For the question “how much can you realistically earn from an AI niche website in 2026”, the defensible answer is anywhere from $0 to several thousand dollars a month, but neither the typical result nor a guaranteed timeframe can be established from reliable population-wide data. A site receiving 20,000 pageviews with a modelled $5–$20 page RPM would make only $100–$400 a month from display ads. That is the arithmetic behind the headline, not a prediction that a new site will attract those visitors.
The old business model was straightforward: publish answers, appear in search, earn a click, then sell advertising or refer a buyer. AI now lowers drafting costs, while AI answers in search can reduce the number of visitors arriving to read the draft. That simultaneous reduction in production costs and potential distribution makes the 2026 economics more difficult to explain through simple “posts multiplied by traffic” projections.
This guide separates five things often confused in niche-site income reports: potential search demand, captured visits, monetisable pageviews, recognised revenue and profit after costs. It then adds a sixth: the value of an operator’s unpaid hours. Rather than claiming a particular monthly income is normal, it offers calculations that readers can reproduce with their own Search Console, analytics and affiliate dashboards.
All dollar amounts in the forecast tables are illustrative USD scenarios, not measured average revenue for AI sites. Verified platform rules and research findings are attributed to published sources. No operator revenue screenshots, ad dashboards or checkout transactions were available for independent audit in this research.
What Counts as an AI Niche Website?
An AI niche website is a focused publishing property that uses artificial intelligence for some part of research, drafting, editing, image work, information architecture or operations. It might review specialist software, teach a technical workflow, cover a tightly defined hobby or provide original calculators. It is not automatically a website about AI; the “AI” can describe the production workflow rather than the topic.
That distinction matters economically. A site about AI writing tools competes against vendors, reviewers, forums and increasingly the assistants themselves. A website about independent bicycle maintenance may use exactly the same drafting software, yet face a different audience, product basket, affiliate structure and rate of product updates. The right comparison is between comparable search intents and customer behaviours, not between all sites using ChatGPT.
How Much Can You Realistically Earn From an AI Niche Website in 2026?
An honest income range has to start with traffic and advertising yield, because the site does not earn simply because it exists. Google’s AdSense documentation explicitly says earnings vary with traffic, subject, user location and ad configuration. There is no official universal “AI niche-site RPM”. Accordingly, the following values are planning assumptions rather than a claim about market averages (Google AdSense Help, n.d.).
| Monthly pageviews | At $5 page RPM | At $10 page RPM | At $20 page RPM |
| 1000 | $5 | $10 | $20 |
| 5000 | $25 | $50 | $100 |
| 20000 | $100 | $200 | $400 |
| 50000 | $250 | $500 | $1,000 |
| 100000 | $500 | $1,000 | $2,000 |
Page RPM is revenue per thousand pageviews: estimated display advertising revenue divided by pageviews, multiplied by 1,000. Do not confuse it with CPM measured per individual ad impression or with sessions, which can include multiple pageviews. Likewise, a premium-ad-network applicant may not be eligible simply because a scenario table indicates a healthy RPM.
Why Earnings Reports Rarely Describe the Average New Site
A newly launched website may earn nothing for extended periods, including after dozens of articles have been published. That fact is obscured by survivorship bias: operators who achieve five-figure months have more reason to publish case studies than those who shut down a domain. Self-reported case studies can still be useful as individual examples but cannot establish a population mean or median without sampling details and independent verification.
The frequently repeated “$1,000 per month after six months” claim should therefore be treated as a target model, not an earnings benchmark. Determine whether the report includes advertisements only, gross affiliate commissions, refunds, outsourcing, previous domain authority and the creator’s labour. Even a seemingly impressive monthly figure can reflect a negative economic return.
The Revenue Formula That Actually Matters
A niche-site model needs at least two distinct engines. Display revenue is roughly pageviews × page RPM ÷ 1,000. Affiliate revenue is qualified outbound clicks × purchase conversion rate × earned commission per conversion, adjusted for reversals. Products and sponsorships require their own units. The combined amount is gross revenue, not take-home income.
Consider 20,000 monthly pageviews, a $10 page RPM and 600 outbound affiliate clicks. Suppose 3% of clicks produce a confirmed commission and each approved commission averages $18. Display ads produce $200, and affiliate earnings are 600 × 0.03 × $18 = $324. Combined gross monthly revenue is $524. If direct operating costs reach $170, accounting profit before tax and owner labour is $354.
