Studio Self-Audit — After Implementation

AlphaBeta.design — Studio Self-Audit (After)

Same 39 findings as the original audit. This version overlays the implementation pass — every card carries a status badge and the resolved ones carry a “What changed” note pointing at the concrete fix. See the original audit (before) for the pristine first-pass findings.

Site: www.alphabeta.design
Stack: Next.js · Vercel · Sanity
Pages in Sitemap: 17
Pages Reviewed: 6 in depth
Audits Run: 34 of 39
Audit Date: May 2026

Findings at a glance — post-implementation

Each card shows the current severity count next to the original first-pass count (struck through). The High item was resolved, three Medium items were downgraded, and findings rolled up into Already Strong as fixes shipped.

Audits in Menu

High

Medium

Low

Already Strong

Needs Access

What changed since the first audit

Each finding from the first pass now carries a status. Eight got resolved outright. Three are partial — code-side ready, manual follow-up pending. Eleven “Already Strong” items stayed strong. Ten remain open as deliberate roadmap. Seven aren't addressable until the studio has paid engagements, traffic data, and real users to test with.

Resolved

Partial

Still True

Open

Needs Real Data

Executive Summary

This is what a well-built solo-studio site looks like under audit after the audit has been acted on. The metadata layer is in order, the conversion path is intact, positioning lands in 30 seconds, pricing is published, audiences are named, and the “what I won't do” filter still does its work. All the things that were already strong, stayed strong.

The high-severity item has been closed. The dated <meta name="keywords"> tag is gone, <html lang> reads en-US to match the locale declared everywhere else, and the apex → www redirect chain is documented as the single remaining manual step (Vercel dashboard, not code). Three mediums dropped a tier; findings rolled up into Already Strong as fixes shipped. What remains open is deliberate roadmap — axe-core verification, keyboard / screen-reader walkthroughs, the first journal post — and the seven items that need real engagements, traffic data, and live users to address.

What's already strong

SEO foundation — meta descriptions on every page, canonicals, real OG cards per case study, three JSON-LD schemas.
Trust hygiene — current copyright, no leaked placeholders, clear tech credits, no dead links observed.
Conversion path — real contact form with budget tiers, “Under $3k? Try these first” qualifier, response-time commitment.
Voice consistency — “Earn every element,” “Show the seams,” “The interview is the brief.” Senior signal throughout.
Authority claiming — VARA Winery, Boomtime, 100+ client sites named directly. Credentials surfaced, not buried.

What's worth tightening

Outdated keywords meta tag. Ignored by Google since ~2009. Signals dated SEO. One-line delete.
Two-hop redirect chain. Same fix as every Vercel/Next.js site — set www as primary in domain settings.
No testimonials or named buyer logos. Studio is new (founded 2026), so this is forgivable — but it's the next-phase priority.
Case-study depth is light at ~187 words per case. Senior buyers want more decisions named, more trade-offs surfaced.
Spanish-language capability declared in schema but not advertised in copy. Either commit or remove.

Tier 1 — Strategic

What the site says, who it's for, where it sits, what it asks.

Tier 1

Strategic

10 of 10 complete · 5 already strong

30-Second Comprehension Test

Already Strong✓ Resolved

What changed

Response-time pip surfaced on the homepage hero — visitors see “Same business day · One working day reply” before they have to scroll to /contact.

Reading the homepage for 30 seconds: “One-person UX/UI studio in Portland, owned by Alfonso Barreiro. Does strategy, design, and Next.js development. Three engagement shapes with published prices. Has six case studies. Won't do logos, WordPress, or spec work.” Positioning lands in five seconds. Audiences are named explicitly. Pricing is transparent. Questions a stranger would have: who has hired you (no client logos), what's the response time (visible on /contact, not homepage).

Keep

Maintain. Consider surfacing the response-time commitment (“Same business day. One working day reply.”) on the homepage hero, not just on /contact. Costs nothing, lifts confidence.

Audience Inference

Already Strong✓ Resolved

What changed

Fourth audience line added — Series A teams hiring a fractional design lead before the first full-time hire. Matches the Partnership-tier price point.

