PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteMobile ad networks connect advertisers with the ad space inside apps. They evaluate each ad request, return the best eligible advert or no fill, and pay the publisher according to impressions, clicks, installs or another contracted outcome. To judge performance, read eCPM (earnings per 1,000 impressions) together with fill rate (the share of requests that receive an ad). Mediation lets several demand sources compete for each opportunity. A higher eCPM can still earn less when its fill is weak, latency is high or ads are not shown.
What a mobile advertising network actually does
An app publisher creates an ad unit—such as a banner slot, interstitial placement, rewarded-video position, native card or app-open placement—and adds the network’s software development kit (SDK) to the app. When that placement becomes eligible to show an advert, the SDK sends an ad request.
1. The request carries eligibility and context
The request can include the ad format, app and placement identifiers, device and operating-system details, approximate location or country, consent state and other policy signals. Demand sources assess whether they can serve an eligible campaign to that opportunity. A source returns an advert or reports no fill. Google describes AdMob as matching ads to app criteria, selecting high-performing demand from its own and third-party networks, handling advertiser billing and paying publishers.
As an Amazon Associate I earn from qualifying purchases.
2. A returned advert becomes an impression only when it is shown
Receiving a response is not the same as producing an impression. An impression is counted under the platform’s measurement rules when the advert is displayed. The user may then click, watch a rewarded video, install an app or complete another contracted action. The buying model determines whether the advertiser pays for an impression, click, install or another interaction.
3. The network settles the transaction
The network or mediation layer records the billable event, subtracts any contractual fees or intermediary charges and remits the publisher’s share. The applicable split depends on the product, demand source, country, contract and whether another intermediary is involved; there is no reliable universal percentage that applies to every network transaction.
eCPM, fill rate and show rate: the metrics you must separate
These metrics use different denominators. Mixing them is one of the fastest ways to misread an app’s advertising performance.
| Metric | Formula or meaning | What it answers |
|---|---|---|
| eCPM | (Total earnings ÷ impressions) × 1,000 | How much the impressions earned per 1,000 displayed adverts |
| Fill rate | Ad-network deliveries ÷ ad requests | How often a request receives an advert |
| Matched-request rate | Matched requests ÷ requests, using the platform’s definition | How often the source finds a candidate advert |
| Show rate | Impressions ÷ matched requests | How often a matched advert is actually displayed |
| Revenue per request | Total earnings ÷ requests | The practical yield of every opportunity sent to demand |
Google defines eCPM as “effective cost per thousand impressions.” It is an observed or estimated yield, not a promise for the next thousand impressions. Compare it over the same ad format, country mix, platform, audience and time window. A rewarded-video eCPM from one country is not a meaningful benchmark for banner inventory in another.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Google’s definition of fill rate is the number of times an ad network delivers an advert divided by the number of times the app requested one. Use the exact denominator shown in your reporting console: request, matched request, impression and show-rate figures are not interchangeable.
Why eCPM and fill rate must be read together
When eCPM is calculated on impressions and fill is expressed as a decimal, expected revenue per 1,000 requests is approximately:
Rank #2
eCPM × fill rate
For example, Google’s worked comparison gives:
| Scenario | eCPM | Fill | Approximate revenue per 1,000 requests |
|---|---|---|---|
| A | $1 | 85% | $0.85 |
| B | $2 | 30% | $0.60 |
The $2 eCPM looks better in a leaderboard, but the $1 source produces more money per request in this example. The approximation assumes the figures use compatible definitions and does not replace a reconciliation against actual earnings.
Other numbers that change the result
- Show rate: a response can arrive after the user leaves the screen, or the placement can be hidden, so not every matched advert becomes an impression.
- Latency: a slow response can delay a screen, cause a timeout or make the app skip the placement.
- Invalid-traffic adjustments: earnings can be corrected when traffic or interactions fail quality checks.
- Consent restrictions: limited signals or non-personalized demand can change auction participation and price.
- Retention: intrusive timing, accidental clicks or excessive frequency can reduce future sessions and advertising opportunities.
How mediation decides which network gets the request
Mediation is the control layer between the app and several ad sources. Instead of relying on one SDK to answer every request, the mediation platform chooses which source to try and records the outcome.
Recommended Free Tools
Traditional waterfall mediation
In a waterfall, sources are ordered by an eCPM value. That value may be manually entered, based on observed performance or supplied by an optimization system. The source at the top is tried first. If it does not return an advert within the allowed time or has no eligible demand, the request moves to the next source.
A high position does not guarantee delivery. Availability, geography, targeting rules and response time determine whether the chain reaches the next source. A stale manual eCPM can also put a weak source above a stronger one.
Real-time bidding
In bidding, participating sources submit bids in parallel. The winning bid is inserted into the mediation ordering according to its eCPM. This allows several sources to compete for the same impression without waiting for a fixed sequence of calls. A hybrid setup can contain bidders alongside waterfall-only sources.
| Characteristic | Waterfall | Bidding |
|---|---|---|
| Timing | Sources are called in sequence | Participating bidders submit in parallel |
| Ordering | Configured or optimized eCPM values | Winning bid is ranked by eCPM |
| Main risk | Stale rankings and added wait time | Fewer eligible bidders in some countries or formats |
| Common deployment | Useful for sources without bidding support | Often combined with waterfall-only demand |
eCPM floors: when a higher minimum price hurts revenue
An eCPM floor is the minimum price at which a source may win. Raising it can improve the price of impressions that clear the threshold, but it is likely to reduce fill. A request rejected by every source earns nothing, so evaluate floors against realized revenue per request or per active user—not eCPM alone.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Rank #4
A safer floor-testing method
- Split traffic by meaningful segments such as country, format, operating system or audience value.
