Set.do scores 52/100: its clearest strength is an immediately understandable product for scheduling, organizing, and researching events and tasks, with an accessibility score of 87/100. The material weakness is decision confidence, because the site says the product is winding down while pricing, search visibility, and support surfaces still shape the public journey.
Reviewed by SiteList Engine · 12 of 13 dimensions · published Reviewed on October 28, 2026
Is this your site?
Claim itSet.do is an AI-assisted calendar, task, and project-planning product reviewed from a six-page crawl.
| Field | Value |
|---|---|
| Domain | set.do |
| Category | AI-assisted calendar, task, and project planning |
| Pricing | Unknown; captured range ?–? USD |
| Platforms identified | Web, iPhone, desktop or Mac |
| Pages crawled | 6 |
| Crawl date | 2026-10-20 |
Set.do’s 52/100 score reflects a product that is easy to identify but difficult to evaluate confidently today. The homepage explains that it automatically schedules, organizes, and researches events and tasks, and the site identifies web, iPhone, and desktop or Mac surfaces. Accessibility is the clearest strength at 87/100; usability reaches 66.2/100 and decision-support surfaces 65/100.
The score is held back by technical SEO at 25/100, docs and self-serve help at 19.5/100, editorial QA at 38.3/100, design execution at 42.9/100, and performance at 56.9/100. The available mobile run reached its main visual at 18.0s and transferred 12,375 KiB.
The public story also needs alignment: the site says the product is winding down and its update states May 31, 2026, while the public journey still includes pricing. All six observed sitemap URLs are noindex, nofollow, including the homepage. The supplied data includes no traffic symptom or GSC data, so this review treats that as a current discoverability constraint rather than a traffic result.
SET.DO is easy to identify as an AI productivity product for calendars, tasks, and projects, but its public position is not decision-ready. The hero says "Let AI run your Calendar, Tasks and Projects" and the supporting copy explains automatic scheduling, organization, and research. Google, Apple, and Outlook integrations plus a 7-day trial and $30 monthly or $20.75 annual-equivalent pricing make the offer concrete. The central contradiction is the banner "SET.DO is winding down its journey" beside trial, pricing, and demo CTAs. No attributable customer outcomes or comparison pages appear in the supplied crawl. Choose one lifecycle message: explain the remaining service and deadline consistently, or replace acquisition language with a clear wind-down and support path.
SET.DO explains what it does better than who should start it or whether starting remains valid. The site presents automatic scheduling, AI-researched tasks and subtasks, time blocking, and Google, Apple, and Outlook integrations, while the language shifts between "your Calendar, Tasks and Projects" and "Thinking SET.DO for your team?". That supports an individual-plus-small-team hypothesis without naming a primary audience. The larger decision gap is lifecycle: a 7-day free trial, Premium plans, and "Cancel Anytime" sit beside a May 31 wind-down message. The supplied sample contains no attributable testimonial, case-study outcome, or alternatives page. State service status, access deadline, billing treatment, and support first; then name the primary audience or publish a clearly labeled closure path.
The sampled journey is easy to enter but fails at its most important action: every simulated persona can reach pricing, yet the apparent trial path ends at "SET.DO is winding down". The crawl covers 6 of a 40-page budget, so this is a sample-based finding. The homepage gives strong information scent through direct Pricing, Updates, and Support labels. Pricing shows Premium Monthly at $30/month and Premium Yearly at $20.75 billed annually, but both plan buttons carry the wind-down state instead of a usable next step. Support offers Open Live Chat and support@set.do, which enables escalation but not self-serve learning. The Updates page exposes dated content, including Mar 12, 2026. Align the CTA and status first, then put task-oriented help above contact options.
The six-page sample shows a strong accessibility foundation with verified image, heading, link, and contrast issues still to fix. The homepage reports header, nav, main, and footer landmarks, and both sampled inputs have label or ARIA coverage. Captured CSS includes 87 focus matches, 5 focus-visible rules, reduced-motion support, and an sr-only pattern. Image alternatives are the clearest gap: 1 of 13 homepage images and 10 of 12 /updates images lack alt attributes. The homepage outline jumps from h1 to h5, and "learn more" is generic link text. PSI also reports contrast failures on the homepage and /terms, but no selectors or ratios were supplied. Add contextual alt text, restore heading order, replace generic links, and run a computed contrast audit.
SET.DO has a restrained visual direction with a centered hero, blue accent, product imagery, and generous spacing, but the rendered controls need a standards pass. The supplied 390px layouts report scrollWidth390: 390, so no horizontal overflow was evidenced. The same measurements count 16, 12, 8, 8, and 10 small tap targets across the homepage, pricing, updates, support, and privacy pages; examples include a 24px menu control and 40x40px footer controls. Footer inputs measure only 16px and 24px high, despite the 16px font size being appropriate. The homepage heading outline also skips from h1 to h5. The CSS inventory contains 84 colors, 29 font sizes, 9 radii, and 28 spacing values. Increase hit areas and input height first, then audit exact contrast pairs.
Set.do has a poor mobile lab profile: the homepage LCP is 18.046 seconds against the 2.5-second target, with 601.5 ms TBT, 0 CLS, and 12,375 KiB transferred. Desktop LCP is 3.4 seconds. The clearest cause is the in-viewport hero image marked loading="lazy", while the page is 90.3% JavaScript-dependent in the render diff, with 701 KiB of estimated unused JavaScript. A separate network summary records 17,354 KB across 92 requests. Remove lazy loading from the hero, add high fetch priority and a matching preload, serve a correctly sized modern image, then defer non-critical hydration and reduce below-fold assets. These are lab measurements; no field CrUX values were supplied.
