Dypt Review: clear planning, slower mobile (74/100)

Dypt is a web-based AI-assisted task planner for hierarchical work and personal productivity.

Fact	Detail
Domain	dypt.app
Category	AI-assisted hierarchical task planning and personal productivity
Platform	Web
Pricing	Free tier and paid Pro plan; exact prices were not provided in the crawl
Pages crawled	40
Crawl date	2026-10-20

Dypt's clearest strength is the product idea: it turns overwhelming work into smaller starts, and the site explains that mechanism quickly. Usability scored 85.7/100 and accessibility 87/100; the reviewed paths include a clear Get Started action without a credit card, sampled pages reported zero missing image alt attributes, and PSI scored 95 on the homepage and 91 on the sampled article.

The main limits are focus and speed. First impressions and positioning scored 56.9/100, while audience and messaging scored 60.9/100: dypt speaks to a broad life-and-work audience and gives limited commercial decision support. Performance scored 72.5/100. Mobile homepage LCP was 6.3 seconds, compared with 1.8 seconds on desktop.

Writing quality was a solid 76.4/100, with concrete homepage language, a credible founder voice and an ordered accounting article. Risk and stability scored 89.3/100, and technical SEO scored 84.6/100 across the 40-page crawl. The result is a fair 74/100 review: a promising planning product whose next gains depend on sharper audience definition and a faster mobile experience.

Dypt has a distinctive task-breakdown promise, but the homepage does not yet identify the primary buyer. The hero says "Too big to start? Divide it." and explains breaking work into smaller steps, while the body says "Use dypt to organise your life" (homepage). The story broadens the context to a senior leader, startup founder, parent, work, and projects, but that remains a set of use cases rather than one audience decision (story). The product mechanism is well supported by hierarchy, calendar, AI planning, sharing, nested subtasks, tags, and deadlines (homepage, feature pages). Add an immediate category anchor such as AI-assisted hierarchical task planner, then name one starting segment and explain the trade-off against a manual list or generic task manager.

Dypt explains the job quickly, but it leaves audience choice and commercial decisions broad. The homepage describes breaking tasks into smaller steps, and feature pages add nested subtasks, deadlines, tags, and AI-generated breakdowns (homepage, feature pages). The emotional job—moving from stuck to a small next action—is also clear in the hero. Starting is unusually direct: "Start free. No credit card required" (homepage). The decision gap is plan detail: the supplied evidence mentions free and premium plans and monthly AI actions for Pro, but provides no prices or limits (homepage, AI breakdown). Name one priority segment, show exact plan economics beside the start CTA, and add attributable customer evidence with a concrete workflow outcome.

A first-time visitor can understand the offer and reach signup quickly, but mobile performance delays the first interaction. The homepage pairs "Too big to start? Divide it." with "Get Started" and "Start free. No credit card required"; the CTA leads directly to /signup (homepage). Blog, Docs, Story, and Pricing are visible discovery routes. Help is less direct: the footer link leads to a page with Getting Started, documentation, import/export guidance, subscription settings, and support, but the primary navigation exposes Docs rather than Help (help). The mobile homepage records 6.3 s LCP, 9.5 s Time to Interactive, and a 5,476 KiB payload. Defer non-critical work and place Getting Started beside Docs.

Dypt has a solid sampled accessibility baseline, with a viewport restriction as the most important verified fix. The homepage, feature page, blog, help page, documentation page, and AI feature page expose header, navigation, main, and footer landmarks; sampled image checks report zero missing alt attributes. PSI reports accessibility scores of 95 for the homepage and 91 for the sampled article. The homepage and article captures use user-scalable="no" or a restrictive maximum scale, which can prevent comfortable zoom. PSI also reports insufficient contrast on the sampled article, a visible-label/name mismatch on the homepage, one non-descriptive article link, and a documentation outline beginning h2 → h1. Remove zoom restrictions, correct the confirmed contrast and heading issues, and verify controls in the accessibility tree. This crawl does not cover focus behavior or screen-reader announcements.

Dypt presents a coherent visual system, but mobile interaction details are not yet hardened. The homepage has one H1, a clear H1-to-H2/H3 outline, a dark canvas, rounded cards, blue-purple accents, and no measured horizontal overflow at 390px. The mobile audit reports 75 small tap targets, including a 40×40px menu control and 24×6px carousel buttons; several other controls are 36px high. Sampled body text and inputs are 14px, below the 16px form-control target. The stylesheet inventory contains 393 colors, 66 font sizes, 31 radii, and 83 spacing values across 435,917 CSS bytes, but those counts include library and unused rules and do not prove visible inconsistency. Remove maximum-scale=1, give controls 44×44px hit areas with separation, and set mobile inputs to at least 16px.

The mobile homepage is slowed by a heavy critical path: LCP is 6.3 s versus 1.8 s on desktop. The mobile blog article reaches 6.6 s, while the topics page reaches 2.6 s, indicating a recurring mobile issue rather than one isolated URL (PSI homepage, PSI article). The homepage lab run records 5,476 KiB total weight, 825 KiB estimated unused JavaScript, and 590 ms of render-blocking savings. Server response time is 265 ms and CLS is 0, so server tuning and layout-shift work are not the first priorities. Identify the exact LCP node and request, then split route-specific code, defer below-the-fold modules, classify the 24 S3 requests totaling 7,695 KB, and re-test the same mobile trace.

