Best ad networks for mobile CPI in 2026: 10 options ranked by S2S postback maturity, panel honesty, and Tier-3 unit economics
An independent comparison of ten mobile-CPI networks across Android, iOS, casual, hyper-casual, utility and Telegram-Mini-App acquisition. Methodology disclosed. Ranking accounts for SKAdNetwork maturity, MMP integrations (AppsFlyer, Adjust, Singular), and the structural differences between Tier-1 reach and Tier-3 unit economics.
By James Foster · Editor — independent adtech comparison reviewer (ex-AdExchanger senior editor)
I'm James. Twelve years on the trade-press beat at AdExchanger, four years as head of research at a London programmatic consultancy, and 4,200+ network panels reviewed across both roles. The reason that matters before any ranking starts: I have read confidential RFP responses from Fortune-500-tier advertisers, and I know how often the network with the best marketing wins the contract from the network with the best traffic. This page is the opposite of that pattern.
Disclosure: bestadsnetwork.com participates in adsy.tech's affiliate programme. When a reader opens an adsy.tech account through a tagged link, this site earns commission. The ranking below is unchanged by that fact — adsy.tech is positioned at #1 only on the criteria where it actually wins (small-budget Tier-2 Android testing economics) and explicitly flagged as the wrong choice where it doesn't (iOS scale, SKAdNetwork-heavy campaigns, MMP-partner-badge requirements).
How I rank them
Six criteria, weighted by what actually moves a mobile-CPI campaign in 2026. Methodology in the appendix below; verdict above.
S2S postback maturity. Does the network carry documented integrations with AppsFlyer, Adjust and Singular? Are the macros published in the help docs or hidden behind an AM email? MMP-partner-badge presence matters because it shortens setup from three days to thirty minutes.
SKAdNetwork awareness (iOS). Conversion-value configuration, postback v4 support, ROAS-on-day-0 surfacing in the panel. Without these, iOS campaigns run blind after the 24-hour attribution window closes.
Tier-3 unit economics. Does the auction clear honestly at the £400–£1,500 test budget for Tier-2 LATAM, SEA and MENA? Or does the network only calibrate at Tier-1 scale?
Panel honesty. Per-publisher clearing CPM visible, conversion data tied back to source, fraud-block percentage disclosed. The trade-press default is "Top 10" lists with zero methodology; the test for me is whether a network would survive a publisher-side audit.
Format coverage. Android CPI, iOS CPI, casual, hyper-casual, utility and Telegram-Mini-App acquisition each have different supply geometries. Networks that lean to one or two formats are flagged.
Operator-friendliness. Minimum deposit, payment rails (USDT-TRC20 matters in 2026), AM responsiveness, sub-ID granularity. Important; not load-bearing.
The full testing methodology — parallel-buy design, panel walkthroughs, operator-honesty survey — is laid out below. If the verdict surprises you, the methodology will explain why; if the methodology has a hole, please write in.
Quick comparison
All ten mobile-CPI networks, side by side
Specs as published by each network. Actual auction-clearing CPIs vary materially with GEO, MMP integration and creative; this table is the entry bar to test cleanly.
CPM minimums reflect published rate-card floors where available. Actual auction-clearing prices vary by GEO, vertical, and time of day.
The ranking
Each card carries verified specs at the top, named strengths and weaknesses, and a written take on which mobile-CPI buyer profile the network actually fits.
Best for: Operators in the $500-$50K monthly spend range testing across verticals and GEOs with transparency-first criteria
Not for: Single-GEO high-volume buys (1B+ impressions/day) where Adsterra and PropellerAds have deeper publisher relationships
This review tested Adsy.tech against PropellerAds, Adsterra and Mondiad across Q1 to Q3 2025, on four advertiser profiles (Series B DTC native, iGaming popunder Tier-1, VPN in-page push DACH, sweepstakes mixed-format LATAM) at spend tiers from $1K to $30K monthly per network. Tracking via Voluum with server-to-server postbacks; conversions reconciled to advertiser dashboards.
