AdSense is Google’s self-serve ad network: you paste one script tag, Google auctions your ad slots in real time, and you keep a share of what advertisers pay. Since the March 2024 switch to an eCPM (per-thousand-impressions) model, you earn on impressions rather than clicks, and your published revenue share is 80% of what advertisers pay after Google’s buy-side fee. In 2026 the approval bar is higher, AI Overviews have squeezed search-driven publisher traffic, and per-visitor revenue is lower than it was three years ago. Yet for a real content site, AdSense is still the lowest-friction first dollar of revenue on the web. Here is who that “real content site” actually is.
TL;DR
- AdSense pays 80% of advertiser spend on a per-impression (eCPM) basis since March 2024, not per click.
- Realistic RPM (revenue per 1,000 pageviews) for indie content sites in mid-2026: $1-$5 for general topics, $5-$15 for tech/health, $15-$40 for finance, $40-$80+ for insurance and legal — all assuming tier-1 traffic.
- Payout threshold is $100 (US); Google issues payment between the 21st and 26th of the month once you clear it.
- Apply only if you have 20+ original, in-depth articles, a domain that is 3-6 months old, steady search traffic, and no product of your own to sell.
- Skip it for SaaS landing pages, affiliate review sites, portfolios, or anything under ~20 substantive articles.
How AdSense pays in 2026
AdSense launched in 2003 as a pay-per-click network. That era is over. As of June 2026 the mechanics are:
- Per-impression (eCPM), not per-click. Since March 2024 you are paid for every 1,000 ad impressions, matching how the rest of the display-ad industry prices inventory. Clicks still happen, but your reporting and payouts are built on eCPM.
- 80% revenue share. When ads are bought through Google’s own buy-side, your effective take lands closer to ~68% after the buy-side fee; through third-party demand the sell-side share is 80%. Either way, Google keeps a cut and you keep the rest.
- Earnings depend on three levers: reader country, topic, and ad density. A US finance reader is worth many times a tier-3 entertainment reader.
Most indie content sites land between $1 and $15 RPM once you blend devices and countries. That means traffic, not the network, is the constraint — you need real volume before AdSense pays the rent.
What different niches actually pay (mid-2026, tier-1 traffic)
| Niche | Typical RPM | Notes |
|---|---|---|
| Insurance | $40-$80+ | Highest-paying category; US-focused content |
| Legal (injury, mesothelioma) | $50-$80+ | Spiky; a few keywords carry the average |
| Finance / investing | $15-$40 | Best risk-adjusted niche for new sites |
| Software / tech | $5-$15 | Steady advertiser demand |
| Health | $5-$15 | Some YMYL review friction |
| Recipes / lifestyle | $2-$6 | High volume, low CPM |
| General entertainment / gossip | $1-$4 | Plan around the floor |
Two structural facts to plan against: tier-1 traffic (US, UK, CA, AU, DE, JP) earns far more per visitor than tier-3, and Q4 (Oct-Dec) typically pays 30-50% more than Q1 because of holiday ad budgets. Do not extrapolate a December RPM into a January forecast.
The 2026 headwind you should price in
Through late 2025 and into 2026, many publishers reported sharp eCPM drops (commonly cited in the 50-70% range on some properties) as Google’s AI Overviews answered more queries directly and cut click-through to articles. This is real and worth budgeting for: AdSense in 2026 rewards sites that earn direct and brand traffic, not just thin search-intercept pages that AI now summarizes for free. Treat AdSense as one layer, never the whole plan.
Should you apply? A 60-second self-check
Apply if all of these are true:
- You run a content site with 20+ original, in-depth articles (Google now favors 20 excellent posts over 50 mediocre ones).
- Your domain is at least 3-6 months old with 90+ days of real history.
- Traffic is mostly organic search from tier-1 countries, not paid social bursts.
- You have nothing else to monetize — no product, course, or service.
- You accept that a few hundred dollars a month is the realistic ceiling until you pass ~100k monthly visits.
If you have a SaaS landing page, an affiliate review site (which usually monetizes better via commissions), a portfolio, or fewer than ~20 substantive articles, skip AdSense. It will either reject you or pay you in cents.
How to apply, step by step
- Check your traffic honestly. Open Google Analytics or Plausible. Under ~1,000 sessions/month and the math does not work yet — grow first, apply second.
- Read the AdSense Program policies once. They are short, and the rejection reasons that bite indie sites (low-value content, navigation, unclear ownership) are nearly all in there.
- Fix the table stakes: an About page with a real named person, a Contact page with a working email, a Privacy Policy that names cookies and ads, clear navigation, and zero broken pages. You must own the site and be able to edit its HTML
<head>. - Apply at adsense.google.com with your domain, then paste the verification snippet into your site’s
<head>. As of 2026, reviews run 24 hours to 4 weeks, usually 1-2 weeks. - Start small after approval. Do not drop in 8 ad units on day one. Place 2-3 thoughtful units, watch Core Web Vitals and user behavior for two weeks, then iterate. Auto ads can over-insert; cap them.
- Set the payout pipe up early. Earnings are quoted in USD and paid by EFT/wire when your balance clears the $100 threshold, with payment issued the 21st-26th of the month. Add your tax form and a USD-capable payment method before you cross the line.
Common pitfalls
- Applying with a handful of thin articles. 2026 reviews actively reject AI-spun, recycled, or shallow content. Depth beats count.
- Assuming revenue scales linearly with traffic. Blended RPM moves with your topic and country mix, not just pageviews. Adding tier-3 traffic can lower your average even as totals rise.
- Stacking ads on a SaaS site. Display ads cheapen the brand and split conversion attention. If you have a product, pick the product.
- Confusing approval with revenue. Approval only means you can show ads. Whether they pay depends on audience and placement.
- Letting a “paid” balance stall. If your balance crosses $100 but the payout never moves, it is almost always a tax form, address, or PIN-verification gap, not an AdSense bug (why a “paid” balance gets stuck).
FAQ
- How much does AdSense actually pay per 1,000 pageviews in 2026?: For typical indie content sites, $1-$15 blended RPM. The low end is general/entertainment content with tier-3 traffic; the high end is finance, software, or B2B with tier-1 traffic, and insurance/legal can clear $40-$80+. Plan around $3-$5 unless your own data says otherwise. Desktop and mobile often diverge 2-3x; when they do, the cause is usually placement, not the network (why mobile vs desktop fill rates diverge).
- Does AdSense still pay per click?: No. Since March 2024 it pays on an eCPM (per-1,000-impressions) basis, with publishers keeping 80% of advertiser spend after the buy-side fee. Clicks still influence advertiser bids, but your payout is impression-based.
- Can I get rejected and try again?: Yes. Most rejections cite “low-value content,” “navigation issues,” or “site ownership unclear.” Fix the named issue and reapply — a fresh-application rejection is not a permanent ban.
- Do I need an LLC or business entity?: No. Individuals can apply with a personal name and address (you must be 18+). You will submit a tax form (W-9 in the US, W-8BEN elsewhere) before payout, but it is a personal-name form.
- What pays better than AdSense if I qualify?: Ezoic accepts very small sites (it works from roughly 3,000 monthly visits) and usually beats AdSense once you have steady traffic. Mediavine requires 50,000 monthly sessions (its lighter “Journey” tier lowers this to ~10,000) and pays more per visitor at scale. Both demand higher content quality than AdSense.
- Will AI Overviews kill AdSense income?: They have already cut search-intercept traffic for many publishers. The sites that hold up earn direct and brand traffic and answer questions AI cannot fully summarize. Build for a reader who comes back, not just a one-click search visit.