Craft that outbuilds the agencies. Proof that no buyer can see.

A site built like a masterclass, selling an unverifiable business

ThreeSided is the rare company whose website is technically better than most agencies could build. Fully static Astro, per-page metadata on all 85 pages with zero duplicates, a correctly interlinked schema graph, one third-party script, a real 404, clean redirects — top-decile execution, and we audit a lot of sites.

And yet: of 22 portfolio projects, zero are case studies for paying clients. Nine real client names appear only as unlinked thumbnails with the line "More on each on request." Every testimonial is from a manager or colleague, not a buyer. The business repositioned as an independent Las Vegas studio on July 17, 2026 — but every off-site profile still describes a 2-10 person Bay Area WordPress shop, four different locations are live across the web, and a branded Google search returns threesided.com in roughly fifth position behind Facebook, an unrelated Instagram account, a stale Yelp listing, and a different company at threesided.net. The site even argues with itself: one page sells retainers and $25k–$80k builds; another page refuses both.

[ Data verified ] Live fetch of all 85 threesided.com pages (raw HTML + JSON-LD), whois, GitHub API, profile checks (LinkedIn, Yelp, The Org, Behance, Crunchbase), live SERP composition on 16 queries, Search Atlas keyword API — pulled July 26, 2026.

The seven questions, answered

Twenty years of enterprise web leadership, ten days into a new identity

Verified: ThreeSided is Adam Aragon — Independent Product Designer + Engineer, Las Vegas, NV, solo, "usually one engagement at a time." Twenty years and ten roles (the homepage says eleven — one of the site's documented inconsistencies), starting in 1999 with Flash/ColdFusion work on Department of Defense projects, then senior web-strategy roles across FinTech, cybersecurity, marketing automation, enterprise planning, and K-12 edtech — including Head of Web Strategy & Development at Illumio. Domain registered March 2003; the LinkedIn company page says founded 2002; the independent-studio relaunch shipped July 17, 2026 with the journal post "Repositioning and Reality."

The offer is structured as three angles — Strategy & analytics, Build & integrate, Care & polish — each with a genuinely substantial service page. Pricing is published and refreshingly honest ("Ranges are honest — actual range, not low-anchor numbers we then talk you up from"): sprints $8k–$18k, projects $25k–$80k, retainers $6k–$15k/month. The terms page is among the most buyer-friendly we've reviewed. What's missing is a decision: who the ideal client is, which pitch leads, and whether retainers and full builds are in or out — because right now different pages say different things.

Three things competitors can't copy

• One engagement at a time. Structural scarcity that proves senior attention. Our competitive sweep found no one marketing this constraint — it's an unused weapon that answers the buyer's biggest fear about agencies.
• Enterprise-scale receipts. +180% traffic / +60% organic on a re-platformed cybersecurity stack; +27% organic via AI-targeted SEO at 1M+ weekly sessions; a 55,000-URL enterprise redesign; +35% MoM CRO lift in fintech — all on the career page, all currently anonymized.Needs confirmation
• The site itself. WebGPU compute shaders, hand-built CSS, 24 live demos, a public design-system file. For technical buyers, the showcase really is the demo.

The demand is national-niche — and two lanes are standing open

Volume in these consulting lanes is modest but intent is surgical and the market pays real money per click. Strategy: roughly 20% local (Las Vegas entity + hospitality) / 80% national niche (web strategy, AI integration, performance, migration). The most important findings here are SERP structure, not raw volume.

[ Data verified ] Search Atlas keyword API, July 26, 2026. Volumes are US monthly averages; difficulty 0–100. SERP-structure notes from live result-composition checks the same day.

What that demand is worth

Assumptions stated: sprint midpoint $13k and project midpoint $50k (the site's own published bands); retainer midpoint $10k/month if retainers stay; 20–30% close on qualified fit-calls once the proof layer exists; 1.5–3% visit-to-inquiry capture on commercial pages; current inbound assumed near zero given the visibility findings (not measured — analytics access would confirm).

Planning scenarios on stated assumptions, not forecasts — results vary. The strategic point of the aggressive case: at "one engagement at a time," demand past capacity converts into pricing power, which is the best position a solo studio can occupy.