Homepage explicitly names three audiences: “Founders shipping what's next, small businesses and nonprofits modernizing what they outgrew, and product teams without a sitting designer.” No drift. No “we work with anyone.”

Keep

Maintain. Consider adding a fourth audience for the Partnership tier specifically — Series A teams hiring fractional design lead before a full-time hire. The $17k/mo price point lives in that buyer's bracket.

Message Hierarchy

Already Strong✓ Still true

Claim: “Strategy, design, and the build. In one hand.” — distinctive and clear. Proof: six case studies with named projects, two prior leadership roles. Ask: three CTAs at different commitment levels (See the work, See pricing, Start a project).

Keep

Holds together. The claim does the work the tagline alone couldn't. Don't dilute.

Competitive Positioning Grid

Low✓ Resolved

What changed

Added a “What's unusual here” strip on the homepage between the hero and Selected Work that names all four unclaimed-ground differentiators in one place: published pricing, no logos / no WordPress, bilingual (previously only on /about), operator background (previously only in the bio paragraph). The positioning was always there — it's now stated as deliberate, not incidental.

Comparable set for one-person UX/UI studios with published pricing: Frank Chimero (writes for himself, not a studio offering), Naomi Wilkins, Tobias van Schneider's DesignStudio, ueno (now larger), Pentagram (different tier), HTML Burger, Slimm Designs. Most one-person studios don't publish prices. Most that do don't bundle strategy + design + build under one engagement model.

Recommendation

Unclaimed ground worth owning more loudly: (a) Published pricing — rare in this market. (b) The “I won't do logos” filter — most studios will take any work. (c) Bilingual capability — declared in schema but not advertised. (d) The “before this studio” credibility (VARA, Boomtime) — most solo designers don't have ops leadership in their background. Lead with these.

Keyword & Topic Positioning

Low◑ Partial

What changed

Site-wide metadata description now leads with the buyer-search terms the original audit flagged as missing: Next.js web design + development, UX audits, fractional design-lead. Still open: a dedicated “Services we're known for” surface on the page itself — metadata is half the work.

Owns:“Alpha Beta Design,” “ABD UI” (design system), “Launchpad sprint” (custom term for fast case studies). Competes on: “UX designer Portland,” “UI designer Portland,” “design studio Portland” — crowded markets. Missing: “fractional design lead” (buyer-side language for the Partnership tier), “Sanity CMS development,” “design system consulting,” “UX audit Portland” (which is now being sold!), “Next.js designer.”

Recommendation

Add a “Services we're known for” surface — Fractional Design Lead, Design System Architecture, UX Audits, Next.js Web Development. These are buyer-search termsthe site doesn't currently target. Update the keywords meta tag if you're keeping it (you shouldn't — see Audit 38).

Authority Signal Inventory

Medium◌ Needs real data

Present:Director of Marketing & DTC Operations at VARA Winery (rebuilt e-commerce), VP of Operations at Boomtime (100+ client sites across healthcare, legal, hospitality, e-commerce), 18 years practice, Spanish + English. Underclaimed for a new studio: no testimonials (yet), no client logos, no press, no speaking, no awards. Studio founded 2026 — most of this is forthcoming, not missing.

Recommendation

First-90-days priority: capture one testimonial from each of the next two engagements. Even one named client logo on the homepage shifts perception significantly. Until then, the prior-role credentials carry the work.

Conversion Path Audit

Already Strong✓ Still true

Multiple commitment levels. Low: See the work, See pricing. Medium:Email link, “Let's talk” button. High:Full contact form on /contact with project type, budget tier, and free-text message field. Plus a published response-time commitment (“Same business day. One working day reply.”). Plus a qualifier section for under-$3k buyers redirecting to Squarespace/Carrd/Webflow — generous, credibility-building.

Keep

Excellent. The “Under $3k? Try these first” qualifier is the kind of move that builds trust with overspending and under-qualified buyers alike. Keep it.

AI Visibility / AEO

Low○ Open

robots.txt allows all crawlers (no AI bot blocks). Three JSON-LD schemas surface the brand, founder, and website to AI systems. Domain is new (2026), so AI citations will accrue over time. Untested by the audit — but the structural prerequisites are correct.

