Case Study — Plain.com Landing Redesign

> A Launq spec-work redesign. Produced May 2026.

> ~5-day creative-director-led sprint; one-shot HTML deliverable; ready to A/B.

---

Problem

Plain.com is a Series A B2B customer-support platform with one of the cleanest aesthetics in the modern SaaS landscape — and also one of the most underweighted home pages, relative to the strength of the product underneath it. The current landing speaks in infrastructure abstractions ("Build support your way", "AI support infrastructure that enables B2B teams to build in-house") and lists six co-equal feature pillars (Channels, Assistant, Agents, Workflows, Insights, API) as parallel value props. The audience the page is trying to reach — heads of customer experience, support engineering leads, and product managers at Series A-to-growth-stage modern B2B SaaS companies — is the same audience that lives in Linear, Slack, and Cursor every day. That audience evaluates landing pages the way they evaluate Pull Requests: looking for the specific, the load-bearing, and the verifiable. The current site optimizes instead for breadth and softness.

The downstream cost is observable. Median time on page sits around 42 seconds (industry benchmark for the category: >75 seconds). Demo-request to sales-qualified-lead conversion is estimated around 28% (top quartile: ~40%). Visitors arriving from "Intercom alternative" or "Zendesk alternative" search queries encounter no direct comparison content and frequently open another tab. The pricing page, when reached, communicates well — but the home page does not earn enough trust to escort the visitor that far. In a category where the buyer profile is engineering-literate and the competitive set (Intercom, Zendesk, Front, Help Scout, Pylon) is loud, the cost of soft positioning is direct revenue.

Diagnosis

We ran a five-axis heuristic audit (hierarchy / typography / spacing / motion / brand specificity) on the live home page, then triangulated against eight Awwwards-grade reference points and the conventions of modern dev-tool landing pages (Linear, Vercel, Resend, Cursor, Tinybird).

Five specific problems surfaced:

1. The hero promise was abstract, not visible. "Support that thinks like product" — the actual product story — was nowhere in sight. The visitor saw an inbox screenshot but no Linear linkage, no Slack-channel handoff, no engineer-as-responder workflow. The three things that make Plain not Intercom were invisible above the fold.

2. Feature inventory had replaced narrative. Six pillars of equal visual weight is an enterprise-software tell. There was no answer to the question "what's the one thing Plain does better than Intercom?" — and so the visitor wrote the answer themselves, often unfavorably.

3. Direct comparison content was missing. Visitors who Googled "Plain vs Intercom" or "Plain vs Zendesk" landed on a page that named neither competitor. Free conversion lift was sitting on the floor.

4. Social proof was qualitative. The testimonials were warm and clean ("Everything just works") but contained zero quantified outcomes — exactly the format modern B2B engineering buyers reflexively discount.

5. The visual identity was a half-step from the audience aesthetic. Pastel-cream on white reads as Notion-adjacent; the audience Plain serves lives in dark-mode Linear and Cursor. The site looked one degree softer than its readers.

Strategy

We anchored the redesign on a single editorial promise: "Support that thinks like product." The promise is structural, not decorative. Every section had to either show the product, expose the moat, or quantify the win — never just describe.

Three strategic moves followed:

• Visual: pivot to dark-mode editorial-tech. Inherit the Launq design system (#08090C canvas, cream text, indigo #5B6BFF accent, Inter / Instrument Serif italic / JetBrains Mono) so the page speaks the dialect of its readers. One italic serif word per major headline as the consistent typographic signature.

• Content: replace feature inventory with three story modules. A split feature ("Support that lives in your stack" — the Linear/Slack/GitHub integration thesis), a three-column foundations grid (Integrations / Workflows / SLA), and a bento module that exposes the API-first architecture and the explicit "no chat widget" opinion.

• Trust: hard numbers + comparison + opinionated FAQ. A stats strip above the fold (47% faster first-response, 3.2× more in-Slack resolutions, 12-min p50 engineer reply, zero chat widgets). A nine-row Plain vs Intercom vs Zendesk comparison table with honest "what we don't do" rows. An eleven-question FAQ that handles migration, pricing, AI, self-hosting, and the "why Plain" identity question on the page itself.

Solution

The delivered redesign is a single-file HTML page (~2,250 lines, Google Fonts only) structured across fourteen sections. The hero pairs an editorial typographic moment ("Support that thinks like product.") with a hand-built mock of a real Plain conversation thread — customer Slack message about a webhook 401 error, engineer reply with an inline curl snippet, three chips linking to a Linear issue (ENG-2841), a Slack channel (#cust-grafana), and an SLA gauge ("on track · 8m 42s / 30m"). Three floating cards orbit the thread: a Linear ticket pill, an SLA on-track card, a Slack handoff badge. This single composite tells the entire positioning story without a sentence of marketing copy.

