YouTube CPM Rates in 2025: How Location Shapes Earnings

YouTube advertising payouts in 2025 look very different depending on where viewers are watching. Median CPM (cost per mille, or cost per thousand impressions) figures from 50 countries put the global midpoint at $2.91 per thousand ad views. The United States comes in at $11.95, the highest in the dataset. Pakistan is at the other end of the scale with $0.42. Geography still plays a big part in deciding what a creator can make from their audience.

Where the Highest and Lowest Rates Fall

Top earnings are concentrated in North America, Western Europe, and parts of Oceania. The United States sits in first place. Australia follows at $8.93, then Norway at $8.19. Switzerland, the United Kingdom, and Denmark each post more than $7.00, suggesting strong advertiser competition in those markets.

Several Asian and African countries remain well below the global average. Pakistan, Bangladesh, and Egypt all report less than $0.55. India, Vietnam, and Indonesia stay under $1.00. These figures reflect smaller ad budgets, lower spending power, and frequent use of ad blockers.

Latin America and Eastern Europe are more mixed. Mexico and Brazil record between $1.30 and $1.64. Poland and Czechia are near the midpoint. Romania, Croatia, and Bulgaria fall in the $1.75 to $1.94 range, showing steady but limited advertiser activity.

Country Name	CPM median (USD)
United States of America	$11.95
Australia	$8.93
Norway	$8.19
Switzerland	$8.02
United Kingdom	$7.60
Denmark	$7.43
New Zealand	$6.72
Canada	$6.65
Belgium	$6.52
Netherlands	$6.44
Germany	$6.43
Sweden	$6.30
Austria	$4.95
Finland	$4.77
France	$4.54
Ireland	$4.29
Singapore	$3.62
Italy	$3.59
Japan	$3.41
Hong Kong	$3.31
Czechia	$3.20
Spain	$3.14
United Arab Emirates	$3.03
Poland	$2.94
Israel	$2.91
Republic of Korea	$2.91
Portugal	$2.78
Greece	$2.33
Saudi Arabia	$2.18
Hungary	$2.10
NaN	$2.08
Croatia	$1.94
Romania	$1.94
Bulgaria	$1.75
South Africa	$1.69
Mexico	$1.64
Brazil	$1.33
Malaysia	$1.31
Serbia	$1.28
Philippines	$1.22
Thailand	$1.22
India	$0.96
Turkey	$0.87
Indonesia	$0.87
Vietnam	$0.82
Sri Lanka	$0.63
Nepal	$0.58
Egypt	$0.53
Bangladesh	$0.52
Pakistan	$0.42

How the Numbers Were Collected

The data, complied by IsThisChannelMonetized, comes from YouTube analytics on videos uploaded during 2024. Countries with fewer than 1,000 views were left out. Any entries showing no revenue or no views were removed. Median values were used to avoid skew from unusual spikes or drops, giving a more stable picture of the market.

CPM and RPM in Brief

CPM means the cost advertisers pay for every 1,000 ad impressions. RPM is what the creator actually earns for the same number of views after YouTube’s 45 percent cut. CPM will always be higher. Both figures depend on who’s watching, where they’re located, the topic of the content, and how many advertisers are bidding in that niche.

Rates by Content Type

The subject of a video can have as much impact on income as the viewer’s location. Based on USD estimates:

• Shorts: $0.02 to $0.15 CPM
• Entertainment and lifestyle: $1.36 to $1.82 CPM, RPM under $1.00
• Gaming: $4.55 CPM, $2.50 RPM
• Education: $9.09 CPM, $5.00 RPM
• Finance and digital marketing: $14.55 to $36.36 CPM, with higher RPMs

Why Some Views Earn Nothing

Not all views lead to revenue. Ads might be blocked, rights holders could claim the income, or YouTube may run ads without sharing earnings if a channel is outside the Partner Program.

Ways to Lift RPM

Creators can’t control every factor, but they can:

• Make videos longer than eight minutes to allow mid-roll ads
• Target audiences in countries with stronger CPM rates
• Focus on topics that attract high-value advertisers
• Keep viewers watching for longer periods

The Takeaway

The 2025 CPM rankings show a widening gap between mature ad markets and emerging economies. Where an audience lives and the kind of content produced remain the biggest drivers of revenue. Growth depends on consistent quality and keeping viewers engaged over time.

Notes: This post was edited/created using GenAI tools.

