RSGE
Case Study

One supplier. Ten services. Zero fines.

Most companies dealing with workplace compliance end up with a patchwork of providers: one for occupational safety, another for fire safety, a third for HR paperwork, someone else for equipment inspections. Every handoff is a gap where something falls through - and in compliance, gaps mean fines.

RSGE was built to close that gap. A Romanian HSE firm with over 12 years in operation, they are a genuine one stop shop: SSM, PSI, RSVTI, HACCP, HR outsourcing, environmental protection, and more - all under one contract, one team, one phone number. Over 485 clients have renewed after year one. 19,300 employees trained. 600+ regulatory inspections attended. And a guarantee written into every contract: if your company gets fined while under their watch, RSGE pays it.

That's an extraordinary offer. The site wasn't making it.

The brief: design a new website from scratch - information architecture, visual identity, and copy direction - that would finally match the scale and confidence of the business and turn existing traffic into leads.

Skills, tools, deliverables.

A credible business, invisible online.

The existing site ran on Avada (WordPress) with Slider Revolution, WPML, and years of accumulated plugin debt. Dense, unstructured, and slow. Three problems were costing them leads every day.

The one stop shop advantage was invisible. "1 Furnizor, 10 Servicii Complete, 0 Griji, 0 Amenzi" is a sharp, genuinely differentiated promise. But it was lost in the noise. Visitors couldn't quickly understand that they could replace their entire compliance vendor stack with a single call to RSGE. The site was listing services like every other firm, not selling the model.

Ten services, no orientation. With 10 distinct offerings across safety, HR, and equipment, a visitor needed a map, not a dropdown. A business owner arriving from Google knew they had a problem - they didn't know which service solved it. Decision fatigue, and they're gone.

The site was failing on mobile before a single word loaded. A post-launch PageSpeed audit confirmed the damage: TTFB of 4.7 seconds, LCP of 6.2 seconds, Core Web Vitals: FAILED. Slider Revolution was shipping ~400 KB of JS and CSS to render a static hero image. On a lead-generation site, this is revenue evaporating silently.

What needed to happen.

• 01 Make the one stop shop model impossible to miss. The core proposition - one contract covering 10 service lines - had to land above the fold, before any service detail.
• 02 Map 10 services without overwhelming the visitor. Every service needed a clear entry point, grouped by domain, so a visitor could self-identify their need in under 5 seconds.
• 03 Surface the guarantee prominently. RSGE pays client fines if a sanctioning occurs under their watch - a contractual commitment that competitors can't match. It needed to be a design moment, not a footnote.
• 04 Establish trust before explaining the offer. 12 years, 485 renewals, 19,300 employees trained, clients like Jouve and Paragon - all of this had to come early, before the sales pitch.
• 05 Diagnose and document the performance crisis. TTFB 4.7 s, LCP 6.2 s, Core Web Vitals FAILED. A post-launch technical audit and remediation plan was needed - not just a redesign.

What was done and how.

• 01 Hero built around the proposition, not the brand. "1 Furnizor / 10 Servicii Complete / 0 Griji / 0 Amenzi" runs large and heavy above the fold - four lines, each eliminating a pain point. Single CTA. Phone number alongside for visitors who won't fill a form. No slider, no animation, no competing elements.
• 02 Services grouped into three pillars. Protecția Muncii (SSM/PSI), Resurse Umane, and RSVTI/Equipment - each with its own card and direct link. Users route themselves. The dropdown was replaced with a scannable grid that works as a map, not a menu.
• 03 Guarantee given its own visual section. "Plătim amenda" is pulled out of the body copy and treated as a standalone statement with its own background break. It stops the scroll. It's the moment a skeptical visitor converts from browser to lead.
• 04 Social proof positioned before the service listing. Client logos (Jouve, MERE, Paragon, Baumontage) and proof numbers placed immediately after the hero - 485 renewals, 19,300 trained, 600 inspections, 12 years. Trust established before a single service is described.
• 05 Five-stage performance audit produced with the technical team. Root causes mapped jointly with the developer and hosting architect: no CDN, no page cache, Slider Revolution adding 400 KB per page, PHP 7.4, images without WebP or correct srcset. Remediation plan sequenced by impact - Cloudflare + WP Rocket first, image pipeline second, JS cleanup third. Target agreed: LCP < 2.5 s, TTFB < 800 ms, Core Web Vitals PASSED.

Sell the model first. Then the services.

The central insight driving the design: RSGE's competitive advantage isn't any single service - it's that you never have to deal with anyone else. One contract. One team. Everything covered. The design had to sell that model before it sold anything else.