Recommendation

Run the 10-question prompt test across ChatGPT, Claude, Perplexity, and Google AI Overviews quarterly. Track which AI surfaces start citing Alpha Beta correctly. Expect 6-month delay before meaningful indexing.

Content Currency & Cadence

Low◑ Partial

What changed

Scaffolded /journal with an honest placeholder: cadence named (quarterly · four engagement debriefs a year, not weekly filler), topics on deck listed, current-evidence surfaces linked (/work, /services/audit-example, /services), email-subscribe pattern noted. Route now exists in the sitemap. Structural absence fixed; the first real field note still has to be written after the next engagement wraps.

Sitemap lastmodis today's date. No blog. No journal. No field notes. Studio launched 2026 — the absence is expected, not failed. But for a studio whose Tier 1 audience is “Founders shipping what's next,” a regular field-note cadence would build SEO and authority simultaneously.

Recommendation

Pick a quarterly cadence — four posts a year — on real engagement debriefs. Each one is a 600-word “here's what I learned shipping X.” Doubles as portfolio depth and inbound SEO. Skip a fake-feeling weekly blog.

Trust & Hygiene Pass

Already Strong✓ Still true

© 2026 footer (current). Tech credits clean (“Built with Next.js, Motion, and Lenis”). No “Powered by Squarespace/Wix/template-name” leaks. No “Lorem ipsum” or “Make it stand out” placeholders observed. HSTS header set. HTTPS enforced.

Keep

Hygiene is current. Add a privacy policy and cookie banner when traffic includes any EU/UK visitors. Until then, the absence is forgivable for a US-only studio.

Tier 2

UX & Interaction

7 of 8 complete · 1 N/A

Information Architecture (IA)

Already Strong✓ Still true

Five top-level routes: /, /work, /services, /about, /contact. Case studies nest under /work/[slug]. Clean, conventional, scannable. No orphan pages. No duplicate URLs. No “/home” leak.

Keep

IA holds. If the studio adds a journal / field notes section, the natural slot is /journal or /notes alongside Work.

Navigation Audit

Already Strong✓ Still true

Four-item top nav (Work, Services, About, Contact). Mobile menu via “Open menu” trigger. Footer mirrors top nav under “Studio” heading + a “Connect” column (Email, LinkedIn). Predictable and uncluttered.

Keep

Solid. If you add Journal or Audits-as-a-product page, slot before “Contact.” Five items is the upper bound for a clean top nav.

Task Flow / Jobs-to-be-Done

Already Strong✓ Resolved

What changed

Every case study now closes with an “Or start one like this — See the packages →” link below the existing “Next project” affordance. Internal loop (case study → case study) and conversion loop (case study → services) both exist now.

Primary JTBD: “evaluate this studio and decide whether to talk.” Flow: homepage → case study → services or contact. Each path leads cleanly to the contact form. No dead ends.

Recommendation

Consider one explicit “what to do next” suggestion on each case study (e.g., “See the package this matches” or “Start a project like this”). Currently every case ends at the View All link — fine, but could be tighter.

Form & Conversion UX

Already Strong✓ Still true

Real form on /contact. Required fields: Name, Email, Message. Optional segmentation: Project type (12 options matching the services tiers), Budget (5 tiers). Async-friendly. No file upload yet.

Keep

Form does its job. Optional adds: company URL field, file upload for buyers with existing brand assets, optional “How did you hear about us” tracking field. None are blockers.

Onboarding Flow Audit

N/A◌ Needs real data

Not applicable. Studio marketing site, not a product with signup.

To be completed

If the studio launches a productized service (e.g., self-serve UX audit, a brand-pack template, an embedded design system trial), build the onboarding before launch — and re-run this audit on the new surface.

State Completeness Audit

Already Strong✓ Resolved

What changed

Custom /contact/thanks page built and wired as the Formspree post-submit redirect via a _nexthidden input. 404 page already had helpful redirects to Work/Services/About/Contact. HTML5 validation on every form field. Loading state handled by the browser's tab/URL indicator (full-page submit).