Read next: Office, Retail, and Support Jobs Projected to Shrink by 2030

by Irfan Ahmad via Digital Information World

Office, Retail, and Support Jobs Projected to Shrink by 2030

Employer forecasts point to a sharp drop in demand for a wide range of traditional roles over the next five years, with clerical positions expected to see the largest cuts.

The figures come from a global survey of more than a thousand companies employing over 14 million people, compiled in the World Economic Forum’s latest jobs report. It ranks the occupations most likely to lose ground between 2025 and 2030.

Postal clerks face the steepest decline, with projections showing a fall of around 40 percent. Bank tellers follow closely with a 35 percent drop, while data entry specialists are expected to contract by roughly a third. These jobs rely heavily on repetitive processing tasks, which employers are increasingly assigning to automated systems.

Retail and administrative work is also under strain. Cashiers, ticket clerks, and administrative assistants all feature high on the list, with estimated declines ranging from about 28 to 30 percent. Bookkeeping and payroll clerks, along with material-recording staff such as stock keepers, could see reductions of more than one-fifth. Transportation attendants and conductors appear just behind them with a 21 percent forecast fall.

Some occupations tied to customer outreach and sales are also projected to shrink. Door-to-door sellers, street vendors, and similar roles are expected to drop by about 20 percent. Graphic designers sit at the same level, while claims adjusters, legal officials, and legal secretaries are each predicted to see losses between 17 and 19 percent. Telemarketing is projected to fall by around 16 percent.

Technology is a driving force behind many of these changes, though real-world results have not always matched early promises. Some systems promoted as fully automated have required significant human oversight to work reliably, and certain AI-driven services have been scaled back when they failed to handle practical challenges. Even so, most employers in the survey anticipate steady advances in automation, pushing them to reshape their staffing plans.

The report suggests that by 2030, workplaces will be more dependent on technology for routine operations, while human staff will be concentrated in areas where adaptability, problem-solving, and complex interaction are essential.

Read next: Fraud Losses Among Seniors See Steepest Increase in Four Years
by Asim BN via Digital Information World

Palestinians in Gaza Struggle to Keep Fundraising Accounts Alive on Bluesky

Palestinians in Gaza who depend on online fundraising to get food, medicine, or other essentials are finding it harder to keep their accounts active. Bluesky, a social media platform that many use to reach potential donors, has been removing profiles linked to these campaigns within days of their creation. For people relying on the money to meet daily needs, losing an account means starting over almost immediately.

Some have opened dozens of profiles over a few months. Each time one is taken down, they rebuild from scratch, trying to reconnect with supporters who might have seen their earlier posts. That often means following large numbers of people in a short period or tagging past contacts in new posts, which makes the accounts look automated to Bluesky’s systems and increases the risk of another removal, as spotted by TheGuardian.

Grassroots Checks on Genuine Campaigns

In May last year, a volunteer based in Germany began informally confirming that certain Gaza-based fundraisers were real. The process started after she helped a contact raise money successfully and then began hearing from others in similar situations. Over time she kept a list of verified accounts, each allowed to use a small badge in their profile and posts to signal that someone had checked their identity.

Her vetting can involve a video call, a referral from someone she already trusts, or documents proving the person’s location and circumstances. While this reassurance helps donors feel more confident, it does not stop Bluesky’s automated filters from removing accounts. Many people still wait weeks for verification, and not everyone is approved.

Daily Costs and Scarce Supplies

The need for these campaigns is immediate. Aid groups say food shortages in Gaza have reached extreme levels, with families unable to secure even basic items. When flour, milk, or diapers are available, prices are far higher than before the blockade. A single day’s donations can make the difference between someone eating or going hungry.

• Also read: Musk Scores Lowest U.S. Favorability in Gallup Poll, Netanyahu Confronts Gaza Genocide Allegations

Because payment processors do not operate in Gaza, fundraisers must rely on intermediaries abroad to run the campaigns and send the money. This requires complete trust in strangers, and some have lost significant amounts when a campaign host withheld part of the funds.

Pressure for Platform Changes

Similar verification groups exist on other platforms such as Instagram, X, and Tumblr. Yet even with confirmation from these networks, Bluesky’s systems often flag genuine campaigns as spam. More than 7,000 users have signed open letters urging the company to adjust its moderation rules for fundraising accounts linked to Gaza.

