Graphy Review: capable charts, slow delivery (66/100)

Graphy is a data-visualization and charting software product for internal reporting teams and developers embedding interactive charts in products.

Fact	Value
Domain	graphy.app
Category	Data visualization and charting software
Pricing	Unknown; price range not supplied
Pages crawled	37
Crawl date	2026-10-08

Graphy is a legible charting product with a coherent workflow from creation to sharing and embedding. Its strongest supplied scores are technical SEO at 86/100, risk and stability at 82/100, first impressions and positioning at 78/100, and audience and messaging at 76/100.

The review is held back by delivery and execution. Performance scores 38/100: /ai reaches its main content at 18.8 s in the mobile lab path, while the developer route reaches 9.5 s against a 2.5 s LCP target. Design execution scores 58/100, decision-support surfaces 58/100, accessibility 62/100, editorial QA 61/100, and writing quality 64/100. The result is a product that explains its category and workflow clearly, but asks mobile visitors to wait too long and leaves several presentation and evaluation details under-resolved.

Graphy is clearly a charting and data-visualization product, but its primary buyer remains implied. The homepage promises sharing interactive charts, while pricing identifies Plus as best for internal reporting. The developer surface adds a useful distinction: a config-first chart API for React with a built-in end-user editor, including a comparison with Recharts and Chart.js. That distinction should move into the main product claim. State the audience and difference together: internal reporting teams get polished, editable charts from a config-first API. This would align the internal-reporting promise with the developer story without adding more positioning language.

Graphy explains a coherent workflow from chart creation to sharing and embedding, but it presents two audiences without clearly routing either one first. Pricing makes the reporting path concrete: Free supports five charts, Plus adds unlimited charts and custom colors, and Business adds Google Sheets refresh, password-protected links, and brand kits. Developer pages present a second path for teams embedding chart capabilities in a product. These are credible entry points, not one blended audience. Put reporting teams and developers into distinct homepage routes, each with its own next action. The message will become easier to act on without removing either workflow.

A first-time visitor can identify Graphy quickly, but the supplied capture does not verify the first-chart task. The homepage says “The AI Graph Maker” and promises to turn messy data into beautiful graphs; Pricing and Blog are visible in navigation, and the newest sampled substantive post is “Vega-Lite Alternatives in 2026,” dated August 6, 2026. Yet the supplied interactive inventory captured no links, buttons, or forms on the homepage. That is a verification gap, not proof that the live product cannot be used. Validate the path from homepage through sign-up and first chart, while addressing the mobile wait: LCP is 18.8 s, TTI 21.0 s, and TBT 1,404 ms on /ai.

Graphy’s supplied Lighthouse accessibility score is 91 on /ai, but the audit identifies several concrete issues. Dialog and alertdialog elements lack accessible names; SVGs exposed with an img role lack text alternatives; and the root html element has no lang attribute across the sampled pages. CSS greps found 11 outline:none or outline:0 occurrences, only one :focus rule, and no :focus-visible rule. These findings warrant focused remediation, not a claim that every control is inaccessible. Add lang="en", name each dialog, label meaningful SVGs, and replace outline suppression with visible focus styles. Add reduced-motion handling after testing the 154 animation declarations. Empty DOM samples for landmarks and headings are not treated as absence because artifact coverage is partial.

Graphy has a coherent visual direction, but its implementation is difficult to maintain and under-sized on mobile. Screenshots show a restrained pale canvas, blue-violet product imagery, strong display typography, and a clear long-form story; the 390px render reports no horizontal overflow. The computed inventory nevertheless contains 185 colors, 31 font sizes, 26 radii, 26 font families, and 289 !important declarations. Seven mobile interactive elements miss the target: primary buttons are 250×40px and cookie controls 294×32px, while body, button, input, and link text compute to 14px. Consolidate the visual scales, raise controls to at least 44px, use 16px input text, and fix the h3-to-h5 heading skip. Mobile performance work should follow because 18.8 s LCP dominates the visual experience.

Performance is the clearest technical constraint in the mobile lab path. /ai reached main content at 18.8 s against the 2.5 s LCP target; developer/chart-api reached 9.5 s and pricing 6.6 s. The /ai server-response metric was 5 ms, so the dominant problem is post-delivery browser work: 2,914 KiB in PSI, 1,404 ms TBT, 2.4 s JavaScript bootup, and an estimated 1,324 KiB unused JavaScript. The developer network summary reached 12,125 KiB, including 6,256 KiB from playground.graphy.dev and 4,568 KiB from framerusercontent.com. Split code by route, defer playground and below-fold modules, optimize images, and enforce budgets.