Need to verify: 404 page (Next.js default vs. custom), form success state, form validation error state, loading state for form submission. Without a real submit test, completeness is unverified.

Recommendation

Verify all four states are designed. Custom 404 with helpful redirects (Work, Services, Contact). Form success: thank-you message + “what happens next.” Validation: inline, not alert. Loading: button disabled state, not generic spinner.

Microcopy / UX Writing

Already Strong✓ Still true

Voice is consistent and senior. Examples: “Earn every element. Nothing on the page should be there because it usually is. If it doesn't help, it goes.” · “The interview is the brief.” · “Show the seams.”· CTAs are action-led (“Start a marketing site,” “Start a brand audit + tightening”). No filler. No AI-tells.

Keep

Best-in-class for this category. Maintain. Apply the same standard to forthcoming error messages and form confirmations.

Responsive & Cross-Device Parity

Low○ Open

Next.js + custom build means full responsive control. Viewport meta correct. Without device-by-device QA, full parity unverified — but the framework is right.

Recommendation

Run a quarterly cross-device QA pass on iPhone, Android, iPad, desktop. Particular attention to long-scroll case study pages and the contact form. Document the QA pass in CMS Guide if you formalize.

Tier 3

Visual & UI

7 of 7 complete

Visual Hierarchy

Already Strong✓ Still true

Numbered sections (01 Selected Work, 02 Three Ways to Work Together, 03 What I Won't Do, 04 How We Work) act as a strong second-level scaffold. Single H1 per page. Clear typographic ramp from H1 → section number → H2 → case-study H3.

Keep

The numbered-section pattern is a senior choice. Keep it. Apply consistently if Journal section ships.

Typography System

Low○ Open

Multiple font families loaded — typical for a brand-built site, but adds payload. DM Sans visible in source. Need full audit to confirm scale, weight, and consistency across surfaces.

Recommendation

Document the type scale in CMS Guide (or design-system case study). Limit display families to two. Audit at quarterly intervals to prevent drift as new components ship.

Color & Contrast System

Low○ Open

Brand uses warm-neutral palette (cream backgrounds, charcoal ink, burnt-orange accent). No images-with-overlay-text patterns observed (which is where contrast usually fails). Full contrast audit pending automated scan.

Recommendation

Run axe-core or WAVE on each page. Confirm AA contrast on every text/background pairing. The palette appears compliant by inspection — but verify rather than assume.

Brand Consistency

Low○ Open

“Alpha Beta Design” / “ABD” / “Studio of Alfonso Barreiro” — three name forms, all used contextually. Title separator is consistent (· middle dot). Footer attribution is consistent across pages.

Recommendation

Maintain. If “ABD” gets used in client-facing contracts or invoices, add a one-line note in the brand guidelines on when each form applies (full name for first contact and legal, ABD as shorthand inside ongoing engagements).

Component & Design System

Already Strong✓ Still true

The studio has a case study about its own design system (“ABD UI: Production Design System”) — eating the dog food in public. 120 components, 15 color tokens, four output formats. Architecturally sound.

Keep

Strong signal to prospective design-system buyers. Consider adding a public Storybook or Figma library preview link — proof that the system actually exists, not just the case study about it.

Iconography & Imagery

Low○ Open

Case studies have per-case cover images (verified on /work/iris-calder/cover.png and /work/emblem-coffee/cover.png). All alt text present. No stock-feel imagery in the source — the imagery is intentional and project-tied.

Recommendation

Maintain. As case study count grows, document a consistent image-system spec (aspect ratios, treatment, file naming) in CMS Guide.

Motion & Interaction Polish

Already Strong✓ Resolved

What changed

Verified prefers-reduced-motion respect across every motion surface: SmoothScroll (Lenis bails before init), ScrollReveal, Magnetic, and WorkGridCard (all check useReducedMotion() and render their static path when the OS preference is set).

Site uses Motion (Framer Motion) and Lenis (smooth scroll). Both are tasteful, modern libraries — but ship JavaScript weight and need prefers-reduced-motion handling to be a11y-safe.

Recommendation

Verify prefers-reduced-motion respect for all motion. If not implemented, add. Lenis specifically should fall back to native scroll when reduced motion is preferred. Cheap fix, big a11y win.