That calculation has a specific vulnerability: an affiliate network can report clicks today and pay later after validation. A return, a changed cookie attribution window or a terminated merchant programme can change the actual payout. Revenue should be reconciled to approved and collected transactions, not the most flattering dashboard estimate.
The Three-Part Conversion Funnel
Model a purchase journey as search visibility → qualified onsite attention → external merchant conversion. A page may rank well but receive few clicks when search results answer the question. A page may receive visitors but attract readers wanting free information rather than a purchase. A purchase may occur without attribution to your affiliate link. Tracking only impressions or pageviews hides these separate failure points.
A more disciplined implementation records Search Console clicks for each query group, GA4 landing-page sessions, outbound affiliate events and programme-confirmed earnings in one monthly worksheet. For operators evaluating the research and content layers, a separate overview of AI SEO tools and workflows provides useful context; it does not replace revenue instrumentation.
Traffic Volume, Intent and Geography Change the Calculation
Two sites with identical pageviews can generate very different revenue. A reader searching for the best accounting software for a ten-person company is closer to a purchase than someone asking for the meaning of a general technology term. Audiences based in different ad markets have different advertiser demand, and product prices affect commission value. Device mix, seasonality, viewability, consent rates and article depth add further variation.
To compare pages sensibly, segment traffic by commercially useful behaviour: informational questions, product comparisons, alternatives, tutorials with a natural tool recommendation and returning subscribers. Report revenue per landing-page session for each group. The aim is not to fill every article with affiliate links; it is to understand which useful pages also enable a transaction.
For a publisher with an AI-tools niche, an AI tools for entrepreneurs comparison may be relevant to a reader evaluating a software stack. That is a stronger contextual recommendation than inserting unrelated offers into an article where the reader has no buying need.
| Visitor segment | Primary intent | Likely monetisation fit | Main measurement |
| Broad definitions | Answer a question | Ads; newsletter | Earnings per 1,000 pageviews |
| Tool alternatives | Evaluate vendors | Affiliate; lead generation | Approved value per outbound click |
| Specialist workflow | Complete a task | Templates; affiliate; product | Purchase rate by landing page |
| Returning readers | Stay informed | Newsletter; sponsorship | Revenue per engaged subscriber |
Advertising Networks: Current 2026 Access Rules
Outdated eligibility claims distort projections. The official Mediavine requirements page currently lists at least $5,000 in annual advertising revenue for its core programme, rather than relying on the historical 50,000-session threshold seen in older blog posts. Journey by Mediavine lists at least 1,000 qualifying premium-country sessions within 30 days, plus content quality, analytics and other review criteria. These are admission requirements, not guarantees of a particular advertising rate (Mediavine, 2026).
Raptive’s September 2026 support documentation lists a minimum of 25,000 pageviews each month. In the 25,000–99,999-pageview bracket, at least 50% of traffic must come from specified English-speaking markets; it also requires original content with meaningful human involvement, a domain at least six months old and correct analytics configuration. At 100,000 pageviews or more, the listed geographic share is 40%, with other conditions still applicable (Raptive, 2026).
| Option | Public threshold checked | Other meaningful limits | Price / publisher economics |
| Google AdSense | No universal published traffic minimum | Site and policy review; earnings vary | No fixed publisher RPM |
| Journey by Mediavine | 1,000 premium sessions / 30 days | GA4, quality and traffic review | Payout rate not fixed publicly |
| Mediavine main programme | $5,000 annual ad revenue | Audience-first content and brand safety | Site-dependent |
| Raptive | 25,000 monthly pageviews | Country mix, site age, human contribution | Site-dependent |
An editor comparing different AI tools for business should apply the same discipline to advertising vendors: check the primary documentation on the day of application. The general lesson is that eligibility, approval and revenue performance are different concepts. A qualifying site can still be rejected, while an accepted site may experience variable yield.
Affiliate Revenue: High Potential, High Variance
Affiliates can contribute more than display ads where the content answers an explicit buyer question, but commission schedules, approvals and disclosure obligations differ by merchant. Do not use a single assumed “industry affiliate rate” across software, physical products and financial services. Check the actual merchant agreement, eligible countries, cookie window, reversal policy and payout schedule before projecting returns.
