Dealsize Review: weak product, fixable friction (54/100)

Field	Evidence
Domain	dealsize.ai
Category	AI-assisted trading research and market-analysis software
Audience	Traders
Pricing	$4.99/month Retail, $25/month Desk and $99/month Institutional; 7-day free trial for Retail; query limits 25, 75 and 150.
Crawl scope	6 pages audited from the supplied corpus, 2026-10-12

Dealsize is identifiable as an AI-assisted trading product, but the homepage does not explain that category in its visible H1. The visible score table shows a mixed result: first impressions, usability, decision-support surfaces and technical SEO are the stronger dimensions, while performance is the clearest weakness at 32/100. Mobile p75 LCP is 6.17 seconds and the supplied lab run reaches 17.69 seconds. Accessibility, design execution, audience messaging, writing quality and editorial QA also remain weak.

The supplied pricing surface shows three tiers with visible monthly prices and concrete limits, but its plan differences and recommendation are not sufficiently defended. The next gains come from clearer proposition copy, more scannable pricing and content, and faster mobile delivery.

DealSize is identifiable as an AI-assisted trading product, but the homepage H1 does not say so. The visible heading is “How can I help you?”, while the interface names Trading Signals, Portfolio, Watchlist, Markets, News, Calendar and ChartDS. The title adds “Made for Traders,” and pricing describes AI-Nalysis for portfolio, news, events, strategies and instruments. Put that category and audience in the hero, then answer the pricing FAQ about differentiation with a specific supported distinction.

A trader can infer DealSize’s purpose from “Made for Traders,” Trading Signals and the market-analysis prompts, but the homepage still makes visitors solve the category question themselves. Pricing is more concrete: Retail costs $4.99/month after a 7-day free trial, Desk costs $25/month and Institutional costs $99/month, with query limits of 25, 75 and 150. Add best-for guidance for each tier and explain the trade-off behind the plans. The visible differentiation question remains unanswered in the supplied copy.

A motivated visitor can reach Pricing and News in one navigation step, but the core tasks stall before a decision or signup. The above-fold prompt, “How can I help you?”, describes an assistant interaction rather than the product proposition. Pricing exposes Log in but no verified signup path in the supplied crawl inventory; this does not establish whether signup exists outside the captured surface. News shows General News, Stocks, Forex and Crypto without a substantive article link in the supplied copy. Mobile loading adds friction: the homepage records 17.7-second LCP and 2,281ms TBT. Add a clear CTA and article-level links, then optimize the load path.

DealSize has a useful structural baseline, including lang="en", main landmarks, no positive tabindex values and observed focus-visible rules. It still fails important access checks. The homepage has one missing image alt, homepage and News repeat h1, and Calendar reports h1 → h1 → h3. PSI flags contrast, invalid ARIA relationships, unnamed buttons or links, and a viewport with user-scalable=no and maximum-scale=1. Remove the zoom restriction, restore one sequential heading tree, add the alt decision, and repair measured contrast and control names.

DealSize presents a recognizable dark trading workspace, and the 390px render has no horizontal overflow. Mobile interaction sizing is the main production weakness: the sidebar toggle is 32×20px, the search input is 356×30px, and market-card controls are 26×26px. PSI also flags contrast, button-name, link-name and ARIA relationship failures. Replace role=button market cards with native controls, provide names for icon actions, remove zoom blocking, and bring interactive targets toward 44×44px before adding visual complexity.

The homepage is slow for mobile users in both field and lab evidence. Mobile CrUX p75 LCP is 6.17 seconds, while the supplied lab run reaches 17.69 seconds LCP, 8.20 seconds Speed Index and 2.28 seconds TBT. The homepage transfers 6,887 KB across 200 requests, including 63 blocking requests; pricing transfers 6,021 KB and news 7,849 KB. Trace the hero LCP path, split and defer JavaScript, and audit third-party tags. CLS is healthy at 0.01, so render and interaction speed come first.

DealSize’s copy contains useful product and pricing vocabulary, but the page system is difficult to scan. Pricing contains 323 words in one paragraph, while “How can I help you?” repeats across home, news, pricing, chart and calendar. The brokers page jumps from H4 to H3, and three sampled pages share the title “dealsize.ai | made for traders” and one meta description. Break pricing into plan, feature and FAQ blocks, use one page-specific H1 per route, and give the brokers empty state a next action.

The pricing modal gives buyers three visible tiers and meaningful capacity differences, but it behaves more like a feature list than a buying guide. Retail, Desk and Institutional are shown, with Desk marked Recommended, yet the supplied copy does not explain which trader profile fits each plan or why Desk is recommended. Add concise Best for guidance and state the trade-off for each tier. Reduce repetitive checkbox rows to decision axes such as query volume, charts, models and connectivity, then verify yearly billing and mobile treatment.

The sample shows a mechanical QA gap rather than a lack of product material. Pricing is 323 words in one paragraph; News, Calendar and Brokers contain multiple H1 elements; three routes share title or description text; and Calendar plus chartDS/XAUUSD point canonicals to /. These are release checks with clear pass criteria. Enforce one H1 and sequential levels, unique metadata, route-appropriate canonicals, and a mobile scannability review for pricing before publishing changes.

DealSize has a workable crawl foundation: robots.txt allows the site, sampled pages return 200, HTTPS redirects from the HTTP root, and the invalid-URL probe returns a real 404. Canonical coherence is the larger issue. Calendar and chartDS/XAUUSD return 200 but canonicalize to the homepage, while /pricing/ also returns 200 beside /pricing. Choose self-canonicals or explicit non-index intent for application views, redirect the slash variant in one hop, and validate the hreflang cluster. Pricing has 141 raw words versus 327 rendered, so keep key commercial content in server-rendered HTML.

Dealsize is a starting point for traders who want an AI-assisted market-analysis surface and visible plan limits. The product's main weaknesses are fixable: the homepage needs a clearer category explanation, mobile delivery needs substantial improvement, and pricing and editorial structure need to support faster comparison. The site is best suited to traders evaluating an AI-assisted research workflow; it needs sharper explanation and faster delivery before the buying path feels dependable.

This review uses the supplied dimension score table and evidence from a crawl of 6 pages on 2026-10-12. The published score is 54/100, rounded from the persisted numeric score for article copy. The supplied profile also records that Google Search Console was not connected and access was unavailable. Read How SiteList scores for the review method and evidence model.