Bluesky has said it wants users in Gaza to be heard while following its guidelines. Those working with fundraisers report that appeals against account closures rarely get responses, and many profiles vanish within a few days of being set up. For people trying to raise enough to buy food for tomorrow, the loss of an account is not only a technical setback but also a break in the fragile connection to donors who might keep them alive.

Notes: This post was edited/created using GenAI tools. Image: DIW-Aigen.

Read next: 

• Where Are The Youngest And Oldest Business Owners In The US? [Infographic]

• Why Many Computer Users Are Better at Spotting Malware Than They Think

• Data Shows Where Americans Spend the Most Time Connected
by Irfan Ahmad via Digital Information World

Data Shows Where Americans Spend the Most Time Connected

A new set of figures comparing internet access and online habits has placed Utah in first place nationally, showing the state ahead on several measures of connectivity. The results come from an analysis of federal and search engine data that ranked all fifty states according to their access to high-speed internet and the volume of social media activity from their residents.

Researchers drew on information from the Federal Communications Commission, the U.S. Census Bureau, and Google’s search statistics, combining the findings into a score out of ten. They measured the proportion of households with broadband connections, the share that could use gigabit-level speeds, and how often people searched for social media platforms online, adjusting the search numbers for population size.

Utah’s Numbers

Utah’s final score was 8.7. Almost 93 percent of households there have broadband access, and more than 72 percent are connected to gigabit internet. Search behaviour also pushed its ranking higher, with nearly 119,000 social media-related queries for every 100,000 residents in a typical month.

Who Followed

New Hampshire came in second on the list, with an 8.3 score, while Connecticut and Washington tied for third at 8.2 each. Washington stood out for having the country’s highest broadband subscription rate at 93 percent, while Connecticut led in gigabit coverage at just over 81 percent. Maryland ranked fifth with 8.1, and Oregon followed closely at 8.0. Oregon also claimed the highest search rate in the nation at more than 130,000 social media queries per 100,000 people each month.

Colorado earned 7.8 and shared seventh position with Nevada and Massachusetts, the latter showing search numbers above 126,000 per 100,000 residents. New Jersey was next at 7.6, California scored 7.5, and Georgia rounded out the top ten with 7.3.

Strong Metrics, Lower Rankings

Some states performed well in one measure but not across the board. North Dakota, for example, recorded the highest proportion of gigabit connections in the U.S., almost 85 percent, yet its overall score was 19th due to lower online search activity. Kentucky, with gigabit in 73 percent of homes, still ranked 36th overall.

The Other End of the Scale

Mississippi sat at the bottom of the list with a score of 2.2. Broadband subscriptions there reach only 81.9 percent of households, and search activity for social media platforms is the lowest recorded, at about 81,500 monthly queries per 100,000 people. West Virginia followed with 2.6, Louisiana scored 3.2, and both New Mexico and Alaska came in at 3.4. Alaska’s result reflects limited high-speed availability, with gigabit connections in fewer than eight percent of households.

What the Data Shows

The gap between the most and least connected states is wide. Places with stronger infrastructure tend to see higher use of social media, while those with slower or less reliable access show lower engagement. The figures suggest that where high-speed connections are common, online activity becomes a more regular part of daily life.

H/T: Flipsnack.

Read next: Why Many Computer Users Are Better at Spotting Malware Than They Think
by Irfan Ahmad via Digital Information World

Musk Scores Lowest U.S. Favorability in Gallup Poll, Netanyahu Confronts Gaza Genocide Allegations

Elon Reeve Musk, chief executive of Tesla, Inc. and SpaceX, ranks as the least liked public figure in the latest Gallup survey. He has a favorable rating of 33 percent and an unfavorable rating of 61 percent. Only 6 percent express no opinion. His net score is minus 28 percentage points. Public opinion has shifted after his political endorsements during the MAGA era, weaker growth in Tesla’s earnings, and a series of high-profile controversies.

Benjamin Netanyahu, prime minister of Israel, follows closely with a net favorability of minus 23 points. Twenty-nine percent rate him positively, while 52 percent rate him negatively, and 19 percent have no opinion. His decline in U.S. public opinion coincides with the ongoing genocide in Gaza, criticism of military tactics, and stalled cease-fire negotiations.

