Performance Benchmarking: Managed Ad Networks
When selecting a managed ad network, understanding their impact on your site's performance is crucial. We've benchmarked the top ad networks to provide a clear picture of how they affect page speed, CPU time, interaction jank, and payload size.
All data points and performance percentages presented in this resource are derived from Bidkernel's internal automated Lighthouse benchmarking suite, which simulates realistic user interaction flows across publisher sites. See the Testing Methodology section for more details.
Here is a comprehensive rundown of each network's performance, listed in alphabetical order.
Aditude Link to section
Aditude leverages aggressive lazy-loading techniques that actually result in a reduced initial payload size (-0.27 MB on average) compared to control sites. However, this optimization comes at a cost to interactivity, as the delayed execution significantly increases interaction jank (+223ms) and causes a noticeable drop in overall Lighthouse scores.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 79.8 | 65.3 | -14.5 pts | -18.2% |
| CPU Time | 11317 ms | 23242 ms | +11925 ms | +105.4% |
| Interaction Jank | 158 ms | 381 ms | +223 ms | +141.4% |
| Payload | 2187 KB | 1910 KB | -277 KB | -12.7% |
| Requests | 373 | 697 | +324 | +87.1% |
Concept Link to section
Concept is the top performer in our benchmarks when it comes to preserving baseline Lighthouse scores, introducing a mere 3.7% penalty. They maintain excellent site responsiveness with virtually no added interaction jank (+15ms), though their overall network payload is slightly heavier than average.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 94.7 | 91.1 | -3.5 pts | -3.7% |
| CPU Time | 4996 ms | 11933 ms | +6937 ms | +138.9% |
| Interaction Jank | 59 ms | 74 ms | +15 ms | +25.3% |
| Payload | 405 KB | 2983 KB | +2578 KB | +637.2% |
| Requests | 168 | 259 | +91 | +54.0% |
Ezoic Link to section
Ezoic provides a highly comprehensive monetization stack, but it introduces considerable performance overhead. Our tests show significant increases in CPU processing time (+19.3 seconds) and a heavy network footprint (+895 requests), leading to a steep 27.2% reduction in Lighthouse performance scores.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 84.3 | 61.3 | -22.9 pts | -27.2% |
| CPU Time | 11376 ms | 30663 ms | +19288 ms | +169.6% |
| Interaction Jank | 262 ms | 585 ms | +323 ms | +123.3% |
| Payload | 536 KB | 2551 KB | +2015 KB | +376.3% |
| Requests | 236 | 1131 | +895 | +379.3% |
Freestar Link to section
Freestar balances a robust ad delivery system with relatively contained CPU overhead (+12 seconds). However, publishers should be prepared for a substantial network footprint, as Freestar introduces the heaviest payload in our tests (+3.35 MB) alongside significant interaction delays (+348ms).
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 92.5 | 63.5 | -29.0 pts | -31.4% |
| CPU Time | 9231 ms | 21226 ms | +11995 ms | +129.9% |
| Interaction Jank | 159 ms | 507 ms | +348 ms | +219.7% |
| Payload | 207 KB | 3640 KB | +3434 KB | +1662.8% |
| Requests | 102 | 807 | +705 | +692.3% |
Mediavine Link to section
Mediavine is a highly optimized network that excels at minimizing JavaScript execution time (+5.9 seconds CPU load). Their lightweight footprint results in very little interaction jank (+62ms), allowing them to preserve Lighthouse performance scores significantly better than the industry average.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 96.4 | 85.5 | -10.9 pts | -11.3% |
| CPU Time | 6635 ms | 12570 ms | +5935 ms | +89.4% |
| Interaction Jank | 89 ms | 151 ms | +62 ms | +69.8% |
| Payload | 127 KB | 1094 KB | +968 KB | +763.4% |
| Requests | 68 | 271 | +203 | +298.7% |
MonetizeMore Link to section
MonetizeMore delivers an incredibly lightweight ad stack, adding only half a megabyte to the page payload while keeping request counts remarkably low. This streamlined approach minimizes interaction delays and excellently preserves Lighthouse performance scores with under a 10% penalty.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 92.1 | 83.4 | -8.7 pts | -9.5% |
| CPU Time | 4087 ms | 13768 ms | +9681 ms | +236.9% |
| Interaction Jank | 115 ms | 163 ms | +48 ms | +42.0% |
| Payload | 314 KB | 879 KB | +565 KB | +180.1% |
| Requests | 79 | 210 | +131 | +164.5% |
NitroPay Link to section
NitroPay maintains a moderate payload size for a comprehensive ad delivery stack, but it struggles with execution efficiency. The heavy JavaScript processing requirements (+19.6 seconds) create noticeable interaction delays, resulting in a substantial hit to overall page performance metrics.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 91.8 | 62.9 | -28.9 pts | -31.4% |
| CPU Time | 6196 ms | 25748 ms | +19552 ms | +315.6% |
| Interaction Jank | 318 ms | 554 ms | +236 ms | +74.3% |
| Payload | 670 KB | 2280 KB | +1610 KB | +240.5% |
| Requests | 83 | 527 | +445 | +537.8% |
Playwire Link to section
Playwire provides an efficient execution profile with moderate CPU processing (+9.1 seconds) despite offering a wide array of monetization features. While their interaction jank is kept reasonably low, they do introduce a high volume of network requests that could be further bundled or optimized.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 84.8 | 69.9 | -14.9 pts | -17.6% |
