim, it happens in Premiere.
Adobe Podcast. Audio cleanup. Bad room, echo, or a rough recording goes in and comes out listenable. It's the step that keeps audio quality from being the reason a good piece underperforms.
Development and technical infrastructure
GitHub. Holds the custom code we write and, just as important, backups of our workflows, so a broken server is an annoyance and not a disaster. The day you need a backup is not the day you want to start making them.
Supabase. The database behind client automations, and our own CRM, once a Google Sheet stops being enough. When something needs to store real data, look it up quickly, and hold up under load, that goes here. We reach for it earlier each year, because starting on a real database beats migrating off a spreadsheet later.
Vercel. Where the front-end pieces get deployed, client builds and our own tools alike. This is what puts something online and keeps it there.
If you're starting from zero
You don't need this whole list to start. Most of it, we earned the need for over time.
Here's what actually matters on day one: Claude, your full Google Suite, and n8n. That's the stack. Docs, Sheets, Gmail, and Drive will cover more than you think, use all of it from the start instead of scattering across five different apps. n8n is genuinely one of the best places to learn automation, because it's visual, it's forgiving to learn on, and things breaking is part of how you actually get good at it. You'll hit failures. That's the job, not a sign you picked the wrong tool.
For finding your first clients, Google Maps is enough. You don't need Outscraper or Apify yet, those earn their spot once you need real volume or a more precise pull than a manual search gives you, but starting out, Maps will get you everything you need depending on what you're selling. Same goes for Instantly, useful for mass outreach, but whether it's the right move depends heavily on your industry, it's not a universal day-one tool.
One honest warning: we'd stay away from Cursor early on. It's the tool we use daily for our own web dev work now, but when we were starting, it became a trap. You can build literally anything in it, and Lucas and I fell into building instead of selling, spending our time deep in a tool instead of out finding clients. If you're starting out, your job is sales and outreach, not building something impressive nobody's paying for yet.
This list will look different in six months. Some of these tools we'll outgrow the way we outgrew a spreadsheet for CRM work, some new ones will earn a spot we can't predict yet. That's normal. The point was never to lock in a stack forever, it's to use the right thing until it stops being the right thing, then change it.