• Also read: Israel Recorded Millions of Palestinian Calls and Stored Them on Microsoft Cloud Without Consent, Raising Surveillance and Human Rights Concerns

Mid-Range Negative Scores

Donald Trump, former U.S. president, records a net score of minus 16 points, with 41 percent favorable and 57 percent unfavorable. Marco Rubio, U.S. senator from Florida, matches that score with 31 percent favorable and 47 percent unfavorable, though he has a higher share of respondents with no opinion at 22 percent.

Joe Biden, president of the United States, JD Vance, senator from Ohio, and Gavin Newsom, governor of California, each hold a net rating of minus 11 points. Biden’s numbers are 43 percent favorable and 54 percent unfavorable, with 3 percent offering no opinion. Vance has 38 percent favorable and 49 percent unfavorable, with 13 percent expressing no view. Newsom’s figures are 30 percent favorable, 41 percent unfavorable, and 28 percent no opinion.

Pete Hegseth, television host and commentator, sits at minus 12 points, with 26 percent favorable and 38 percent unfavorable. Thirty-seven percent have no opinion of him.

Near-Neutral Ratings

Robert F. Kennedy Jr., independent presidential candidate and environmental lawyer, is at minus 5 points, with 42 percent favorable and 47 percent unfavorable. Alexandria Ocasio-Cortez, U.S. representative from New York, is at minus 4 points, with 34 percent favorable and 38 percent unfavorable. Emmanuel Macron, president of France, is at minus 1 point, with 30 percent favorable and 31 percent unfavorable.

Positive Rankings

Bernie Sanders, senator from Vermont, holds a net score of plus 11 points, with 49 percent favorable and 38 percent unfavorable. Ukrainian president Volodymyr Zelenskyy is higher at plus 18 points, with 52 percent favorable and 34 percent unfavorable.

At the top of the list is Pope Leo XIV with a net favorability of plus 46 points. He has 57 percent favorable, 11 percent unfavorable, and 31 percent no opinion. His broad appeal reflects support across political lines and recognition for his humanitarian role.

Notes: This post was edited/created using GenAI tools. 

Read next: AI Language Models Show Brain-Like Understanding of Visual Scenes
by Arooj Ahmed via Digital Information World

AI Language Models Show Brain-Like Understanding of Visual Scenes

A group of researchers has shown that advanced language models can reflect how the human brain interprets what it sees. The work used detailed brain scans and computer models to study how people understand complex, real-world scenes. It combined neuroscience with machine learning to find patterns linking brain activity to artificial intelligence outputs.

Participants viewed thousands of images while lying in a 7-Tesla MRI scanner. The images came from a public photo database and included varied situations such as street views, people at work, animals in their habitats, and objects in familiar places. Each picture had several human-written captions describing its content.

Matching AI to Brain Signals

The captions were processed through MPNet, a large language model designed to turn sentences into compact numerical representations called embeddings. These embeddings were compared with brain activity patterns using a method known as representational similarity analysis. In higher-level visual areas, the AI-generated patterns aligned closely with the human brain’s responses.

A second test used the brain data to predict the language model embeddings, then matched these predictions to a large library of captions. This allowed the researchers to reconstruct short textual descriptions of the images people had seen.

Importance of Full Context

Models that used only lists of objects, single words, or limited parts of speech showed weaker alignment with brain activity. Embeddings created from full sentences performed best, suggesting that combining all the information in a caption is important for reflecting how the brain processes meaning.

Building Vision Models from Language

The team also trained recurrent convolutional neural networks to take images as input and predict the language model embeddings for their captions. These networks matched brain responses more closely than many top computer vision systems, despite being trained on far fewer images. When directly compared to otherwise identical networks trained only to classify objects, the language-trained versions produced richer internal representations that explained brain activity better.

Interpreting the Results

The results suggest that the brain may convert visual information into a high-dimensional form similar to how a language model encodes text meaning. This process appears to keep both object details and the wider relationships between elements in a scene.

Future Use

The researchers say this approach could help design AI systems that see and interpret more like people do. It may also lead to improvements in brain–computer interfaces and visual aids for people with sight loss. The work offers a possible common framework for studying complex meaning in the brain by connecting insights from vision research, computational modelling, and language processing.

Read next: OpenAI launches GPT-5, a unified stack that blends fast chat and deep reasoning

by Asim BN via Digital Information World

OpenAI launches GPT-5, a unified stack that blends fast chat and deep reasoning

OpenAI released GPT-5 on August 7, 2025. The company says the new family of models combines quick conversational replies with heavier reasoning in one system. The rollout moves ChatGPT away from a menu of separate models and toward a single, adaptive experience for millions of users.