Graphy’s strongest writing names the input, action, and result: “Upload your data into Graphy (Google Sheet, Excel, or even a CSV URL)” gives readers a concrete path. That standard is uneven across commercial pages. The heatmap page repeats the homepage promise, “From messy data to beautiful graphs in a click,” instead of leading with the heatmap job. Pricing has a 710-word extracted paragraph and no H1, which makes comparison harder to scan. The developer page contains useful substance—“Bar, line, pie, funnel, heatmap, and more. Pick a type, pass your data, done.”—but its heading tree jumps from H2 to H6 and H5. Preserve the practical workflow style, split dense content into paragraphs and lists, normalize headings, and give each graph page a chart-specific opening and next step.

Graphy’s pricing surface supplies plan information but does not visibly guide a visitor to the right choice. The supplied evidence does not identify audience segmentation, a defended rationale, or an easy override for recommendations. That leaves a multi-plan SaaS decision framed around plan names rather than workflow, scale, collaboration, and implementation constraints. There is useful evidence elsewhere: a seven-product Vega-Lite alternatives comparison records Graphy’s minimal code, drag-and-drop editor, AI chart generation, and storytelling strengths alongside a lower customization ceiling and no public plugin system yet. Connect those balanced distinctions to pricing. Add a Best for label, one named trade-off, and a rationale per tier, then show a visible update date. Mobile comparison behavior still needs direct verification.

Graphy has unusually useful comparative evidence because the supplied analysis records both strengths and limitations across seven products. It names polished branded charts, a drag-and-drop editor, AI generation, and storytelling, while also recording customization and plugin limitations. That balance supports fit-based guidance. The credibility gap is operational: the pricing freshness record has no lastmod or published date, and the supplied pricing evidence does not show a comparable methodology or update signal. Add a dated “last checked” label, identify whether claims are measured, vendor-stated, or editorially assessed, and carry the material trade-offs into plan guidance. The result should be a documented decision path, not a universal-winner claim.

The supplied evidence shows no sitewide indexability failure, and no traffic-drop symptom was supplied. robots.txt allows crawling, public samples return 200, HTTPS and host consolidation use single-hop 308 redirects, the soft-404 probe returns 404, and raw and rendered content align. The main boundary risk is intentional: /ai is noindexed and reaches visualize.graphy.app through two hops, while duplicate titles appear across app and signup URLs. Public acquisition pages should remain crawlable and independent of that workspace path. Template cleanup also matters: 18 pricing images and 11 developer images lack alt text, and sampled pages contain heading and metadata inconsistencies. GSC is disconnected and Wayback returned no CDX rows, so this scan cannot establish historical traffic loss or causation.

The supplied sample contains content defects a final editorial QA gate should catch. The developer page jumps from H2 sections to H6 feature labels, while pricing has no H1. The heatmap page repeats the homepage opener, “From messy data to beautiful graphs in a click,” so its first paragraph does not do enough chart-specific work. The copy-paste article is a useful benchmark because it names a problem, gives concrete steps, and uses a comparison table; its Google Sheet, Excel, and CSV URL workflow is specific and actionable. Add publication checks for one H1, sequential heading levels, structured paragraphs, distinct templated openings, and reconciled blog schema. These are sample-level editorial QA findings from the supplied pages, not a sitewide content verdict. The partial crawl does not support a sitewide AI-content or citation verdict.

The sampled technical SEO foundation is sound. robots.txt allows crawling and declares a sitemap; the nonexistent URL probe returns 404; HTTP and www consolidate to HTTPS in one 308 hop; and sampled pages show no meaningful raw-versus-rendered content gap. The main caution is the application boundary: /ai is noindexed and redirects to visualize.graphy.app through two hops. Keep that workspace out of search if intentional, but direct public discovery to a crawlable marketing URL and shorten the chain where practical. Template hygiene remains unfinished: 18 pricing images and 11 developer images lack alt text, while /manifesto has 11 H1 elements and pricing and graph pages have missing H1s. Resolve duplicate titles and descriptions and give each indexable template one descriptive H1.

Graphy works best for teams that need a clearly presented charting workflow, especially internal reporting teams and developers embedding interactive charts in products. Positioning, audience messaging, risk and stability, and technical SEO provide a sound foundation.

The fix list is focused: reduce the mobile delivery delay, tighten design execution, and make the pricing surface more useful for decisions. The supplied evidence also points to accessibility and editorial QA issues that need a deliberate final pass. At 66/100, Graphy is understandable and technically crawlable, but its public experience is not yet as efficient or polished as its product story.

This 13-dimension review uses the supplied crawl of 37 pages from 2026-10-08, the public dimension score table, and the stated site profile. The assessment covers positioning, messaging, usability, accessibility, design execution, performance, writing, decision support, content integrity, risk and stability, editorial QA, and technical SEO, with the published overall score rounded to 66/100.

No separate enrichment-pending list was supplied; Google Search Console was not connected or accessible for this review. See How SiteList scores for the review method and data conventions.