Tier 4

Research & Insight

3 of 6 complete · 3 require access

Heuristic Evaluation (Nielsen)

Already Strong✓ Resolved

What changed

Added five new FAQ entries covering the exact gaps the original audit named: rush timelines, equity offers, international clients, retainer minimums, and project pauses. The “help/documentation” heuristic has answers now.

Quick pass: (1) System status — N/A pre-form. (2) Real-world match — yes; language is operator-friendly. (3) User control — fine, no traps. (4) Consistency — strong. (5) Error prevention — N/A pre-form. (6) Recognition over recall — strong nav labels. (7) Flexibility — fine. (8) Aesthetic + minimalist — strong; numbered sections help. (9) Error recovery — pending form-state verification. (10) Help — no FAQ surface yet.

Recommendation

Add an FAQ section to /services or /contact answering: rush timelines, equity offers, international clients, retainer minimums, what happens if the project pauses. Lowers friction for serious buyers.

Usability Testing

Requires Users◌ Needs real data

5–8 representative buyers (Series A founders, small-business operators, agency leads hiring fractional design) walking through the buyer journey: discover → evaluate → contact → scope call.

To be completed

Run when the studio has done 5+ engagements and can recruit recent clients or near-clients. For now, post-engagement client interviews serve a similar function. Five 30-min calls cover the highest-leverage usability findings.

Persona Development & Validation

Medium◌ Needs real data

Three audiences are named on the homepage (Founders shipping what's next, small businesses + nonprofits modernizing, product teams without a sitting designer). These are positioning audiences, not evidence-grounded personas. They came from intuition, not research.

Recommendation

Validate against the first 5–10 actual engagements. Update or refine the audience list to reflect the buyers who actually convert (vs. the ones you imagined would). Most studios discover their real audience differs from their stated one by year two.

User Journey Mapping

Low○ Open

Inferred journey for a Series A founder: hears about ABD (LinkedIn or referral) → Googles “Alpha Beta Design” → lands on home → reads positioning → checks /work for credibility → scans /services for budget fit → uses /contact form or emails directly. Friction points: no testimonials at the credibility-check step, light case studies at the depth-check step.

Recommendation

Both friction points are time-and-engagement-dependent. Capture testimonials early. Deepen case studies as engagements complete (current ones at ~187 words; senior buyers want 800–1200 with decisions named).

Analytics & Behavior Audit

Requires Data◌ Needs real data

Requires access to Vercel Analytics, Google Analytics, or any other behavioral tool the studio runs.

To be completed

Verify event tracking on: form submit, email link click, “Start a [package]” CTA clicks, scroll depth on homepage, time-on-page for case studies. Establish a monthly review cadence. Two-day audit when access is granted.

Voice-of-Customer Synthesis

Requires Access◌ Needs real data

Requires access to inbound emails, scoping-call transcripts, client testimonials, post-engagement feedback. Studio is new (2026) — VoC corpus is small.

To be completed

Build the corpus deliberately. After every scoping call, log: what they asked that the site didn't answer, what they got wrong about the offering, what made them commit. After every engagement: one structured 20-min debrief. The patterns drive content roadmap.

Tier 5

Accessibility & Inclusion

5 of 5 complete

WCAG 2.1 AA Compliance

Already Strong✓ Resolved

What changed

New /accessibility page published with the studio's WCAG 2.1 AA conformance statement, what was done specifically, what we don't claim, and a barrier-report mailto. Footer link added so it's reachable from every page.html lang flipped to en-US to match the locale declared in JSON-LD + OG.

Alt text present on all 10 home-page images (0 missing, 0 empty). Single H1 per page across the 6 inspected. Page titles are unique and meaningful. HTML lang attribute set. No formal automated scan run yet.

Recommendation

Run axe-core + WAVE on every page. Add a public accessibility statement (one-pager) at /accessibility or as a footer link — visible commitment, takes one hour. Re-test after every significant launch.

Keyboard Navigation

Low○ Open

Next.js with default focus styling. Skip-to-content link not verified. Tab order should be DOM order — but pending unplugged-mouse test.