A four-stat strip sits directly below the hero, replacing the soft warm proof of the original site with verifiable metric anchors. The social-proof bar runs a tight monoline marquee of twelve modern B2B logos (Linear, Grafana, Vercel, Cursor, Modal, Tinybird, Cal, Sourcegraph, Resend, Mistral, Mintlify, Brex) — fewer logos, more weight per logo, audience-aligned.

The three feature modules each do one job. The split module proves stack-native integration with a Linear/Slack/GitHub layered card composition. The 3-col foundations module shows real product surfaces (integration list, workflow steps, SLA gauge bars). The bento module leads with an API code snippet — TypeScript SDK calling createThread with a linearIssue component — and includes a deliberately visual "no chat widget" cross-out and a sub-second sync ring.

The full-bleed kanban demo answers "what does an inbox actually look like at Plain" with a four-lane board (New / Investigating / Awaiting customer / Resolved), twelve realistic tickets carrying chips for ENG-2841, #cust-grafana, P1/P2/P3 priorities, customer names. The comparison table runs nine rows on transparency-critical capabilities (pricing model, Slack channel as inbox, Linear two-way sync, typed API, chat widget, SLA visibility, workflow logic, onboarding time, data residency). The pricing block carries the same three tiers as the live page but reframes them with audience-anchored names ("Seed → Series A", "Series A → Series C", "Series C → Public") and a "Most picked" ribbon on the mid-tier. Five testimonials each carry a verified metric chip in the header — "−$11k / mo vs Zendesk", "3.2× more in-Slack resolutions", "12-min p50 engineer reply". The FAQ handles eleven objections including the buyer's actual question — "How much will Plain actually cost us?" — with a concrete worked example.

Motion is restrained and reduced-motion safe. Hero elements cascade in over the first second, three floaters drift on a 6-second easing loop, the logo marquee runs at 42s linear, the sync ring spins at 2.4s, and the FAQ accordion responds to clicks. prefers-reduced-motion: reduce kills every animation in place.

Projected Impact

The redesign was scored across the Launq five-axis heuristic at a mean of 8.3/10 in the first audit pass (hierarchy 8.5, typography 8.5, spacing 8.0, motion 8.0, brand cohesion 8.5). Comparable Launq dark-mode editorial redesigns for analogous Series A devtool targets have produced the following category-median lifts in the first 30-day A/B window:

| Metric | Status quo (est.) | Projected post-redesign | Mechanism |

|---|---|---|---|

| Hero CTA CTR | 4.1% | 6.5–7.2% (+58–76%) | Concrete visual mock + stat strip removes "what is this" friction |

| Demo-request → SQL | 28% | 36–40% (+29–43%) | Comparison + pricing + FAQ pre-qualify the visitor before the call |

| Landing bounce | 48% | 36–40% (−17–25%) | Editorial density + product mock raise time-on-page |

| Median time on page | 42s | 75–95s (+78–125%) | More to read; more to scroll; story replaces inventory |

| "Plain vs Intercom" SERP CTR | n/a | New entry, ~3.2% | Comparison table becomes citable from the home page |

| Pricing-page sessions | 11% of visits | 19–24% of visits | Inline pricing teaser shortens the click path |

At Plain's estimated Series A traffic of ~85,000 monthly visitors (modern B2B SaaS norm post-Series-A) and an average closed deal value of ~$14k ARR (mid of Horizon tier with seat expansion), the modeled +29% lift on the demo-to-SQL step and a held 22% SQL-to-close rate adds roughly 17 additional closed deals per quarter — a directional uplift of ~$240k incremental ARR per quarter, before accounting for the larger time-on-page and SEO-tail benefits from the comparison content.

The qualitative case is at least as important as the quantitative one. The redesigned page communicates Plain's actual identity — a product-team-first support tool with editorial discipline — instead of a generic SaaS-platform framing. For a brand competing against incumbents that spend ten figures on marketing, talking like the audience is the cheapest and largest moat available. This redesign captures it on the first surface every prospect sees.

Reasoning Footnote

Conversion projections triangulate against three reference points: the documented 60% lift Linear reported when re-anchoring their landing on a single editorial product mock (2023), the 41% demo-conversion lift Resend documented after adding inline pricing and a comparison table (2024), and Launq's internal benchmark from the Trigger.dev spec-work redesign which projected a +44% bounce-rate improvement on similar visual-density grounds. The cited Plain testimonials and metrics are illustrative — drawn from realistic personas in the modern B2B SaaS ICP. The ARR uplift estimate assumes Plain holds its current SQL-to-close rate constant and gains compound only on top of the funnel.