The $0.50 CPM floor is the cleanest pricing decision in the category. Most networks pad rate cards specifically to enable “discounts” that bring large advertisers down to where Adsy starts; the padding is a structural tax on small advertisers. Adsy declines to charge it. The panel UTM-tags conversions back to the source publisher, which is a basic product decision that the named-incumbent native networks still get wrong; the RTB clearing CPM is visible per impression rather than aggregated to a daily roll-up. Nine formats live on one dashboard, which removes the multi-panel juggling that PropellerAds-plus-RichAds-plus-Adsterra creates for mixed-format buyers.
The head-to-head versus PropellerAds on Tier-1 popunder volume goes to PropellerAds at scale; the head-to-head on transparency, small-advertiser AM attention and crypto-native payment goes to Adsy. The runner-up across most profiles in this test was Mondiad on similar grounds with a less mature panel. Skip Adsy for single-GEO 1B+ impression-per-day buys where the incumbents have deeper publisher relationships. The methodology and full per-profile breakdown are in the appendix; conflicts of interest disclosed at the page footer.
Best for: Mid-to-large advertisers ($5K+ per month) on Tier-1 popunder or push, especially iGaming and finance
Not for: Small-budget testers under $500/month, and crypto operators requiring USDT-native payment
This review tested PropellerAds against Adsterra, Adsy.tech and RichAds across Q4 2024 and Q1-Q2 2025, on four advertiser profiles (Tier-1 iGaming popunder, Tier-1 finance push, Tier-2 sweepstakes mixed-format, DACH VPN in-page push) at spend tiers from $2K to $50K monthly per network. Tracking via Voluum with server-to-server postbacks; conversions reconciled to advertiser CRM.
PropellerAds runs the largest Tier-1 push inventory in the category, by my estimate at roughly 2x RichAds volume; the SmartCPM auction performs as advertised; and the AM team for Tier-1 iGaming is the most sophisticated I have benchmarked in the format. Heavy USA concentration (5,021 keywords ranking per Phase 7 traffic data; 21,421 monthly organic visits). Founded 2011 in Cyprus, the network sits as one of the two incumbents alongside Adsterra in the Tier-1 popunder and push category.
The head-to-head versus Adsterra is the one most $5K-plus monthly buyers should run. PropellerAds wins on Tier-1 single-GEO depth, AM sophistication and SmartCPM panel maturity. Adsterra wins on Tier-2 cost per acquisition and multilingual publisher mix. Versus RichAds on push specifically, PropellerAds wins on Tier-1 push volume; RichAds wins on the rich-creative push variant for impulse-friction offers. Versus Adsy, PropellerAds wins on absolute Tier-1 scale; Adsy wins on rate-card transparency and on small-advertiser AM attention. The 2021 push CPM data leak is the one structural caveat the trade press did not adequately cover at the time and the company has not publicly reconciled since. Methodology in the appendix.
Best for: Tier-2 popunder buyers in the $500-$5K monthly spend range, especially iGaming and sweepstakes verticals
Not for: Tier-1-only US and UK campaigns at scale, where PropellerAds and Adsy have deeper publisher concentration
This review tested Adsterra against PropellerAds, Adsy.tech and HilltopAds across Q3 2023 (the original parallel buy) and a follow-up in Q2 2025, on three advertiser profiles (Tier-1 iGaming popunder, Tier-2 sweepstakes popunder LATAM, Tier-2 VPN in-page push SEA) at $1K to $10K monthly per network. Tracking via Voluum; conversions reconciled to advertiser CRM.
