Most portfolio sites are galleries waiting for visitors. This one goes out and gets them, and it is the case study
Portfolio web design for consultants, founders, and independents, argued from a live example. gautamkhorana.com runs 3,100+ URLs in 8 languages on Astro + Supabase, publishes on a queue, and works as its own pitch. The same thinking, applied to your name.
START A PROJECT SEE THE WORK3,143 URLs in the live sitemap 8 languages, one custom admin Astro + Supabase + Netlify, around 80 posts published on autopilot
What a portfolio site is actually for
A stranger with budget lands on your site knowing nothing. They give you about 90 seconds to answer three questions: what exactly do you do, and for whom. Why should they believe you, proof, not adjectives. And what happens next, one obvious action. Most portfolio sites answer none of the three. They open with a montage, hide the work behind unlabelled thumbnails, and offer five social links instead of one booking link.
That is a positioning failure, not a design failure. A portfolio is a positioning instrument that happens to be a website. The design's job is to make the position memorable. The architecture's job is to move a stranger from "who is this" to "booked a call". Design that is tasteful but interchangeable fails the first job. Architecture organised around your work instead of the buyer's decision fails the second.
The case study: this site
I would rather show the argument than assert it, so the case study is the site you are reading.
The numbers first. 3,143 URLs in the live sitemap, 752 of them in English, the rest across 7 other languages. Around 80 blog posts shipped through an automated queue (drafted, run through an AI-detection humaniser, scheduled, published) without me opening a text editor. Comparison directories for hosting, CMSs, databases, and search tooling that bring in strangers with buying intent. All of it static HTML on Astro, content in Supabase behind a custom admin with inline editing, deployed on Netlify, with a build-time SEO linter that fails any deploy that breaks metadata or schema.
The architecture gives every surface one job. The services pages and briefs convert. The work page proves. The blog and guides earn search traffic. The life pages and a playable word game make it memorable, because people hire people, and nobody remembers the fifth tasteful minimal portfolio they saw on Tuesday. Every enquiry that reaches my calendar has already seen the pricing, the work, and the stack, which changes the quality of the first call entirely.
And the design takes one deliberate risk (the retro-terminal identity) instead of ten safe choices. You may not love it. You will place it a week later, and that is the point.
The five decisions that decide whether it earns anything
1. The positioning sentence
Who you serve, what you do, one line of proof, above the fold. "I help brands tell stories" is not a position. "Custom CRM builds for teams leaving HubSpot" is.
2. One action
A booking link or an email address. Not five social icons and a newsletter modal fighting for the same click.
3. Proof architecture
Case studies written as decisions and numbers, not screenshot grids. What was broken, what you did, what changed, what it cost.
4. A memorability budget
One deliberate design risk that makes the site placeable a week later. Spend it somewhere visible.
5. Distribution
The gallery model waits for referrals to check you out. The machine model adds search surfaces (guides, comparisons, a real blog) so the site acquires strangers on its own. This is the difference between a portfolio and an asset.
Three shapes I build
The credibility site. 5-10 pages: position, work, about, contact. Astro, custom design, schema markup, done properly in 4-6 weeks. 6,000-18,000 USD. Right for consultants and fractional executives whose deal flow is referral-led and who need the reference point, not traffic.
The publishing portfolio. Everything above plus a writing system: blog or guides architecture, topic strategy, an SEO foundation built for AI search as well as Google, and a publishing pipeline you will actually use. 8-12 weeks. 18,000-45,000 USD.
The acquisition machine. Everything above plus programmatic surfaces: comparison pages, directories, tools, multilingual where your market warrants it. This site is the reference build. From 45,000 USD, scoped individually.
When this is the wrong call
Four honest cases. Under two years of work to show: a one-page profile and a sharp LinkedIn beat a portfolio project, save the money. Need pipeline inside 30 days: a portfolio compounds over quarters, buy outreach or ads first and build this second. You know you will never publish: buy the credibility site and skip the content engine, do not pay for a blog you will not feed. Budget under 3,000 USD: a Framer or Webflow template is genuinely the right answer, and I will say so on the call rather than take the brief.
Frequently asked questions
What does portfolio web design cost in 2026?
Template customisation (not what I sell): 500-3,000 USD. A custom credibility site of 5-10 pages: 6,000-18,000 USD. A portfolio with a publishing system and SEO foundation: 18,000-45,000 USD. Acquisition-machine builds with programmatic surfaces: from 45,000 USD. Price moves with case-study depth, custom illustration or motion, and how much content system you want. Page count matters less than people think.
How long does a portfolio website take?
Credibility site: 4-6 weeks. Publishing portfolio: 8-12 weeks. The variable is almost always content: clients who write or approve their case studies on schedule launch on time. Copy first, design second is the order that holds deadlines.
Custom build or Webflow, Framer, Squarespace?
Honest take: for a credibility site under 3,000 USD, use a template and spend the difference on photography. Custom earns its price when you want search distribution (fast static pages, schema, a content system) and code you own outright. This site is custom Astro because it publishes at volume in 8 languages, and a template cannot do that job. Match the tool to the ambition, not to fashion.
Do I need a blog on my portfolio site?
Only if you will feed it. A stale blog (three posts, the last one from 2023) reads worse than no blog at all. If you can publish monthly, or want a pipeline that publishes for you, the search surface becomes your cheapest acquisition channel. If not, skip it. The credibility site converts referrals fine without one.
What actually makes a portfolio site convert?
Five decisions: a positioning sentence above the fold, one action, proof written as numbers and decisions, one memorable design risk, and distribution surfaces so strangers arrive on their own. The full argument is on this page. The live demonstration is the site around it.
Can you rebuild my existing portfolio without losing my traffic?
Yes. Full redirect map, metadata transport, schema preserved, 90 days of monitoring after launch, the same discipline as any migration. If your current site ranks for anything, that equity survives the redesign.
What to bring to the call
Your current site, if one exists. The three pieces of work you are proudest of, with the numbers behind them. Who hires you, meaning the actual job title that signs. And your day rate or project rate, so we scope the site to the deal size it needs to win. Thirty minutes later you will know which shape fits, the timeline, and the price band.