A niche site that recommends a software subscription must also consider whether commissions are paid once, recurring, or only for specified tiers. A $100 subscription is not equivalent to a $100 commission. Some promotions pay per qualified lead rather than sale; others exclude existing customers or certain placements. Affiliate economics can change suddenly when a merchant revises its programme.
A Worked Affiliate Sensitivity Table
| Outbound clicks / month | Approved conversion | Commission / sale | Illustrative earnings |
| 100 | 2% | $15 | $30 |
| 300 | 3% | $20 | $180 |
| 600 | 3% | $18 | $324 |
| 1,000 | 5% | $30 | $1,500 |
Each row is a mathematical example, not a documented conversion benchmark. Change one assumption at a time to discover the critical dependence: at 600 clicks, moving from 3% to 1% conversion reduces the illustrative $324 to $108. This sensitivity is why a publisher should evaluate actual merchant outcomes before forecasting growth from additional articles.
Commercial pages must retain editorial independence. Test or document the product before recommending it; describe limitations; disclose commissions conspicuously; and provide alternatives when they fit better. Publishing low-quality affiliate round-ups solely to extract search clicks is not a durable business plan.
The Cost of Running a Niche Site Is More Than AI Credits
Domain registration, hosting, email, analytics, editorial review, original images, research access, contracted writing and technical maintenance can all become real costs. AI subscriptions are only one line. A low subscription price can hide the hours needed to correct fabricated claims, retest tools, update screenshots and resolve search or accessibility issues.
A simple cash budget might allocate $15 for hosting and infrastructure, $25 for tools, $30 for email and $100 for research or freelance assistance each month. That totals $170 before taxes and labour. These are illustrative allocations, not quoted current vendor prices: product plans, seats, rate limits, API usage and taxes require checkout verification. Operators can compare possible tools in the small-business AI stack guide, but must use the plans actually bought in their accounts.
Now include the owner’s time. If the owner spends 25 hours monthly and values time at $20 per hour, the economic labour charge is $500. In the earlier $524 revenue example, cash profit of $354 becomes an economic loss of $146 after charging for labour. No money necessarily leaves the bank for that unpaid time, but ignoring it makes a business appear more attractive than alternative uses of the same hours.
Break-Even Analysis for Two Business Models
A display-first site with $150 monthly fixed cash costs and a $10 realised page RPM must generate 15,000 monetisable pageviews to cover those costs before tax or labour. A site earning $20 per confirmed affiliate sale needs eight approved sales to exceed $150. These models can coexist, but the assumptions are fundamentally different. The second model may require less traffic if purchase intent is concentrated; it may also yield nothing if approvals never arrive.
Break-even should be checked monthly and on a rolling quarter, because traffic and approval cycles are uneven. A standalone ROI and break-even calculators review illustrates why it is important to inspect input assumptions and missing formulas in quick online tools.
How AI Search Changes the Traffic Opportunity
In its research published in July 2025, Pew Research Center found that people clicked conventional Google result links in 8% of visits when an AI summary appeared, compared with 15% when no AI summary appeared. Its dataset concerned observed browsing by 900 US adults, rather than a controlled experiment for every niche. This finding does not mean that a particular publisher lost exactly 47% of its traffic. It does show why historical click-through expectations deserve stress testing (Chapekis & Lieb, 2025).
Cloudflare introduced a crawl-to-refer metric showing that AI crawling and referral traffic are not the same. Its 2025 year-in-review documented very different crawl and referral patterns across AI platforms. The metric has limitations, including missing referrer data for some native applications. Neither a citation in an AI response nor a bot crawling a page is equivalent to a monetisable visit (Cloudflare, 2025).
A strong content portfolio therefore needs some queries where a reader benefits from reaching the original website: a functioning tool, original measurements, downloadable data, current documented tests, personalised calculations, photographs, or nuanced purchases. Plain definitions are easier for an answer engine to satisfy without a click.
Search Dependence as a Revenue Risk
If 85% of a site’s sessions come from organic search, it should model an abrupt reduction rather than assuming uninterrupted growth. An illustrative $1,000 monthly site that experiences a 40% proportional traffic drop could fall to roughly $600 if monetisation rates stay fixed. Actual changes can be larger or smaller because losing commercially valuable pages affects revenue disproportionately. Monitor revenue by landing-page cohort and acquisition channel, not only by total users.
Google’s guidance says that generative AI can help organise content but mass production without added value may violate scaled-content rules. Google also published advice in May 2026 emphasising original, useful material in generative search. AI is not the disqualifying characteristic; the lack of distinctive user value is the issue (Google Search Central, 2026).