The playbook already exists — solo operators are winning these SERPs

The peer set: boutique & solo

Category leaders — patterns to borrow, not fight

Market structure: "agency"-phrased AI and dev queries are directory-mediated (Clutch, DesignRush, marketplaces) — not winnable and not worth chasing. "Consultant"-phrased queries — web strategy, AI integration, performance, migration, technical SEO — are vendor-page SERPs where individuals demonstrably rank. Every result for "best AI integration consultants" queries is a listicle naming firms, which means the AI-citation layer is placement-driven: publish the comparison content, get named in it, or both.

The honest balance sheet

Technically excellent — and it argues with itself

Let's be clear first: this is not a broken website. It is fully server-rendered (Astro v7 on Netlify), all 85 pages return 200 with unique titles, descriptions, canonicals, and per-page OG images; there are zero duplicate titles, zero canonical mismatches, exactly one H1 per page, clean single-hop redirects, and a correct 404. The critical render path is roughly 8.3KB of CSS and 2KB of JS. Any pitch that opens with "your SEO is broken" would be wrong.

Fix order: reconcile the seven contradictions (a copy edit, not a rebuild), promote the commercial pages to top-level nav, add the booking + email paths, self-host the client images. Days, not months.

Perfect hygiene, zero circulation

On-page fundamentals are excellent — this audit found the cleanest metadata layer we've reviewed at this size. The defects are structural and strategic:

• The internal-link desert. 76 of 85 pages contain no contextual links. All 29 journal posts link only to /contact and back to /journal — zero post-to-post links, zero links to related project pages, zero links to the service each post proves expertise in. All 22 project pages have zero body links to any commercial page. The site's best asset passes no authority and no attention to the pages that sell.
• Content aimed at peers, not buyers. The journal is genuinely active (29 posts May 12–Jul 24, 2026; the latest three days before this audit) and genuinely good — but by its own framing it serves developers ("nerd-sniping welcome"). Not one post addresses a problem a buyer would search.
• Zero off-page equity. Effectively no referring domains beyond stale directory profiles describing the old business. No press as ThreeSided, no reviews anywhere, no directory presence, a GitHub account with 0 stars and 1 follower.
• Small technical cleanups: 18 journal titles overrun 60 characters via a double "— Journal — threesided" suffix; ~21 meta descriptions run long; schema mainEntityOfPage and Service URLs omit the trailing slash their canonicals use; /contact/thanks/ sits in the sitemap despite noindex; only one lastmod in the sitemap; 26 pages carry no JSON-LD (including /work/ and /contact/); two indexed legacy URLs (/our-people, /company-history) return 404 and should be 301'd.

[ Data verified ] Full crawl of the 85-URL sitemap, robots.txt, RSS, redirect and 404 behavior checks, per-page metadata + JSON-LD extraction — July 26, 2026.

Ask an AI who ThreeSided is, and it describes a company that no longer exists

The on-site entity work is genuinely strong — a correctly interlinked Person → Organization → WebSite graph with stable @id references, BlogPosting schema on every post, Service schema on every service page. The problem is everything outside the site: an AI engine cross-referencing ThreeSided finds a 2-10 person Hayward, California WordPress shop founded in 2002 (LinkedIn, Yelp, The Org, Crunchbase), a second person listed on the team, four different locations, and a live profile stating the company focuses on — verbatim, typo included — "Social networkung and SEO."

The brand SERP compounds it: threesided.com sits around fifth for its own name, behind the stale Facebook and Yelp profiles, an unrelated Instagram, and threesided.net. The @threesided handles on X, Instagram, and GitHub belong to a Toronto front-end developer — category-adjacent confusion at the exact moment an AI engine tries to disambiguate.

Fix path: enrich the Organization schema (address, logo, foundingDate, sameAs to every controlled profile); enforce one canonical description across every off-site profile — or retire the profiles that can't be corrected; 301 the dead indexed URLs; publish llms.txt; add FAQPage schema to the 12-question services FAQ; and adopt a consistent disambiguator ("threesided.com" or "ThreeSided Studio") everywhere. AI Search Readiness moves from 4/10 toward 8/10 in one phase — most of the hard work (the on-site graph) is already done.