- Record baseline revenue per request, fill, show rate, latency and retention before changing the floor.
- Change one floor or floor group at a time and leave enough traffic for a stable estimate.
- Compare total earnings and user outcomes with the baseline, not just the winning eCPM.
- Keep a dated change log so seasonal demand is not mistaken for an optimization gain.
Sparse traffic can leave an optimization system in a pending or low-confidence state. Avoid making aggressive decisions from a few dozen impressions or from a single unusually strong day.
What percentage do ad networks take?
There is no single current percentage that applies to all mobile advertising networks. The advertiser’s payment may cover an impression, click, install or another event. A network or mediation platform then applies its contractual fee and pays the publisher. The result varies by:
- the network product and ad format;
- the demand source and whether it is a direct or third-party buyer;
- the publisher’s country and contract;
- whether a separate mediation or exchange intermediary is involved; and
- adjustments for invalid traffic, refunds or other contract terms.
AdMob documentation confirms that publishers can be paid for clicks, impressions and other interactions, but the available primary documentation does not establish one universal current split for every AdMob or third-party transaction. Use the terms attached to the specific account and product rather than quoting a generic “the network keeps X%” figure.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to compare networks and mediation platforms
Build a comparison around the outcomes your app needs. A source with the highest headline eCPM may be a poor choice if it fills rarely, adds crashes or serves unsuitable creatives.
| Area | Questions to ask |
|---|---|
| Realized yield | What are revenue per request, per impression and per daily active user? |
| Availability | How do fill, matched-request rate and show rate vary by country, format and season? |
| Demand mechanics | Does the source support bidding, waterfall placement, floors and automated optimization? |
| Format coverage | Are banner, interstitial, rewarded, app-open, native and rewarded-interstitial placements supported where needed? |
| Operations | What are SDK size, latency, crash risk, adapter maintenance and reporting detail? |
| Policy and privacy | Does the setup support consent management, app-ads.txt or equivalent supply-chain controls, brand-safety review and ad-quality controls? |
| User experience | Can you control frequency, timing, accidental clicks and reward conditions? |
Network capability is not uniform. Google’s source directory notes that format coverage, bidding support and automatic data collection vary by network and country; some sources require manual eCPM configuration. If you connect Google demand through AppLovin MAX, its documentation says you need a valid AdMob account and recommends creating new placement IDs for accurate MAX reporting.
A practical implementation sequence
- Instrument the baseline. Track requests, matched requests, impressions, show rate, clicks, revenue, response latency and crashes by ad unit, country, operating system and format. Preserve a common event ID or timestamp so app analytics and network reports can be reconciled.
- Choose a primary source. Start with one dependable network and only the formats and geographies your app can support. Verify test ads before releasing production traffic.
- Add compatible demand partners. Install mediation adapters, check SDK versions and confirm that each partner’s reporting dimensions map to your own event names.
- Complete privacy and policy work. Obtain consent where required, publish the necessary privacy disclosures, configure supply-chain controls and test restricted-consent paths. A request that cannot legally or technically participate should not be counted as a normal performance failure.
- Run a controlled comparison. Keep a baseline group while testing mediation. Compare revenue per request, fill, show rate, latency, crashes and retention. Do not judge the test on eCPM alone.
- Adjust gradually. Change waterfall order, floors or bidder participation in small steps. Record the date, segment and reason for every change.
- Remove unproductive sources. A partner that adds latency, poor creative quality, policy exposure or maintenance work without improving realized yield is a cost, even if it occasionally reports a high eCPM.
AppLovin recommends beginning with at least six networks with strong global demand in addition to its own demand when establishing a baseline. That is vendor guidance, not a universal requirement: a smaller, well-instrumented set can be more appropriate for a regional or specialized app.
Diagnosing low fill or falling revenue
Fill is low across every country and format
- Check that the ad unit ID, app ID and SDK configuration are correct in every build.
- Confirm that requests are sent only after the placement is ready and that test mode is not suppressing production demand.
- Inspect consent state, age or content restrictions and any policy-limited serving status.
- Look for timeouts, connectivity failures and adapter initialization errors.
Fill is acceptable but revenue is weak
- Separate eCPM by country, platform and format; a blended number can hide a low-value segment.
- Check show rate to find responses that never become visible impressions.
- Review floors and waterfall rankings for stale values that block lower-priced but available demand.
- Inspect ad quality, accidental-click reports and retention for signs that the placement is damaging user behavior.
Revenue falls after adding mediation
- Compare equivalent dates and traffic segments; seasonality can look like a configuration effect.
- Verify that adapters, placement IDs and reporting joins are correct. Misaligned IDs can make one source appear to have no revenue.
- Measure added latency and crashes. A theoretical auction gain is not useful if fewer placements reach the impression stage.
- Roll back one change at a time, then reintroduce sources under a controlled experiment.
Scale, seasonality and realistic expectations
Country mix, format, device quality, consent state, demand competition, placement design and time of year all influence results. No responsible network comparison can promise a particular eCPM, fill rate or publisher share without those details.
Google AdMob reported nearly one million advertisers and 200 billion ad requests per month globally in 2017. Those figures are historical context, not a current estimate of platform scale.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →The most useful operating dashboard therefore puts revenue per request and revenue per active user beside eCPM, fill, show rate, latency, crashes and retention. That view reflects what the app actually earns and what users experience.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