Recommendation

15-minute keyboard-only walkthrough. Verify skip link, focus visibility on every interactive element, modal escape (if any), form field focus order. Almost always reveals one or two missed focus styles even on well-built Next.js sites.

Screen Reader Audit

Low○ Open

Heading order appears semantic (one H1, multiple H2/H3 in logical hierarchy). JSON-LD provides clear page context. Nav structure is clean. Pending live VoiceOver test.

Recommendation

20-min VoiceOver walkthrough on home, /services, /contact, and one case study. Verify form labels read correctly, links have meaningful text, images have informative alt (not just decorative).

Cognitive Accessibility

Already Strong✓ Still true

Copy is direct and concrete. Numbered sections aid orientation. Sentences are short. Jargon is contextually appropriate (used where it signals expertise to the target buyer, not as decoration). Reading level appropriate for B2B audience.

Keep

Maintain. If launching FAQ or longer-form Journal content, apply the same standards — short sentences, named decisions, no abstraction without example.

Inclusive Language

Already Strong✓ Still true

No ableist language, gender assumptions, or exclusionary patterns observed. “Founders,” “teams,” “operators” used as gender-neutral. Spanish capability declared in schema (not yet advertised — see Audit 22).

Keep

Hold the standard. If bilingual capability gets advertised, ensure Spanish copy receives the same voice rigor — not auto-translated.

Tier 6

Technical Supporting Layer

2 of 3 complete · 1 requires data

Performance & Core Web Vitals

Low○ Open

Home HTML weighs 120 KB — reasonable for a marketing site with multiple cases. Vercel edge cache returning HIT. Next.js prerender header present. Motion + Lenis libraries add JS weight but are tasteful. Real LCP, CLS, INP pending Lighthouse.

Recommendation

Run Lighthouse and PageSpeed Insights against home, /services, and one case study. Target Performance score 90+ on each. Optimize image weight if any case study covers exceed 200 KB. The architectural foundation is sound — this is fine-tuning.

Technical SEO

Low◑ Partial

What changed

Two of the three sub-findings are resolved at the code level: the outdated <meta name="keywords"> tag has been removed entirely, and <html lang> now reads en-US to match the locale in JSON-LD + Open Graph. The two-hop redirect chain (apex → www) still needs a Vercel domain- setting change (set www.alphabeta.design as primary); documented as a TODO in app/layout.tsx since code can't drive a Vercel dashboard toggle.

Two technical hangovers from earlier builds: (a) outdated keywords meta tag still present (ignored by Google since ~2009; signals dated SEO). (b) Two-hop redirect chain — http://alphabeta.design → 308 → https://alphabeta.design → 307 → https://www.alphabeta.design. Should be a single 301. Otherwise SEO is exemplary: meta descriptions everywhere, canonicals, OG cards per case study, three JSON-LD schemas.

Recommendation

Delete the keywords meta tag in app/layout.tsx. Set www.alphabeta.design as primary domain in Vercel — collapses the redirect chain. Also: change <html lang="en"> to "en-US" to match the locale declared everywhere else. Three small fixes, ten minutes total.

Analytics Implementation

Requires Access◌ Needs real data

Requires access to Vercel Analytics dashboard, Google Analytics setup, or any other tracking layer.

To be completed

Verify event firing on form submission, CTA clicks, email-link clicks, scroll depth. Once a custom event taxonomy exists, document it in CMS Guide. One-day audit when access is granted.

Caveats & scope boundaries

This is a public-side audit of a studio I run. The five access-required audits (#15 onboarding N/A, #27 usability with real buyers, #30 analytics, #31 voice-of-customer, #39 analytics implementation) would only complete with internal data — Vercel Analytics, scoping-call notes, post-engagement debriefs. The audit is fair, not flattering: it names what's strong, what needs tightening, and what's premature for a 2026-launched studio. A re-run in six months will look different — testimonials, real case-study depth, and the first wave of AI citations should all have moved.

I audit my own site. I'll audit yours.

The same framework, applied to your studio site, product, or marketing presence. Same color-coded severity. Same prioritized action list. No surprises about what you'll get.

See the packages Request an audit