• Hero built around the proposition, not the brand"1 Furnizor / 10 Servicii Complete / 0 Griji / 0 Amenzi" runs large and heavy above the fold. Four lines, each removing a pain point. Single CTA - "Cere ofertă" - with a phone number alongside. No carousel, no animation, no distraction.
• Guarantee given its own visual moment"Plătim amenda în cazul în care compania ta va fi sancționată" - guaranteed by contract. Not a feature bullet. Its own visual break, its own section. The moment a skeptical visitor stops scrolling and starts reading.
• Social proof placed before the first service is describedClient logos - Jouve, MERE, Paragon, Baumontage, Bike Geek - plus 485 renewals, 19,300 trained, 600 inspections, 12 years. Immediately after the hero. Trust is the prerequisite for any decision.
• 10 services collapsed into 3 scannable pillarsProtecția Muncii (SSM/PSI), Resurse Umane, RSVTI/Equipment - each with its own entry point. A user arriving from Google self-identifies in seconds. No dropdown archaeology.
• Dark, anchored, structured visual toneConfident without being cold. Heavy grotesque headings for authority, open body text for compliance content users need to read, deliberate whitespace to make a service-dense site feel navigable.

The site went live. Then the real work started.

After launch, a full technical audit was run using Google PageSpeed Insights (mobile, May 2026). TTFB at 4.7 seconds, LCP at 6.2 seconds, Core Web Vitals: FAILED. The redesign had solved the design problem. This was an infrastructure problem, and it needed a different conversation. I documented the findings, brought them to the developer and hosting architect, and together we agreed on a five-stage remediation sequence ordered by impact-to-effort ratio.

• Stage 1 - TTFB and cachingNo CDN, no page cache - every request hitting the origin server cold. Fix agreed with the architect: Cloudflare free tier with Cache Everything rule plus WP Rocket. Projected TTFB drop: 4.7 s to under 600 ms for cache hits.
• Stage 2 - LCP and imagesHero image was the LCP element, served as a full-size JPG with no preload hint. Documented for the developer: convert to WebP (-50% to -70% weight), add fetchpriority="high" and preload on the LCP image, fix srcset for client logos displayed at ~150px but served full resolution.
• Stage 3 - JavaScriptSlider Revolution loading ~400 KB of JS and CSS on every page to render a static image. Recommendation to the developer: rebuild the hero in plain HTML/CSS. Additionally: defer non-critical JS, replace 176 non-composited animations using top/left/width with transform/opacity for GPU rendering.
• Stage 4 - WordPress hygieneFlagged with the hosting architect: PHP 7.4 without OPcache, no object cache, accumulated unused plugins. Upgrade path agreed: PHP 8.2, OPcache enabled, Redis Object Cache, WP-Optimize for database cleanup.
• Stage 5 - SEO and accessibilityMissing alt attributes, no LocalBusiness JSON-LD, auto-generated meta with duplicate content. Documented and handed to the team: descriptive alt on all images, structured data via Rank Math, hand-written meta per page.

Target agreed with the team: LCP < 2.5 s · TTFB < 800 ms · Performance Score ≥ 90 · Core Web Vitals: PASSED.

Cross-functional from day one.

The engagement ran across multiple teams: client managers who owned requirements and sign-off, a WordPress developer and front-end architect responsible for build and infrastructure, and a prompt engineer pairing on copy and the post-launch technical brief. My role was to hold the design direction, translate between constraints, and make sure every decision was something the team could ship and defend.

• Worked directly with the developer and architect from the IA stageNot just handing over Figma files - walking through layout decisions, flagging components that needed custom implementation, adjusting design when the technical reality required it.
• Hero went through eight iterations in two daysArchitect flagged breakpoint issues at 1024px, developer confirmed what was buildable in the theme's row system, design converged from both ends simultaneously.
• Presented and defended a recommendation to client managementThe brief asked for all 10 services prominent. The data didn't support it. Mapped the user journey against the original experience, brought it to the management team, got the 3-pillar architecture approved.
• Performance audit produced jointly with the technical teamRoot causes mapped together with the developer and architect. I structured and authored the brief; technical sequencing was validated with the people who would execute it.
• AI used as a leverage tool, not a shortcutLLMs drafted, stress-tested copy, and summarised audit data into client-readable briefs. The judgement, framing, and trade-offs stayed human.

Design delivered. Performance documented. Client equipped to execute.

RSGE's advantage was always the model - one contract, one team, everything covered. The old site made you work to understand that. The new one leads with it. From hero sequence to services architecture to the guarantee section, every design decision reinforces the same idea: you don't need anyone else.

• One-stop-shop proposition surfaced above the foldfour-line hero replaces a generic services list, single CTA throughout.
• 10 services collapsed into 3 scannable pillarsvisitors self-route in under 5 seconds, no dropdown archaeology.
• Contractual "we pay your fine" guarantee given its own momentpulled from body copy into a standalone visual break.
• Social proof front-loaded485 renewals, 19,300 trained, 600 inspections, 12 years, Jouve/MERE/Paragon logos placed immediately after the hero.
• Five-stage performance roadmap produced with the technical teamroot causes mapped jointly with developer and architect, brief sequenced by impact: TTFB < 800 ms, LCP < 2.5 s, Performance ≥ 90, CWV PASSED.
• Client equipped to execute independentlyevery recommendation paired with the specific plugin, setting, or code change required.
• Cross-functional delivery across design, engineering, and client managementdecisions made with developers, architects, and client managers, not handed to them.