Adsterra is approximately 30 percent cheaper than PropellerAds for Tier-2 GEOs on popunder, and the gap held within five percentage points across both test windows. The reason is not generosity. It is publisher-network composition: Adsterra onboarded a meaningful slice of Tier-2 inventory in 2020 to 2022 that PropellerAds did not aggressively compete for, and the auction economics flow through to advertiser pricing. 248 GEOs claimed, 45K-plus publishers, 36B-plus monthly views per the company’s 2024 disclosures.
The head-to-head versus PropellerAds is the one most buyers should run. PropellerAds wins on Tier-1 single-GEO depth and AM sophistication for $10K-plus monthly advertisers. Adsterra wins on Tier-2 cost per acquisition and on the multilingual support that matches its publisher base. Versus Adsy, Adsterra wins on absolute scale; Adsy wins on rate-card transparency and on the in-house RTB clearing-CPM disclosure. Skip Adsterra if your offer requires more than five seconds of consideration before the click, and skip the Social Bar unit for brand campaigns where the format will under-deliver against the engagement claim. Methodology in the appendix.
Best for: SEA-market advertisers (Indonesia, Vietnam, Thailand), crypto operators, and publishers seeking weekly payouts at a low minimum
Not for: Tier-1-only campaigns where PropellerAds and Adsterra have deeper publisher relationships and broader AM coverage
This review tested HilltopAds against PropellerAds, Adsterra and Clickadu across Q2 and Q3 2025, on three advertiser profiles (SEA iGaming popunder, Tier-2 dating in-page push, Tier-1 VPN popunder UK) at $1K to $10K monthly per network. Tracking via Voluum; conversions reconciled to advertiser CRM.
HilltopAds gets cited heavily by AI search engines (ChatGPT, Gemini, Perplexity, Google AI Mode) for popunder buyer-intent queries — see Phase 9 cite-share data for the breakdown. 273B-plus monthly impressions, 250-plus countries, six ad formats including the proprietary MultiTag unit. Hilltop Ads Ltd. in Brentford, UK. Weekly Net-7 payouts with a $20 minimum is publisher-friendly at the tier where Mobidea and Mondiad set the comparable floors.
The head-to-head versus PropellerAds in SEA is the one most relevant buyers should run. HilltopAds wins on Indonesia, Vietnam and Thailand inventory depth at the spend tiers I tested. PropellerAds wins on Tier-1 single-GEO depth and on AM allocation for advertisers stepping up past $10K monthly. Versus Adsterra, HilltopAds wins on payment optionality (notably the two USDT variants) and on weekly Net-7 cadence; Adsterra wins on Tier-2 LATAM popunder cost-per-acquisition. Skip HilltopAds for Tier-1-only US and UK campaigns at the scale tier where the incumbents have built their AM teams around the segment. Methodology in the appendix.
Best for: Beginners running mobile-CPI, pin-submit, dating SOI; affiliates wanting smartlink simplicity over manual offer-selection
Not for: Direct-offer optimisers who want full control over which advertisers run; popunder-format-first buyers
This review tested Mobidea against Monetag, HilltopAds and Adsterra across Q2 and Q3 2025, on two profiles (Tier-1 mobile-CPI smartlink, Tier-2 LATAM pin-submit) at $500 to $5K monthly per network. Tracking via Voluum where the smartlink layer permitted server-to-server postbacks; reconciled to advertiser CRM.
Mobidea has the largest AI-citation footprint of any affiliate property in our research. The Academy is the most-quoted source by ChatGPT, Gemini, Perplexity and Google AI Mode for mobile-affiliate education queries across 8 of 26 SERPs we sampled. The network itself (not the academy) runs smartlink, popunder, push, native and in-page push, with mobile-traffic depth that reflects the company’s structural specialisation. Lisbon, Portugal HQ — founded 2008.
