How to start a web design agency in 2026 (no fluff guide)
The real path from zero clients to a $5K/mo agency: pick a niche, find local businesses without a website, write a 4-line cold email, send 50 a week. No portfolio required.
Most “how to start a web design agency” guides are written by people who have never run cold outreach to a real local business. This one is. After two years of helping freelancers and agencies find local SMB clients at scale, the path that actually works is shorter, cheaper and more boring than the YouTube version. Here it is.
The 4-step starting sequence
Don't pick a niche first. Don't build a portfolio first. Don't set up an LLC first. Do this:
- Find 20 prospects in one evening. Open the No-Website Finder, pick a city + niche, copy 10-20 businesses into a spreadsheet. Total time: 30 minutes. Total cost: $0.
- Write a 4-line cold email. Steal a template from our Cold Email Library. Replace the tokens. Add one specific observation per prospect.
- Send 50 emails in week one.Don't wait for the domain to be perfect, don't wait for the portfolio. Send.
- Niche down after 100 sends. Whichever vertical replied best, double down. The Niche Pickergives you a directional shortlist if you're stuck.
That's it. Everything else - the LLC, the brand identity, the portfolio, the website - can be done after your second client. Most new agencies fail in the inverse order: 6 months on the website, 0 clients sent.
The math that should be on a sticky note above your monitor
Cold email to local SMBs without a website converts at roughly:
- 50 personalized sends per week => 5 replies => 1-2 closed deals/month
- Average local SMB site project: $1,200-$3,500
- So one hour of focused outreach earns ~$300/hr at the start
- That goes up as your portfolio fills and you can quote higher
At 50/week, your first client lands inside 30-60 days. At $5K/mo retainer revenue, you're full-time freelance in 9-12 months. That's the realistic timeline. Anyone selling you faster is selling a course.
Why local SMBs without a website are the easy mode
Three reasons:
1. The value prop writes itself.“You show up on Google Maps but you're invisible to anyone who clicks through to learn more” is a 10-second pitch. No spec sheet needed.
2. Decision-making is single-threaded.The owner decides. There's no committee, no IT department, no procurement. First yes wins.
3. Reply rates are 3-5x higher.Generic cold email to businesses that already have a (mediocre) site replies at 1-3%. Targeted email to no-website businesses replies at 8-15%, because the ask isn't “buy something better,” it's “here's a thing you're missing.”
Pick a niche after, not before
The single biggest mistake new agencies make is picking a niche before they have data. You think you want to do dentists, you spend three weeks on a dental-themed portfolio site, you send 20 emails, get 0 replies, and now you're three weeks behind with nothing to show.
Better sequence: send 100 cold emails across 4 verticals (e.g. dentists, plumbers, restaurants, salons), see which one replies, then commit. The dataset is small but it's your dataset. The market is whispering at you - listen first, theorize second.
See Pick a niche for your web agency for the 12-niche scoring matrix we use.
Pricing: stop charging $300 for a website
New web designers wildly under-price because they anchor to Fiverr instead of to the actual value they create. A clean local SMB site adds $5K-$50K/year in revenue to the business. Charge accordingly.
Three tiers that cover 90% of local SMB work:
- 1-page conversion site: $800-$1,800. Hero, services, social proof, contact, booking. 2-week delivery. Most- requested deliverable.
- Multi-page site (4-7 pages): $2,500-$6,000. Hero, services, about, blog, gallery, contact, FAQ. 4-6 week delivery.
- Retainer (hosting + monthly content/SEO): $200-$800/month. Lock these in from day one - 80% of long-term agency revenue.
Detailed pricing logic at How much to charge for a small business website.
The minimum stack (free or $30/mo)
You don't need a $300/mo tool stack to start. You need:
- Gmail (free) for cold sending. Cap at 30/day from a personal inbox.
- No-Website Finder (free) to source prospects.
- Cold Email Templates (free) for starting points.
- Website Health Check (free) to generate audit-angle hooks.
- Mock Site Preview (free) to attach a visual hook to every email.
- Figma (free for solo) for the actual design work.
- Framer or Webflow ($20/mo per site) once you close.
The 90-day plan
Days 1-7: ICP + 50 prospects + 3 templates. Send the first 25 emails.
Days 8-30: 50 sends/week. Track replies, refine subject lines and openers based on what gets opened.
Days 31-60: First client should land here. Build the site. Charge for it. Capture the case study.
Days 61-90: Use the case study to lift reply rates. Niche down. Start charging 30% more. Add the retainer to every quote.
Mistakes that kill 80% of new agencies
Building the “perfect” portfolio first. Three projects in your target niche, real or speculative, is enough. Spend the time you'd spend on a fourth project on cold outreach instead.
Sending generic emails.“I do beautiful websites” replies at under 1%. “Saw your 4.8 rating from 230 patients - your site doesn't have a booking widget” replies at 12-20%. The first line decides everything.
Asking for too much in email one.“Can we hop on a 30-min discovery call this week?” gets ignored. “I sketched a mockup, 5 minutes to look?” gets a reply. Tiny ask, big yes.
Quitting on month 2. Cold email is a 60-day game, not a 7-day game. Most replies happen on email 2 or 3, not email 1. Discipline beats genius here.
Run this on autopilot
Every step above, automated.
Prospea finds local businesses, pulls verified contacts, writes the first email, and sends the follow-ups. Free plan: 20 leads/month. No credit card.