A single family, four sizes

GPT-5 arrives as a family rather than a single product. The lineup includes the full-capability GPT-5, a higher-capacity GPT-5 Pro for extended reasoning, and two smaller options called GPT-5 Mini and GPT-5 Nano. Mini and Nano trade some depth for speed and lower cost. Pro adds parallel compute during inference so it can handle longer, multi-step problems.

OpenAI is making GPT-5 the default model in ChatGPT for free and paid accounts. Free users will see usage limits and may be routed to Mini when they hit caps. Plus subscribers get higher limits. Pro subscribers at $200 per month receive unlimited or prioritized access to GPT-5 Pro. Team, Enterprise and Edu customers will gain broader Pro access in the days after launch.

Automatic routing, fewer choices

ChatGPT now decides which version of GPT-5 to use for each request. A routing system chooses quick replies or the deeper “thinking” mode when tasks need more processing. The aim is to remove the friction of asking users to pick a model for each task. The chat interface also offers four preset conversational styles, labeled to help users set tone without extra prompts.

On the chat platform OpenAI will phase out older selectable models for most users. Enterprise API customers can still call legacy models for now. That gives developers time to update systems that relied on prior model behavior.

How it performs on tests

OpenAI shared benchmark results showing gains in coding and complex reasoning. On SWE-Bench Verified, which tests real-world coding tasks, GPT-5 scored 74.9% on first attempts. That placed it ahead of some recent competitors on that metric. A GPT-5 Pro configuration with tools reached about 42% on Humanity’s Last Exam, a difficult composite reasoning test. On a PhD-level science benchmark, GPQA Diamond, the Pro setup recorded near-top performance. Health-focused tests also showed much lower rates of fabricated answers when the model used its thinking mode.

Benchmarks are mixed across workloads. In simulated web navigation tasks, GPT-5 posted strong results in some sections and weaker results in others. Those variations suggest the system gains are real but not uniform.

Fewer fabrications and safer responses

OpenAI reports that GPT-5 reduces the rate of fabricated answers compared with recent reasoning models. When thinking mode is enabled, hallucinations decline substantially compared with earlier versions. The company also introduced safer completion behavior so the system explains limits and avoids abrupt refusals. The model is designed to reject clearly malicious requests more reliably while permitting benign or borderline queries to receive useful guidance within safety boundaries.

New developer controls and bigger context

The GPT-5 API brings controls that let developers tune speed and depth. Teams can set reasoning effort levels to choose between rapid responses and deeper analysis. A verbosity option sets brief, standard or detailed reply lengths. Tools can accept free-form strings where appropriate. Developers can also constrain outputs with grammar or regular expressions and enable brief preambles that describe planned tool calls before they run.

OpenAI lists base API prices at $1.25 per million input tokens and $10 per million output tokens for the main tier, with lower unit costs for Mini and Nano tiers. The context window spans 256,000 tokens, allowing much larger documents and longer conversations without aggressive chunking.

What this is not

GPT-5 advances reasoning and task automation, but it does not meet broad definitions of artificial general intelligence. The model does not learn continuously while deployed and it lacks persistent autonomy. Those limitations mean GPT-5 is a significant capability upgrade rather than a shift to fully autonomous intelligence.

Early reactions and migration issues

Early enterprise testers report faster development cycles, more reliable code generation, and cleaner design choices in generated interfaces. Several developer tools and platforms are already running trials. At the same time some ChatGPT users are upset that they can no longer pick older models they had tuned to specific workflows. OpenAI says legacy models remain available via the API for now, which should ease migration for businesses.

An open and commercial approach

The GPT-5 launch comes alongside an open-weight release called gpt-oss intended for local use and customization. That creates a two-track approach: a more capable, centralized family available inside ChatGPT and a free open model for offline experimentation and lower-cost deployments.

Bottom line

GPT-5 packages multiple performance tiers into a single user experience and adds developer tools to tune cost, latency and accuracy. It improves coding and multi-step reasoning and lowers fabrication rates in many settings. The release simplifies the chat surface for most users while keeping options open for developers who need older models or different trade-offs.

Notes: This post was edited/created using GenAI tools.

Read next: 

• Study Tests Threat-Based Prompts, Finds Results Unpredictable and Mostly Ineffective

• Where Are The Youngest And Oldest Business Owners In The US? [Infographic]

by Irfan Ahmad via Digital Information World