Sample review · 8 September 2026
A clear change list.
Evidence you can check.
Three observed inconsistencies and two copy hypotheses on Voxelia's English homepage, with concrete next steps.
Page reviewed: 3d-voxelia.com/en
This is a self-initiated demonstration on the provider's existing project, not paid client work. No customer results or conversion gains are claimed. AI-assisted analysis was checked against rendered browser content and page metadata. The source website was not modified.
Start with these changes.
- Align the structured FAQ delivery time with the visible answer.
- Translate remaining German controls and alternative text.
- Clarify the two service regions in the on-site capture card.
01 / High priority · ObservedOne delivery question, two different answers.
The expanded FAQ and main page copy state 24–48 hours. The corresponding FAQ JSON-LD still states 12–24 hours.
Rendered FAQ: within 24-48 hours, often faster
JSON-LD: delivered in 12–24 hours
Machine-readable information can create different expectations when reused by other systems. No resulting impact on sales has been measured.
Fix: Confirm the actual delivery commitment, then generate visible and structured answers from the same approved text. Retain “often faster” only if supported by delivery records.
Acceptance check: expand the FAQ and compare its time and conditions with the JSON-LD. No additional FAQ schema is needed.
02 / Medium priority · ObservedGerman controls interrupt the English experience.
The interactive model preview and rotating workflow include untranslated German instructions and image descriptions.
MIT MAUS DREHEN · DRAUFSICHT · DXF-Handoff kopieren · KLICKEN ZUM KOPIEREN DES HANDOFFS
Replacement copy: Drag to rotate · Top view · Copy DXF handoff · Click to copy the handoff.
Roof view alt text: Top view of the roof model.
Acceptance check: inspect the full animation cycle, image descriptions and accessible control names on the English page.
03 / Medium priority · Observed; owner input neededIs on-site capture available in Zurich too?
Section label: RHINE-MAIN & ZÜRICH
Heading: On-site flights across Rhine-Main
The label names two regions; the heading highlights only one. Actual service coverage was not established in this one-page review.
If both are served: No drone imagery yet? We capture it on site in Rhine-Main and Zurich.
If only Rhine-Main is served: Remove Zurich from the label.
Acceptance check: label, heading and destination page match the owner's confirmed service area.
04 / Copy hypothesis · Not a proven defectName the useful output earlier.
The current headline is concise, while the supporting paragraph supplies the drone, solar and CAD context. Compare it with a more explicit version.
Current headline: From imagery to planning.
Headline: Turn drone images into 3D models for solar and CAD planning.
Supporting copy: Send your image set. We reconstruct the model, review it and prepare the agreed format for your next planning step.
Acceptance check: new visitors can correctly explain the required input, promised output and next step. User testing or adequate behavioral evidence is needed to establish which version works better.
05 / Copy hypothesis · Not a proven defectExplain the account step behind the CTA.
CTA: Start 1 free project
Destination: /en/auth/signup
The signup sequence is reasonable, but its first step could be more explicit. No account was created during the review.
Keep the CTA: Start 1 free project
Add microcopy: Create your account, then upload your image set. No credit card required.
Before use, confirm that the upload sequence and free-project conditions match this wording. The destination's full terms were outside this review.
Search snippet copy.
Suggested title:
Drone Images to 3D Models for Solar & CAD | Voxelia
Suggested description:
Turn drone images into reviewed 3D models for solar and CAD planning. Start with one free project and receive the agreed format for your next planning step.
Editorial suggestions, not a ranking or click-through forecast. Search engines may show different text.
Checks that already look sound.
- One H1 in the rendered document.
- A self-referencing canonical for the English URL.
- An index/follow meta directive; this does not prove actual indexing.
- The skip link and #outputs link have existing targets.
- Organization, WebSite, ProfessionalService, WebPage and FAQPage data are present; full schema validation was not performed.
Additional mobile lab snapshot.
A PageSpeed Insights run on 8 September 2026 at 09:57 CEST reported a performance score of 74, LCP of 5.4 seconds, FCP of 1.7 seconds, total blocking time of 150 ms and CLS of 0. The run used Lighthouse 13.4.1, an emulated Moto G Power and slow 4G throttling.
Real-user data were unavailable. These are lab results from one run, not a field Core Web Vitals assessment or a mobile usability sign-off. The report flags render-blocking resources and LCP request discovery as investigation starting points; savings estimates are not guarantees.
Scope and limitations.
The main review covers rendered public content, an expanded FAQ, link targets and DOM metadata. No analytics, full-site crawl, signup submission, payment, security audit or full accessibility assessment was performed. The lab snapshot is supplementary. Missing image width/height attributes were not automatically called layout defects because CSS may reserve the space.
There is no invented overall audit score. Observed evidence supports the individual findings, not a claim of higher revenue.
A small first engagement.
One public page, five prioritized checks, headline/CTA and metadata rewrites, a printable HTML report and one consolidated question round. Fixed pilot prices: EUR 79 (default), USD 99 or CHF 79, paid once. Scope, delivery contact and date are confirmed before payment. Target turnaround: two business days after the brief and payment are confirmed.
Enquire by email ↗AI-assisted and browser-verified. Implementation is not included. Broader redesign strategy requires a separate scope.