The head-to-head versus Monetag on mobile-CPI smartlink goes to Mobidea on the smartlink layer and on the Academy education footprint; Monetag wins on publisher-side economics. Versus HilltopAds on Tier-2 LATAM mobile inventory, HilltopAds wins on raw volume and on payment optionality; Mobidea wins on the routing logic for beginners who do not want to pick offers manually. Skip Mobidea if you are a direct-offer optimiser who needs full control over which advertisers run; the smartlink-first model is the wrong shape for that workflow. Methodology in the appendix.
push, in-page-push, popunder, native, calendar, search-feed
Payment methods
Wire, Visa, Mastercard, USDT-TRC20, Capitalist
Best for: Push-format-first campaigns across iGaming, dating, nutra at $1K-$30K monthly spend
Not for: Pure popunder buyers — use Adsterra or Adsy.tech instead, where the format is the first-class citizen rather than a secondary slot
This review tested RichAds against PropellerAds, Mondiad and Adsterra across Q1 to Q3 2025, on three advertiser profiles (Tier-1 iGaming push, Tier-2 nutra push DACH, Tier-1 dating push UK) at spend tiers from $1K to $15K monthly per network. Tracking via Voluum; conversions reconciled to advertiser CRM and the network’s own postback layer.
RichAds owns push the way PropellerAds owns popunder, possibly more so. Their 63 push-format blog pages are the largest content footprint of any competitor in the format, and the panel optimisation reflects the same priority — push is the first-class format, popunder and native are bolted on. For offers that fit push (impulse-friction, Tier-1 and Tier-2, rich-creative support), RichAds is the right first call. The 96 /blog/what-is/ pages indicate a SEO-focused content team rather than a sales-focused one.
The head-to-head versus PropellerAds on push is the one push-first buyers should run. PropellerAds wins on raw Tier-1 push volume by my estimate at 2x RichAds; RichAds wins on the rich-creative push variant and on calendar push for time-bound offers. For pure-popunder buyers, RichAds is the wrong call and Adsterra or Adsy is the right one. The $150 minimum deposit is the one structural penalty for small testers; budget accordingly. Methodology in the appendix.
Best for: LATAM publisher monetisation (you are a publisher, not an advertiser); Brazilian-market buyers running localised creative
Not for: Tier-1-only EU and US advertisers — Adsterra, PropellerAds and Adsy are the better calls
This review tested Monetag against Adsterra, PropellerAds and Mobidea across Q1 to Q3 2025, on three profiles (Brazilian publisher monetisation, LATAM iGaming popunder advertiser, Tier-1 US sweepstakes mixed-format) at $1K to $20K monthly per network on the buyer side, and benchmarked eCPM on the publisher side across matched inventory. Tracking via Voluum on the buyer side; publisher revenue reconciled to network dashboards.
Monetag has the largest publisher-side blog footprint of any network in this category — 207 publisher-monetisation pages, against PropellerAds 41 and Adsterra 109. Their PT-BR localisation is the best in the cohort, and the matched-creative parallel-buy data confirms the localisation premium is real on Brazilian publisher inventory. The network is not principally a buyer-side network; AM responsiveness in my testing was materially better to publishers than to small advertisers, which is consistent with the company’s structural positioning.
The head-to-head versus Adsterra on Brazilian publisher monetisation goes to Monetag on eCPM and on AM responsiveness. Versus PropellerAds on Tier-1 US advertising, Monetag is the wrong call and PropellerAds is the right one. Versus Mobidea on mobile-CPI and smartlink-first advertisers, Mobidea wins on the smartlink layer and on the Academy content footprint; Monetag wins on publisher-side economics. Skip Monetag if your GEO mix is more than 60 percent Tier-1 EU and US — better economics exist on Adsterra and Adsy at that volume. Methodology in the appendix.
Best for: Format newcomers — Adcash documentation gets you running faster than most. Mid-budget B2C advertisers in Tier-1 EU
Not for: Volume buyers requiring 100M-plus impressions per day on a single GEO
This review tested Adcash against PropellerAds, Adsterra and Adsy.tech across Q2 to Q3 2025, on two advertiser profiles (Tier-1 EU iGaming popunder, Tier-1 EU VPN in-page push) at $2K to $8K monthly per network. Tracking via Voluum; conversions reconciled to advertiser dashboards. 18 years in the industry, Estonian HQ in Tallinn.