Dypt’s strongest writing is specific and human, while the feature templates often fall back to generic productivity language. "Too big to start? Divide it." gives the homepage a crisp promise, and "I'm Dane, the developer behind dypt" gives the story an attributable voice (homepage, story). The accounting article is the clearest model: it opens with a payroll check waiting on client data, then moves from blocker to action in numbered sections. Feature copy is less distinctive, using "AI" and the repeated "Unleash your potential" heading (AI breakdown). The priorities page also contains "help"/"helps" and "recieve"/"receive" errors. Replace generic promises with concrete task moments and run a focused template copy pass.

Dypt shows no supplied evidence of a broad indexing failure, but slow mobile content remains the clearest future traffic risk. The crawl reports robots.txt allowing crawling, a 200 homepage, a 404 nonexistent probe, one HTTP-to-HTTPS redirect, and raw/rendered parity across supplied pages. No sitewide noindex, redirect loop, or mass 5xx cluster is supported. Mobile lab LCP is 6.3 s on the homepage and 6.6 s on the sampled article; no CrUX field data was supplied, so this is a directional engineering signal rather than a traffic-loss diagnosis. The crawl also records /help and /tasks/7228 canonical targets that differ from their fetched URLs, while /story/ serves 200 beside /story. Confirm canonical intent, standardise the trailing-slash policy, and connect GSC and analytics before attributing any future search change.

Dypt’s editorial operation is competent but needs a stronger publishing gate. The accounting article has an author, publication and update dates, a descriptive headline, nine on-page sections, and a visible "Wrapping up" close; its BlogPosting evidence also carries word count and author/date fields (article). Feature pages show the opposite pattern: "cutting-edge AI" and repeated "Unleash your potential" weaken page-specific voice. Mechanical issues are also confirmed: a short Help title, a canonical pointing to /dypt, overlong article titles, and one duplicate description across three paginated blog URLs (help). Add canonical, pagination, title-length, and duplicate-description checks, then require one scenario and proof point for every feature release.

Dypt’s public help page is a useful entry point, but the deeper documentation surface was outside the 40-page crawl. /help links to Getting Started, data import/export, subscription management, direct support, and https://docs.dypt.app; no page from that external host was captured, so guide coverage and information architecture remain unverified (help). The captured help page has no search box or breadcrumbs, although it provides sidebar navigation. Its visible update date is 12 September 2026, while the sitemap lastmod is 23 October 2025. Keep the concise help index, add a prominent search or topic index, make the handoff to external docs explicit, and align the visible update date with sitemap metadata. The crawl also returned 404 for /llms.txt and /llms-full.txt; a concise public file could improve documentation discovery if it fits the publishing plan.

Dypt has sound crawlability and rendering foundations, with mobile performance and URL consistency as the practical priorities. The 40-page crawl reports HTTP 200 responses in the inventory, robots.txt declares https://dypt.app/sitemap.xml, the sitemap contains 83 URLs, and raw and rendered word counts match on supplied pages, including 1,775 words on the homepage and 1,953 on the sampled article. PSI measured mobile LCP at 6,314.6 ms on the homepage and 6,610 ms on the article; the homepage payload is 5,476 KiB. Confirm automated TLS renewal before 28 November 2026, when the sampled certificate is due to expire. Finally, resolve the /help and /tasks/7228 canonical differences and choose one /story trailing-slash policy. Validate structured-data coverage across key templates before making a wider claim.

Dypt works best when it shows its core mechanism: large work can be divided into smaller, deeply nested tasks, with priorities, deadlines, reminders and time estimates. Strong usability and accessibility scores support a straightforward path from the homepage to getting started.

Three fixable weaknesses hold the review at 74/100. The site speaks to a broad life-and-work audience, so it needs sharper audience selection and stronger commercial decision support. Its mobile experience also needs attention: homepage LCP was 6.3 seconds, against 1.8 seconds on desktop. Finally, the visual foundation is coherent but design execution scored 61/100, leaving room to make the product story more distinctive.

Dypt is best suited to individuals and small teams organizing work, projects and personal responsibilities, including founder and knowledge-worker use cases. It is a credible fit for people who need help turning a large task into an organized plan.

This 13-dimension review uses the supplied 40-page crawl from 2026-10-20, product facts from the current site crawl, simulated usability walkthroughs, sampled accessibility checks, performance lab results, content review and technical SEO signals. Scores and quoted metrics are limited to the public dimension table provided for this article frame.

Two dimensions were excluded: 12 · Decision-support surfaces failed, and 13 · Review-content integrity was not applicable. Google Search Console was not connected, so GSC enrichment was pending and no Search Console data was used.

Read the full SiteList methodology for the scoring framework and data notes. This article is a 13-dimension review; excluded dimensions are not represented as article sections.