| CPU Time | 8068 ms | 17204 ms | +9136 ms | +113.2% |
| Interaction Jank | 151 ms | 264 ms | +112 ms | +74.2% |
| Payload | 697 KB | 2323 KB | +1626 KB | +233.3% |
| Requests | 325 | 830 | +505 | +155.6% |
Pub Collective Link to section
Pub Collective keeps their CPU execution overhead surprisingly efficient (+10 seconds) given the complexity of their ad stack. However, publishers will see a significant increase in their network footprint, as the platform adds over 3 MB of payload and hundreds of requests to the baseline page load.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 89.4 | 66.4 | -23.1 pts | -25.8% |
| CPU Time | 6410 ms | 16420 ms | +10011 ms | +156.2% |
| Interaction Jank | 142 ms | 325 ms | +183 ms | +129.1% |
| Payload | 226 KB | 3559 KB | +3333 KB | +1476.6% |
| Requests | 131 | 652 | +521 | +399.1% |
Pubnation Link to section
Pubnation delivers strong overall performance preservation with a relatively low Lighthouse score penalty (-11.7%). Their streamlined delivery ensures that user interaction remains smooth (+86ms jank), though there is still some minor room for optimization in their CPU processing footprint.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 93.3 | 82.4 | -10.9 pts | -11.7% |
| CPU Time | 5039 ms | 14606 ms | +9567 ms | +189.9% |
| Interaction Jank | 84 ms | 170 ms | +86 ms | +103.0% |
| Payload | 225 KB | 1521 KB | +1296 KB | +575.4% |
| Requests | 67 | 397 | +330 | +496.0% |
Raptive Link to section
Raptive offers exceptional JavaScript execution efficiency, keeping CPU overhead remarkably low (+8.2 seconds) for a premium network. While their processing is highly optimized, they do introduce a sizable network payload and moderate interaction delays during the ad rendering phase.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 82.1 | 69.7 | -12.4 pts | -15.1% |
| CPU Time | 8211 ms | 16390 ms | +8180 ms | +99.6% |
| Interaction Jank | 92 ms | 348 ms | +257 ms | +280.2% |
| Payload | 377 KB | 3225 KB | +2848 KB | +755.5% |
| Requests | 73 | 567 | +494 | +680.5% |
Setupad Link to section
Setupad is one of the most efficient networks we tested, boasting extremely low JavaScript execution overhead (+7.4 seconds) and a highly optimized payload footprint. This lightweight delivery keeps request counts low, though slight interaction jank prevents it from perfectly preserving the control scores.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 87.8 | 72.0 | -15.8 pts | -18.0% |
| CPU Time | 7990 ms | 15351 ms | +7361 ms | +92.1% |
| Interaction Jank | 169 ms | 270 ms | +101 ms | +59.8% |
| Payload | 366 KB | 1767 KB | +1402 KB | +383.4% |
| Requests | 167 | 458 | +291 | +174.5% |
Snigel Link to section
Snigel manages to keep its CPU execution costs relatively contained (+12.1 seconds) while delivering a heavy monetization stack. Unfortunately, this comes at the expense of a massive network footprint (+3.09 MB) and severe interaction delays, resulting in the steepest Lighthouse score penalty in our benchmark.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 96.6 | 59.9 | -36.7 pts | -38.0% |
| CPU Time | 10420 ms | 22543 ms | +12122 ms | +116.3% |
| Interaction Jank | 86 ms | 408 ms | +322 ms | +374.1% |
| Payload | 173 KB | 3333 KB | +3159 KB | +1821.2% |
| Requests | 87 | 813 | +726 | +835.1% |
Venatus Link to section
Venatus provides a deeply comprehensive and feature-rich monetization stack tailored for specific publisher verticals. This heavy functionality brings the highest CPU processing load of any network tested (+25.6 seconds), creating significant interaction delays and a severe impact on baseline performance scores.
View Technical Metrics
| Metric | Control (Without Ads) | Test (With Ads) | Average Impact | % Change |
|---|---|---|---|---|
| Lighthouse Score | 91.5 | 60.6 | -31.0 pts | -33.8% |
| CPU Time | 6578 ms | 32215 ms | +25636 ms | +389.7% |
| Interaction Jank | 98 ms | 431 ms | +333 ms | +339.1% |
| Payload | 320 KB | 2845 KB | +2525 KB | +788.3% |
| Requests | 99 | 671 | +572 | +578.6% |
Testing Methodology Link to section
To produce these benchmarks, we ran Lighthouse performance tests against each site under two conditions:
- Control Phase (Without Ads): Ad network scripts and bidders were completely blocked to establish a baseline performance profile.
- Test Phase (With Ads): The ad network was allowed to load normally.
Our testing process involves a comprehensive user flow to accurately capture real-world performance:
- Navigation & User Flow: We navigated to a representative page on the site and allowed the initial page load to settle.
- User Interaction: We simulated typical user interactions (such as scrolling) to trigger lazy-loaded ad units and accurately measure interaction jank (e.g., maximum potential interaction delays).
- Network Payload: Throughout the entire navigation and interaction flow, we captured all network activity to determine the total byte weight and request count introduced specifically by the ad network.
The "Impact" metrics represent the average difference between the Test and Control phases across all sites we tested for that specific provider. These numbers are aggregated averages and highlight the true cost of ad execution beyond the baseline site load.