What a Twelve-Month Launch Can and Cannot Promise
A practical calendar is a sequence of decisions, not a forecast that traffic grows smoothly. In months one to three, identify a narrow audience, inventory purchase-intent searches, produce a small set of useful pages and configure measurement. Spend effort on original tests or calculations that competitors cannot generate by merely rephrasing each other. An operator can use AI productivity tools to reduce repetitive editing, but quality control remains a separate obligation.
In months four to six, inspect impressions, rankings, engagement and early affiliate clicks. Delete or combine overlapping thin pages rather than publishing more variations of the same question. Investigate whether the strongest pages attract a recognisable audience, and build an email subscription that provides a real reason to return.
In months seven to nine, scale only a proven cluster. Aim to improve the conversion of existing traffic before multiplying publishing volume. Where tools are compared, re-check product features, pricing and limitations from official documentation. In months ten to twelve, decide whether to continue, narrow, sell a product to the audience or stop. Those decisions should be made against pre-set cash and time budgets.
There is no honest basis for assigning guaranteed month-by-month earnings to this schedule. A new domain, prior expertise, backlinks, competition, language and geography all affect visibility. Failure to earn in the first year is possible. Conversely, an operator with an existing audience or a proprietary dataset can convert earlier; such a result should not be generalised to a beginner.
Build a Niche That Produces Evidence, Not Just Articles
The least defensible site in 2026 is a large catalogue of interchangeable generic text. The stronger opportunity is a compact publishing business with something verifiable: original product tests, clear calculators, expert interviews, proprietary survey results, open datasets, detailed troubleshooting steps or a community that trusts its corrections. An AI assistant can organise these assets, but cannot retrospectively create real observations that never happened.
Choose one repeatable evidence unit per article. For a software review, capture the plan, date, workflow attempted, time to completion, observed error and alternative tested. For a consumer niche, use original purchase evidence, dimensions, photographs or field experience. For a data site, publish the formula and assumptions that make a calculation auditable. Readers and search systems have a reason to value records that are not duplicated everywhere.
It is also worth tracking how much research a page requires after publication. Software comparisons have update debt: subscriptions change, features migrate, integrations are removed. A page that earns $8 per month but needs two hours of expert review every quarter may destroy value. The neglected cost is the editorial maintenance liability created by every additional URL.
Publishers working with spreadsheets, APIs and customer research can use AI data analysis tools for draft checks and exploratory work, then confirm every published calculation independently. Internal linking should follow the user journey—research, comparison and implementation—rather than a rule that every page must point to every other page.
Three Original Decision Metrics for a 2026 Publisher
1. Revenue per Verified Content Hour
Divide collected revenue attributable to a page cohort by the total research, writing, QA and maintenance hours spent on that cohort. Compare articles published in the same quarter and account for reporting lag. This exposes “cheap” AI pages that require costly fact correction and provides a more meaningful comparison than words per hour.
2. Monetisation-Weighted Traffic Concentration
Calculate the share of total site revenue contributed by the top five landing pages, rather than merely the share of visits. A site with diverse traffic but 75% of revenue tied to one ranking has a hidden dependency. Test the impact of losing that page, its affiliate programme or its highest-paying country segment.
3. Maintenance-Adjusted Revenue Yield
For an article, subtract its attributable update and verification cost from its collected monthly revenue. The question becomes whether another publishing cycle generates more durable returns than improving an existing page. This approach is especially important for frequently changing AI products, whose published limits can become inaccurate quickly.
Scenario Planning: Conservative, Base and Upside
Instead of one aspirational earnings figure, maintain three scenarios. The conservative case uses modest search capture and weak affiliate conversion; the base case uses outcomes measured from the property’s own early cohorts; the upside case requires clear evidence that the acquisition and transaction funnels both improve. The table below demonstrates distinct hypothetical site states, not an expected twelve-month progression.
| Illustrative site state | Pageviews | Page RPM | Affiliate revenue | Gross revenue | Cash costs | Cash result |
| Low traction | 2,000 | $5 | $0 | $10 | $120 | −$110 |
| Developing | 20,000 | $10 | $324 | $524 | $170 | $354 |
| Established | 60,000 | $15 | $1,000 | $1,900 | $400 | $1,500 |
| Upside case | 120,000 | $20 | $2,500 | $4,900 | $900 | $4,000 |
A prudent decision-maker should also run a downside exercise: cut search traffic by half, eliminate the largest affiliate partner, increase editorial workload by one-third and ask whether the website remains worth maintaining. An “upside” case that only works if all favourable assumptions occur simultaneously is not a sensible operating budget.
