Make-it Review: promising platform, slow mobile experience (68/100)

Make-it.ai is an AI-powered platform for building smart garden gadgets.

Fact	Detail
Domain	make-it.ai
Category	AI-powered smart garden gadget design and building
Pricing	Supplied site price range: 9.99–29.99 USD
Pages crawled	38
Crawl date	2026-09-23
Overall score	68/100

Make-it.ai earns 68/100 because its product story is clear, its intended audience is legible, and its visual system is handled with care. The site presents an AI-powered platform for building smart garden gadgets and speaks directly to DIYers with no electronics experience. First impressions, usability, design execution, and technical SEO are among its stronger areas.

The largest weakness is mobile performance: the homepage records a 14.9-second LCP, with the summary attributing the delay to a large, unoptimized hero image and additional script weight. Writing quality is uneven. The project library provides specific component names and outcomes, but the broader copy has structural weaknesses. The comparison tool lists 12 axes without enough decision support, and review content lacks affiliate disclosure, methodology blocks, and drawbacks. The site is broadly crawlable and render-stable, but duplicate host architecture remains a resilience concern.

Make-it clearly presents an AI-powered platform for building smart garden gadgets, led by “Your garden, made smarter.” The “Text to Device” concept differentiates the offer, but the audience remains an implied “Functional DIYer” segment rather than a named role or vertical. Proof is thin: the site cites “10k+ makers building,” yet the supplied evidence contains no named customers or outcome-led case studies. Name the priority audience in the hero, support the category claim with customer names and measurable results, and keep “Make-it” and “AI Garden Gadget Builder” in a consistent brand hierarchy.

Make-it speaks effectively to DIYers without electronics experience, and its language fits a beginner mental model. The cost question is answered with “Start free. Every account gets 2 free projects,” but the supplied evidence does not explain credits or the upgrade path. Credibility also weakens because “10k+ makers building” is not paired with named customers or measurable outcomes. Explain credits and paid tiers where the free offer is introduced, then add attributed customer evidence so the message answers both “Can I use this?” and “Why trust it?”

The homepage gives visitors a direct value path through “Your garden, made smarter.” and “Show us your garden,” but the pricing page leaves the key choice under-explained. Packs have prices, yet features and differentiators are not compared visually, while the “80% less than retail” claim is buried. Labels such as “Marketplace,” “Guides,” and “Blog” may require interpretation from visitors unfamiliar with electronics. Add a side-by-side pricing comparison, explain what “Start Building Free” includes, use benefit-led navigation labels, and standardize the primary button treatment across mobile and desktop.

Make-it’s visual system is coherent: sampled pages use 64px+ section padding, one radius convention, one H1, and a clear scan order. The token system uses custom properties for colors, radii, and shadows. Two accessibility failures are direct and fixable: body text and form placeholders using #9CA3AF on white compute to 2.8:1, below the 4.5:1 requirement. Increase the contrast of body text and form placeholders to meet the 4.5:1 WCAG AA requirement. The mobile pricing “Buy Now” buttons are 20px high; raise them to 44px so the visual system also supports reliable touch interaction.

Mobile performance is the clearest technical barrier: the homepage’s largest contentful paint is 14.9 seconds, driven by a 2.0 MB hero image rendered at 368px despite a natural size of 1344×768. The image has no preload or fetchpriority signal. Third-party requests add weight, including Stripe at 300KB, fonts at 263KB, and Google Tag Manager at 172KB. The audit also reports 3,910KB of unused JavaScript savings potential and 2.1 seconds of bootup time. Preload the critical image, defer non-critical scripts, remove unused JavaScript, and add explicit image dimensions.

Make-it’s project library earns trust through concrete detail: descriptions name parts such as Arduino Nano Every and capacitive soil sensors, explain mechanisms, and state outcomes. The broader structure undercuts that strength. The supplied stats report paragraph_count: 1 across sampled pages, including an 829-word homepage and a 2,096-word Guides index. The About page also uses phrases such as “robust, scalable platform” and “mission critical SaaS infrastructure,” while capitalization alternates between Make-it and Make-It. Break copy into semantic paragraphs, replace generic corporate phrasing with concrete outcomes, and standardize the name site-wide.

The “Arduino vs ESP32 vs Raspberry Pi” comparison gives visitors 12 axes, but it does not turn those facts into a recommendation. Four rows—Price, Type, WiFi/Bluetooth, and Programming language—are all-checkmark rows, so they provide no meaningful difference. The page says “It depends on what you want to build” without segmenting advice by user need. Reorder the table around learning curve and cost, add recommendations for beginners, WiFi-enabled IoT work, and software-heavy projects, replace checkmarks with values such as price ranges, and validate the stacked table at a 320px viewport.

The Arduino, ESP32, and Raspberry Pi comparison does not give readers the context needed to judge its recommendations. The supplied evidence finds no affiliate disclosure, no “how we chose” or criteria block, and no named drawback for any board. Its conclusion—“They’re all right. And they’re all wrong.”—does not explain where each option falls short. Add a disclosure near the top, state the comparison criteria, and give each board one explicit drawback. These additions help readers distinguish commercial relationships, selection method, and tradeoffs before acting on the advice.

The supplied evidence does not show a sitewide suppression mechanism: public pages return 200, the soft-404 probe returned 404, HTTPS is consolidated, and raw and rendered content match across sampled templates. The material resilience issue is duplicate host architecture. Ten sampled garden.make-it.ai URLs return 200 while canonicalizing to make-it.ai equivalents, and duplicate titles and descriptions appear across host variants. Choose one retained host, redirect duplicate routes in one hop, update internal links and sitemap signals, and re-crawl. Search Console is disconnected, so the crawl cannot quantify traffic loss or click changes.

Make-it has a sound crawl baseline: robots.txt permits public content and declares the sitemap, the nonexistent probe returns 404, HTTP and www consolidate to HTTPS in one 301 hop, and sampled raw and rendered word counts match. The main issue is the secondary host: 10 garden.make-it.ai pages return 200 while pointing canonicals to make-it.ai. The sitemap lists 96 URLs, but the crawl covered 38 of a 40-page budget, so remaining URLs are unverified. Consolidate the public host, make retained URLs self-canonical, complete the sitemap pass, and resize the repeated 1,987KB homepage image.

Make-it is positioned for people who want guided smart-garden gadget building without prior electronics experience. The site communicates its core proposition clearly, gives the homepage a prominent CTA, and maintains strong visual consistency.

Two fixes should lead the next pass. First, reduce the homepage's 14.9-second mobile LCP. Second, turn the 12-axis comparison tool into a buying aid with clearer priorities and recommendations. For the Arduino/ESP32/Raspberry Pi comparison reviewed, the site should also strengthen review-content transparency with affiliate disclosure, methodology blocks, and drawbacks. The site is aimed at DIYers, including homeowners, small business owners, and farmers. The next priorities are to reduce LCP and add audience-specific recommendations to the comparison.

This is a 13-dimension review based on a crawl of 38 pages completed on 2026-09-23. The overall score is 68/100, rounded from the persisted numeric score. The review combines the public dimension score table with crawl evidence and site facts, including the supplied site price range of 9.99–29.99 USD.

Google Search Console was not connected, so no Search Console enrichment was used.

Read How SiteList scores for the review method and dimension definitions.