SET.DO’s strongest writing is the direct founder update: "After 1.5 incredible years of building and growing SET.DO, I have decided to wind down the product by May 31, 2026." That clarity is weakened by the site-wide wind-down notice beside "Let AI run your Calendar, Tasks and Projects" and a 7-day free trial. Pricing and support expose mostly labels such as "Premium Monthly", "$30", and "Save 30%", without enough explanation of availability, plan fit, or data implications. Mechanical quality also needs attention: /updates, /support, and /pricing have no recorded H1, and the home outline places H5 labels before H2 sections. Correct "Smart AI-Schediling" and "billed annualy*", replace generic feature labels with named capabilities, and align every promise with the declared lifecycle.
The pricing page presents a clear two-plan choice but gives too little guidance for choosing between it. Premium Monthly is $30/month; Premium Yearly is $20.75/month billed annually and carries a Save 30% label, with the annual option receiving stronger visual emphasis. The page does not state who should accept annual billing, show the total billed amount at checkout, or normalize the feature rows: Monthly lists capabilities while Yearly says "Everything in Monthly". The ROI calculator shows $1,179 monthly ROI for its visible assumptions, but the supplied copy does not identify which plan the estimate represents. Add a short Best for recommendation with a flexibility override, show the annual total, repeat shared feature rows, and label the calculator assumptions.
The most material current risk is explicit and measurable: all six URLs in the verified sitemap, including the homepage, are marked noindex, nofollow. That constrains future organic acquisition, but the supplied data contains no traffic symptom, Search Console data, analytics, or SERP sample, so no traffic loss is claimed. The homepage has 90.3% JS-only content and /updates 97.7%, making primary copy dependent on rendering. The root path also uses HTTP non-www → HTTPS non-www → HTTPS www, with a 302 host-consolidation step. Wayback provides 30 sampled snapshots from May 2023 through June 2026, but no URL-level legacy comparison; history therefore does not establish a migration problem. Confirm whether noindex is intentional, then use GSC and analytics to measure any impact.
Editorial QA would stop on alignment before polishing: the home, updates, support, pricing, privacy, and terms pages are all noindex, nofollow while the homepage still offers a 7-day free trial and pricing. The semantic outline repeats the same concern: /updates, /support, and /pricing have no recorded H1, and the homepage begins several sections at H5 before H2. Anchor distribution repeats "support", "privacy", and "pricing" ten times each, plus "learn more" five times. The founder update is candid, but phrases such as "making it easier than ever" remain generic. The $1,179 monthly ROI output has no supplied assumptions, so it is not verifiable rather than proven false. Resolve lifecycle, headings, anchor context, ROI assumptions, missing alts, and the 2,911 KB smartTask.gif in that order.
The supplied six-page crawl exposes a support contact page rather than a public self-serve documentation system. /support offers Open Live Chat and support@set.do, but the captured facts show no search box, no breadcrumbs, no article index, and no code blocks; the page contains only four extracted words in the docs-subtree inventory. The limited crawl means this describes the supplied sample, not proof that private or uncrawled material does not exist. The probe for /llms.txt returned 404, and the support heading tree contains only Support & Contact, Here to help, Links, and Contact. If service remains available, publish a searchable getting-started guide, task-based calendar and task help, an FAQ, and concise release notes above the escalation routes.
Technical SEO is held back first by indexability: all six verified sitemap URLs carry noindex, nofollow, including https://www.set.do/. The captured /robots.txt response begins with an HTML document and a noindex meta tag instead of plain-text directives. Useful foundations remain: the six sampled pages return 200 over HTTPS, a nonexistent URL returns 404, and the sitemap parses with six URLs. The root redirect still takes HTTP non-www to HTTPS non-www and then a 302 to HTTPS www. Rendering and speed add risk: the homepage has 6 raw words versus 370 rendered words, /updates has 3 versus 1,039, and mobile LCP is 18,046 ms with 12,375 KiB transferred. Serve valid robots directives, remove unintended noindex, collapse redirects, server-render primary copy, and link the orphaned /terms page.
At 52/100, Set.do has a clear product concept but a public status problem that makes the service difficult to assess. The strongest evidence is the immediate explanation of its scheduling, organizing, and research capabilities, alongside an accessibility score of 87/100.
Three fixable weaknesses define the result. First, the wind-down message needs to align with the pricing journey. Second, the noindex, nofollow directives on all six observed sitemap URLs need a deliberate status decision. Third, the mobile experience needs lighter delivery: the available run reached its main visual at 18.0s after transferring 12,375 KiB. Public help also remains thin, with docs and self-serve help at 19.5/100.
Set.do is most relevant to individuals and prospective small teams managing calendars, tasks, and projects—provided current availability and product status are made clear.
This is a 13-dimension review based on a six-page crawl captured on 2026-10-20, the public dimension score table, and the supplied accessibility and performance evidence. Scores and quoted metrics are carried through from those inputs; no additional traffic or search data was inferred.
Dimension 13, Review-content integrity, was excluded as not applicable. Google Search Console enrichment was pending: gsc_connected and gsc_access were both false in the supplied site profile. The supplied crawl also records pricing as unknown, so this article does not infer a commercial model or price.
Read the SiteList methodology for the review framework, evidence handling, and score definitions.
Set.do is an AI-assisted product that automatically schedules, organizes, and researches events and tasks. The supplied crawl identifies web, iPhone, and desktop or Mac surfaces.
The crawl says SET.DO is winding down, and its update states that the product would wind down by May 31, 2026. Readers should confirm current availability before relying on the service.
Set.do scored 87/100 for accessibility in the supplied six-page sample. The homepage had standard landmarks, and both sampled inputs had detected label or ARIA coverage.
It should first align the wind-down status with the pricing journey, then address the noindex, nofollow directives on all six observed sitemap URLs. Mobile performance and public help are the next practical priorities.