A Practical Measurement Setup for the First 90 Days
Start with a simple technical stack rather than a crowded set of dashboards. Verify domain ownership in Google Search Console, install Google Analytics 4 with consent controls appropriate to your audience, establish a lightweight monthly income-and-expense ledger, and use separate tracking identifiers for affiliate programmes where permitted. Capture the date and version of any offer discussed so that an expired price does not silently remain in a buying guide.
In Search Console, group queries into commercial comparisons, practical workflows and answer-only information. Watch impressions and clicks separately: impressions can rise while organic clicks stagnate. In GA4, inspect the landing page and acquisition source before assigning value to the visit. For an affiliate link, configure an outbound-click event with a stable destination category, but do not mistake that event for a sale. Only the merchant or programme can confirm that the referred conversion was valid and commissionable.
Create a monthly reconciliation calendar. On day one, record website costs and subscriptions. After month-end, extract pageviews, qualified sessions and affiliate click counts. When affiliate payouts are finalised, add reversals and confirmed earnings to the revenue ledger. In a fourth worksheet column, record editorial update hours per article, not just publishing hours. This arrangement allows the operator to compare cash profit, economic profit and article-level revenue without pretending that provisional affiliate dashboards represent banked income.
What to Do When Early Data Disagrees With Your Plan
Low impressions indicate that the chosen topics may have too little demand, pages may not be indexed, or the site lacks sufficient visibility. They do not automatically mean that the writer needs more articles. Check indexing and query targeting first. High impressions but low click-through call for a different diagnosis: intent mismatch, uncompetitive search snippets, ranking position or search experiences that satisfy the reader before a click. High onsite traffic but no earnings points toward audience intent, ad eligibility, placement or the merchant funnel.
Treat each diagnosis as a hypothesis to test, not a cause proven by a single metric. For instance, updating ten pages simultaneously and changing affiliate merchants in the same month makes it difficult to attribute any resulting improvement. Choose a cohort, record a baseline, alter one prominent element and observe enough traffic for meaningful comparison. A genuine experiment may find that there are too few visits to draw a reliable conclusion. That is useful information, not a reason to manufacture certainty.
A site should also have a stopping rule. A straightforward example is a six-month maximum cash-loss budget coupled to three leading indicators: non-branded search clicks, newsletter sign-ups from qualified readers and approved affiliate revenue. If none improves despite original content and corrections, reconsider the niche or distribution strategy. A publisher with a limited budget may be better served by expanding a service business or a focused newsletter than by investing indefinitely in more indexed pages.
The distinction between a publishing asset and a content-production habit is commercially important. An asset accumulates returning readers, trustworthy first-party evidence, useful rankings and direct revenue relationships. A habit accumulates URLs. Both may feel productive, but only the first has a plausible path to a durable operating return.
Our Editorial Verification Process
We reviewed Google AdSense Help for RPM terminology and earnings uncertainty; current Mediavine and Journey eligibility statements; Raptive’s September 2026 eligibility document; Google Search Central guidance on generative AI and scaled content; Pew’s observed-search study; and Cloudflare’s crawler/referral analysis. Display and affiliate scenario tables were derived arithmetically from explicitly labelled input assumptions. They are not sampled publisher earnings benchmarks.
We also examined ten discoverable competing or adjacent results covering AI content-site monetisation, affiliate models, earnings anecdotes and niche-site case studies. The recurring weakness was unverified self-reported revenue, assumed RPMs, older ad-network rules, or insufficient treatment of owner time. This article instead separates observed evidence, official programme eligibility and transparent mathematical scenarios. The target site’s sitemap.xml, sitemap_index.xml and post-sitemap.xml could not be retrieved as parseable XML; seven live indexed, contextually relevant internal article URLs were checked and used in body sections. No direct first-hand revenue test was conducted.
This article was researched and drafted with AI assistance for review by the Sami Ullah Khan editorial desk at Perplexity AI Magazine. Primary-source figures and published rules were checked in the linked documentation; the hypothetical revenue models require no claim of field testing. Final editorial approval and any additional independent verification should be completed before publication.