No Google Business Profile — and the old citations point at the wrong state

• No GBP exists for ThreeSided in Las Vegas. No Bing Places or Apple Maps signals either. The local entity is simply absent.
• The citations that do exist are wrong: Yelp lists a Hayward, CA address with a Sonoma-area phone number, (707) 925-2326 — a number whose current ownership needs confirming.Needs confirmation The Org and Behance still say Bay Area. Nothing anywhere says Las Vegas except threesided.com and GitHub.
• The local prize is hospitality. "Restaurant website design Las Vegas" is currently held by machine-generated city-template pages from firms with no Vegas presence — and ThreeSided's portfolio already includes two hospitality venues plus OpenTable/Resy, menu, ordering, and POS integration services. A real local with real hospitality work beats programmatic pages.
• Prerequisite: a publishable Las Vegas address (a suite or registered-agent address works). Without it: no GBP, no LocalBusiness schema, no local pack. This is a decision for Adam, not a task.
• Do not build mass city pages. No proof layer supports them, and the programmatic-page competitors are exactly what ThreeSided should be positioned against.

The proof exists. It's just invisible, anonymous, or unasked-for

One good form, standing alone

The contact form is genuinely well-designed — budget-direction guidance, timeline selector, honeypot, and the best "what to include" copy we've seen on a studio site. But it is the only path in the building: the email address is deliberately un-crawlable, there is no phone, no booking link (the process page promises "a calendar link" only after a manual reply), no lead magnet, and no client proof standing next to the form at the moment of decision.

Offer ladder: fit call → Web Foundation Audit → sprint (one standardized definition) → project (scoped as chained sprints) → care plan, if retainers stay. The ladder exists implicitly today; naming it is the work.

The engine already runs — aim it at buyers

Twenty-nine posts in ten weeks with zero calendar theater is a real content engine — most agencies can't sustain it. The fix is aim, not effort. Five pillars, each feeding an existing service page:

• AI integration in the real world → Build & integrate. Build-vs-buy guides, real cost breakdowns (the unserved band between SaaS chatbots and $10k+ agency pilots), RAG and streaming-UI case notes, and the comparison content AI engines cite.
• Web strategy & analytics for non-enterprise teams → Strategy & analytics. "Consultant vs agency vs hire," GA4 + Looker dashboard patterns, pre-redesign discovery guides.
• Performance & Core Web Vitals → Care & polish. He already writes this ("Lighthouse 100s are mostly a font problem") — add buyer framing and one free diagnostic tool, the proven link-and-lead engine in this niche.
• Legacy migration → Care & polish. Per-stack migration guides (WordPress, Drupal, ColdFusion — he started in it), including honest "when not to migrate" content.
• Hospitality web → Build & integrate + the local layer. Las Vegas restaurant guides, reservation-integration notes, and the accessibility-for-restaurants vertical nobody owns.

Automation that fits how he already works: form → CRM → same-day reply with booking link; pre-call research briefs before every fit call; the Foundation Audit report generator (his walkthrough-video delivery style, systematized); case-study drafts from project notes; review requests triggered at handoff; each journal post repurposed into LinkedIn posts on a cadence. Internal-link retrofit on all 29 existing posts is a one-time, high-value pass.

Elite craft, missing commerce — the score says both

The first five dimensions score current state (higher = stronger today). The last three score headroom (higher = more opportunity to capture). Scores are directional, built from the verified findings in §6–§12 — and deliberately not inflated for technical polish: craft and commercial foundation are scored as the separate things they are.

From contradictory and unverifiable to consistent, proven, and cited

What to do first — ranked by impact

The first five moves

• Decide: retainers in or out, one sprint definition — then fix /process/ and /contact/ to agree.
• Put Services in the top-level nav and a booking link on every commercial page.
• Rewrite the LinkedIn company page and correct or retire the stale Yelp, The Org, and Crunchbase profiles; 301 the dead indexed URLs.
• Pick the 2–3 clients who will go on record, and start those case studies.
• Self-host the nine client thumbnails and enrich the Organization schema (address, logo, sameAs, founding date).