IdeaWell

Project ideas from Hacker News discussions.

Why is Google still serving dodgy ads?

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

1. Profit motive / financial incentive
Many commenters argue Google tolerates scam ads because they generate revenue and removing them would cut profits.

“Google wants the most money, so they don't care to let dodgy ads thru.” – CqtGLRGcukpy
“If they reject more ads they get less money.” – IsTom

2. Legal immunity (Section 230) removes downside
Users repeatedly cite the lack of liability for hosted content as a key reason Google does not act.

“Since they're not liable for damages, they simply don't care enough to check the ads.” – itopaloglu83
“benoau: I think section 230 immunity shielding from any liability is also very motivating.” – benoau

3. Inadequate ad review / reliance on ineffective reporting
The discussion notes that Google could automate review but chooses to ignore user reports unless they reach a threshold, effectively allowing bad ads to persist.

“They would rather have people report them, then they automatically reject the report until enough people report it, then an actual person may review it.” – vintermann
“I have reported so many of those dodgy, scammy, sexualized ads on YouTube, that YouTube/Google has removed the option to report advertisements from my account.” – martin_a

4. Users resort to ad‑blocking for self‑protection
Faced with pervasive malicious ads, many commenters rely on blockers (uBlock Origin, Brave, etc.) or avoid platforms altogether.

“If a site doesn’t let me see it with my ad blocking enabled, I just go elsewhere.” – loloquwowndueo
“walrus01: It doesn't solve everything but the 'full' desktop version of the ublock origin plugin runs just fine in Android on Firefox.” – walrus01

🚀 Project Ideas

Generating project ideas…

ScamAdSpotter

Summary

  • A browser extension that uses lightweight AI models to detect scammy or deceptive ads on YouTube, Google Display Network, and other ad platforms, then automatically files reports and shows users the status of each report.
  • Core value proposition: gives users transparency and confidence that their ad‑reporting efforts are actually acted upon, reducing frustration from ignored reports.

Details

Key Value
Target Audience Everyday web users, especially those who frequently see scam ads on YouTube or news sites and want to help clean up the ecosystem
Core Feature Real‑time ad‑content scanning with an on‑device classifier (e.g., TensorFlow.js) that flags likely scams, auto‑submits a report via the platform’s API, and displays a live “report status” badge on the ad
Tech Stack Manifest V3 WebExtension (TypeScript), TensorFlow.js or ONNX Runtime Web, IndexedDB for caching, optional backend Node.js for report aggregation
Difficulty Medium
Monetization Revenue-ready: Freemium – free basic detection; $2/month for premium features like detailed report history, exportable CSV, and priority support

Notes

  • HN users lament that reporting does nothing: “stevage: Why do you keep reporting them, when clearly that does nothing?” and “snailmailman: I’ve never once seen Google reply to any of my reports”.
  • Provides tangible feedback loops that could spark discussion on ad‑platform accountability and empower users to see the impact of their reports.

AdAttribution Badge

Summary

  • A lightweight proxy/service that injects a clear, non‑intrusive badge (e.g., “Served by Google Ads”) onto every ad creative served through Google AdSense, Ad Manager, or similar networks, making the source obvious to viewers.
  • Core value proposition: addresses the attribution problem highlighted by commenters who want to know when an ad is from Google, thereby increasing pressure on the network to police its own inventory.

Details

Key Value
Target Audience Publishers, ad‑tech professionals, and privacy‑conscious end‑users who want visibility into which network serves each ad
Core Feature Intercepts ad requests/responses (via a local MITM proxy or browser extension) and overlays a semi‑transparent label on the ad iframe or native ad unit, with option to customize text/color
Tech Stack Go or Rust for high‑performance proxy, WebExtension (JS) for client‑side injection, optional WASM for image processing, Configurable via JSON rules
Difficulty Low
Monetization Hobby (open‑source) – can be hosted freely; premium hosted version offered as a service for publishers at $5/site/month

Notes

  • Commenters like “hansvm: …Add that attribution -- blame Google (or… credit Google)… a good header goes a long way toward swaying people's opinions” show demand for clear labeling.
  • Easy to deploy and discuss; could become a reference point in debates about ad‑network responsibility and transparency.

AdFraudWatch

Summary

  • A SaaS dashboard for publishers and advertisers that aggregates impression‑level data from ad networks (via provided logs or API) to identify and quantify scam/misleading ads, offering evidence for disputes and refund claims.
  • Core value proposition: turns the opaque “black box” of ad delivery into actionable analytics, helping stakeholders reclaim wasted spend and pressure networks to improve vetting.

Details

Key Value
Target Audience Medium‑size publishers, ad agencies, and brands that run display/video ads and suspect fraudulent inventory
Core Feature Ingests ad‑delivery logs (CSV, JSON, or via network‑specific APIs), runs heuristic & ML‑based scam detection, surfaces offending creatives, domains, and revenue impact, with one‑click export for chargeback requests
Tech Stack Python backend (FastAPI), Pandas/scikit‑learn for detection, React/Material‑UI frontend, PostgreSQL for storage, Docker deployment
Difficulty High
Monetization Revenue-ready: Tiered subscription – $49/month (basic), $199/month (pro), $499/month (enterprise) based on monthly impression volume

Notes

  • HN discussion notes Google’s lax oversight: “benoau: …they could easily human-review every single ad …” and “drcor: …Google does have an AI reviewer, and it also has false positives…”. AdFraudWatch gives users the data to prove such shortcomings.
  • Provides concrete utility for a technically‑savvy audience likely to discuss and adopt the tool on HN.

ScamAdBlocklist Hub

Summary

  • A community‑driven platform where users submit URLs, screenshots, or hashes of scammy ads; the community votes on validity, and approved entries are compiled into regularly updated blocklists for uBlock Origin, AdGuard, and similar blockers.
  • Core value proposition: leverages collective intelligence to keep ad‑blocking lists current against fast‑changing scam campaigns, reducing the burden on individuals to constantly maintain custom filters.

Details

Key Value
Target Audience Power users of ad blockers, privacy advocates, and anyone who wants to contribute to a cleaner ad ecosystem
Core Feature Web app for submission, moderation, and validation; generates JSON/EasyList‑compatible feeds; provides API for blocklist consumers to pull updates hourly
Tech Stack React/Next.js frontend, Node.js/Express backend, PostgreSQL for submissions & votes, GitHub Actions to auto‑publish blocklists to CDN, optional Cloudflare Workers for edge delivery
Difficulty Medium
Monetization Hobby (open‑source) – can accept donations via GitHub Sponsors; optional premium “verified reporter” badge for $1/month

Notes

  • Users like “snailmailman: When Google broke the original uBlock origin…” and “walrus01: Use ublock origin rather than ABP.” show a strong reliance on community blocklists.
  • Encourages discussion on the efficacy of crowdsourced filtering and could become a go‑to resource referenced in HN threads about ad‑blocking effectiveness.

Read Later