Conclusion
The most useful answer to the earnings question is not a universal monthly number. A niche website can generate no revenue, modest side income or a substantial business, but its result depends on search visibility, the commercial usefulness of its visitors, verified conversions, expenses and how much human work the content requires. AI reduces parts of the production burden; it does not guarantee attention or trust.
The economics have also become less forgiving of generic explanations. Search systems increasingly answer simple questions directly, while advertising platforms assess quality and traffic composition rather than granting identical yields to all publishers. On the positive side, genuinely useful original data, workflows and buying guidance can support several revenue streams even without extraordinary traffic.
For a new publisher, the most rational starting point is a small measurable portfolio, documented assumptions and explicit continuation criteria. Treat any spectacular case study as one possible outcome, not a forecast. The open question for 2027 is how much referral traffic AI-powered discovery will return to independent publishers; that uncertainty belongs in the business model today.
Frequently Asked Questions
How much can you realistically earn from an AI niche website in 2026?
A new site can realistically earn $0, and some established sites can earn hundreds or thousands of dollars monthly. There is no verified market-wide median for all AI niche websites. Calculate your own range from monetisable traffic, realised advertising RPM, approved commissions and operating costs; consult Google AdSense and your affiliate programme for actual account performance.
Can an AI-written website make $1,000 a month?
Yes, but publishing with AI is not sufficient. For example, $1,000 in display revenue alone requires 100,000 pageviews at a hypothetical $10 page RPM. A mixed affiliate and ads model can reach the same figure with different traffic, but conversions and commissions must materialise.
How long does it take an AI niche site to make money?
There is no dependable timetable. New sites can make nothing for months and may never become profitable. Existing audiences and commercially valuable topics can change the result. Use milestones for measured demand, conversions and profitability rather than a promised month-six income.
How many articles do I need to earn $500 per month?
There is no fixed article count. A few highly useful commercial pages could outperform hundreds of generic posts. Track revenue and maintenance cost by landing page instead of using word count or URL count as a revenue forecast.
Is AdSense or affiliate marketing better for an AI niche site?
Advertising works when pages attract monetisable views; affiliates suit readers ready to evaluate or purchase relevant products. A blended model may diversify income. The better choice depends on measured audience intent, country mix and actual programme terms.
Does Google penalise content just because AI helped write it?
Google’s published guidance focuses on value and scaled-content abuse, not a blanket prohibition on AI assistance. Generating many low-value pages primarily to manipulate rankings can violate spam policy. Original evidence, accuracy and usefulness matter more than the mere presence of AI in the workflow.
What is the biggest hidden cost of an AI niche website?
Human research, verification and updating are frequently overlooked. Subscription payments may be modest while maintaining volatile software comparisons consumes substantial hours. Include the operator’s labour when judging whether a site provides a meaningful financial return.
Can AI search reduce a niche website’s earnings?
It can reduce clicks for some queries, although effects vary. Pew observed fewer traditional-result clicks when AI summaries appeared in its study. Model traffic downside, diversify acquisition and publish assets readers need to visit directly.
References
Chapekis, A., & Lieb, A. (2025, July 22). Google users are less likely to click on links when an AI summary appears in the results. Pew Research Center. https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/
Cloudflare. (2025). The 2025 Cloudflare Radar year in review. https://blog.cloudflare.com/radar-2025-year-in-review/
Google AdSense Help. (n.d.). How much will you earn with AdSense? https://support.google.com/adsense/answer/9902
Google AdSense Help. (n.d.). Revenue per thousand impressions (RPM). https://support.google.com/adsense/answer/190515
Google Search Central. (n.d.). Google Search’s guidance on using generative AI content on your website. https://developers.google.com/search/docs/fundamentals/using-gen-ai-content
Google Search Central. (2026, May 15). A new resource for optimising for generative AI in Google Search. https://developers.google.com/search/blog/2026/05/a-new-resource-for-optimizing
Mediavine. (2026). Mediavine and Journey application requirements. https://www.mediavine.com/application-requirements/
Mediavine. (2026, August 7). Two years of Journey by Mediavine. https://www.mediavine.com/blog/two-years-of-journey-by-mediavine/
Raptive. (2026, September 26). Who is eligible for Raptive? https://help.raptive.com/hc/en-us/articles/360032840891-Who-is-eligible-for-Raptive
