Most monetization advice for new publishers assumes you already have a dashboard full of data to work from. New sites do not. The whole point of estimating revenue before launch (or in the first few months) is to decide whether the content investment makes economic sense, and waiting twelve months to find out is not a viable plan.
The honest workflow uses a public revenue calculator as the starting model, then adjusts it with a small set of inputs you can derive from competitor analysis, keyword research, and a clear-eyed reading of your niche. The output will not be precise, but it will be much closer to defensible than a back-of-envelope guess.
This guide walks through that workflow step by step.
Photo by Harun on Pexels
Step 1: Define the Specific Sub-Niche, Not the Broad Category
The single biggest source of error in pre-launch revenue estimates is using a broad niche category when the real content is a narrow sub-niche.
"Personal finance" is not a sub-niche. "Credit card reviews for travel rewards" is. "Technology" is not a sub-niche. "Self-hosted homelab software guides" is. The CPC band for the broad category is an average across many sub-niches, and your specific topic will sit somewhere on that spectrum.
Write down the actual sub-niche your content will cover. Then map it to the closest band:
- High-CPC sub-niches ($5 to $40 CPC): insurance comparisons, mortgage refinance, B2B software, legal services, addiction treatment, online degree programs.
- Mid-CPC sub-niches ($1 to $5 CPC): credit cards, personal finance broad, home improvement, parenting, technology reviews for consumer products.
- Low-CPC sub-niches ($0.20 to $1 CPC): recipes, general news, sports recap, entertainment, gaming, viral lifestyle.
Be honest. Most new sites overestimate where their sub-niche falls on this scale.
Step 2: Pull Comparable Site Traffic Numbers
You do not have traffic data yet, but you can find approximate numbers for sites covering similar topics. A SimilarWeb or Ahrefs traffic estimate for three or four competitor sites in your sub-niche tells you what realistic monthly pageview counts look like after one, two, and three years.
Take the median, not the maximum. The biggest site in any niche is the outlier. If competitor sites land between 20,000 and 200,000 monthly pageviews after two years, the realistic median is closer to 50,000, not 100,000.
This is also the step where you discount any "I will be the exception" thinking. Three out of four content sites never reach 100,000 monthly pageviews from organic search. Plan against the median outcome, not the best case.
Step 3: Plug the Numbers Into a Revenue Calculator
Use a calculator that exposes the niche category and traffic level as inputs. The AdSense Revenue Calculator is one option that lets you adjust niche and traffic projections to see how the model behaves at different assumed levels.
Run the math at three traffic levels:
- Six-month projection (realistic minimum based on competitor early-stage traffic).
- Twelve-month projection (the median competitor outcome at one year).
- Twenty-four-month projection (the median competitor outcome at two years).
Record the RPM estimate the calculator returns at each level. Do not move the niche slider during this step.
Step 4: Apply the Geographic and Traffic-Source Discounts
The calculator's default RPM usually assumes a tier-one geographic mix and an organic-search-dominant traffic profile. If your site does not match those defaults, the estimate will overshoot.
Apply discounts:
- If your audience is less than 50 percent tier-one (United States, United Kingdom, Canada, Australia, Western Europe), reduce the estimate by 25 to 40 percent.
- If your projected traffic mix is more than 50 percent social or referral, reduce the estimate by 20 to 30 percent.
- If your topic skews to a tech-savvy audience (developers, security, gaming), reduce by an additional 15 to 25 percent for ad-block rate.
Stack these discounts. Most pre-launch estimates that turn out to be wrong are wrong because the publisher skipped this step.
Step 5: Sanity-Check Against a Public Competitor's Disclosure
Some niches have published RPM disclosures from successful sites. Search for "Pat Flynn AdSense earnings" or "site:moz.com publisher revenue" or "ProBlogger income report" and read three or four real-publisher writeups in your sub-niche.
The actual published numbers are usually within 30 percent of what a properly discounted calculator estimate would produce. If your estimate is wildly higher than three real-publisher disclosures in the same niche, your inputs are too generous. Background reading on how the auction itself prices clicks, including the Wikipedia entry on pay per click advertising, fills in why the niche bands in public calculators are necessarily averages over a wide range.
Wikipedia's overview of online advertising covers the structural reasons why publisher revenue ranges are so wide and why aggregated industry numbers diverge from individual-site reality.
Photo by Yelena from Pexels on Pexels
Step 6: Build a Range, Not a Point Estimate
The final output should be a range, not a single number. Take the discounted estimate from step 4 and bracket it:
- Conservative case: discounted estimate minus 30 percent.
- Central case: discounted estimate as calculated.
- Optimistic case: discounted estimate plus 30 percent.
Plan the content investment against the conservative case. If the business math only works at the optimistic case, the model is too fragile to commit to.
Step 7: Lock the Estimate and Revisit Quarterly
Save the inputs and the resulting range. Once you have three months of real AdSense data, compare actual RPM to the central case estimate.
If actual is within 25 percent of the central case, the model was honest. If actual is more than 50 percent below, one of the audits (niche, geo, ad block, traffic source, ad density) needs revisiting. Industry monitoring from organizations like the Interactive Advertising Bureau tracks these structural shifts at the macro level, which helps explain why a previously-honest estimate can drift over time as the auction environment changes.
A longer reference on why AdSense RPM varies so much by niche and traffic source walks through the auction mechanics in more depth, which is useful when you are trying to explain to a stakeholder why the estimate range is as wide as it is.
Why This Workflow Matters
The alternative to this workflow is one of two failure modes:
The first is committing to a content investment based on an inflated single-point estimate from a calculator with default settings. When real revenue lands at half the estimate, the project looks like a failure and gets abandoned, even though it might be in line with what a properly discounted estimate would have predicted.
The second is refusing to estimate at all and starting with "we will see what happens." This produces a different failure mode: the publisher invests in content production for twelve months, hits 60,000 monthly pageviews, sees $400 a month in revenue, and realizes too late that the math never made sense for the cost of production.
A defensible pre-launch estimate is not a guarantee of revenue. It is a check on whether the planned investment has a reasonable shot of producing return at the median outcome, not just the best case.
Practical Reminder
The estimate is a model. The model will be wrong. The question is whether it is wrong by 25 percent or 250 percent, and the seven-step workflow above keeps the error in the smaller range. That is enough to support a content investment decision and to know what to track once real data starts flowing in.
Most pre-launch revenue estimates that turn out badly were not bad estimates. They were single-point estimates run with default inputs and no discount applied. Spend an hour on the audits before committing to the budget.
Top comments (0)