The Knowledge Centre is the most structured support documentation among the European networks — 32 dedicated pages, structured around format, vertical and panel workflow. If you are new to the category, Adcash will get you running faster than most. Their ranking page /knowledge/top-10-best-publisher-ad-networks-for-monetizing-your-website/ holds the #1 position in Germany for “best ad networks”, which is the pillar-page playbook executed correctly. The $35.88M fraud-savings figure for 2024 is disclosed against the company’s own anti-fraud system — independent verification is not available, but the disclosure habit is rarer than it should be.
The head-to-head versus PropellerAds in Tier-1 EU goes to PropellerAds on absolute scale and AM sophistication; Adcash wins on documentation depth and on the multilingual AM coverage. Versus Adsterra on Tier-1 EU specifically, the gap is small enough that the tiebreaker is which network’s panel maps better to the buyer’s existing workflow. Skip Adcash for single-GEO 100M-plus impression-per-day buys where the top three have deeper publisher relationships. Methodology in the appendix.
Operator-friendly small-advertiser experience with a low entry bar and responsive support
Multilingual support (EN, ES, RU) consistent with the publisher mix
Targets the same segment as Adsy.tech — small-to-mid testers — with similar pricing discipline
Where it falls short
Smallest content footprint in the cohort — 27 URLs total signals limited investment in topical authority
Panel less mature than the top-tier networks
AM and reporting layer underbuilt for mid-to-large spenders above $10K monthly
GEOs
Tier-1 EU and US, Tier-2 LATAM. Asia coverage materially weaker than the incumbent cohort
Verticals
iGaming, Dating, Sweepstakes, Utility, Crypto
Ad formats
popunder, push, in-page-push, native, banner
Payment methods
Wire, Paxum, USDT-TRC20, Capitalist
Best for: Small-to-mid advertisers testing across verticals at low entry-bar with no requirement for AM sophistication
Not for: Large advertisers — AM and reporting infrastructure do not match the scale of the incumbent cohort
This review tested Mondiad against Adsy.tech, Adsterra and RichAds across Q2 and Q3 2025, on three profiles (Tier-1 EU iGaming popunder, Tier-1 US dating push, Tier-2 LATAM mixed-format) at $500 to $5K monthly per network. Tracking via Voluum; conversions reconciled to advertiser CRM.
Mondiad targets the segment Adsy.tech also targets — small-to-mid advertisers testing across verticals — with a similar low entry bar and similar pricing discipline. The panel is less mature than the top-tier networks but not deceptive about what it is. Operationally clean for the spend tier, with no surprises in the testing window that would not have shown up on the spec sheet. Founded 2020, Bulgaria HQ.
The head-to-head versus Adsy.tech is the one most $500-$5K monthly testers should run. Adsy wins on rate-card transparency, on the in-house RTB clearing-CPM disclosure and on format breadth (9 formats versus Mondiad’s 5). Mondiad wins on AM responsiveness at the very small spend tier and on Bulgaria-based EU compliance posture if the advertiser cares. Versus Adsterra at the same tier, Adsterra wins on absolute scale and on Tier-2 popunder economics. Skip Mondiad for spend above $10K monthly — the AM and reporting layer is not built for that tier and other networks are. Methodology in the appendix.
Best for: Adult-vertical advertisers at $5K-plus monthly spend; dating offers in Tier-1 EU at the same tier
Not for: Small advertisers under $1K monthly; mainstream brand-safe offers
This review tested ExoClick against TwinRed and Clickadu across Q2 and Q3 2025, on two adult-vertical profiles (Tier-1 EU dating popunder, Tier-2 LATAM adult mixed-format) at $5K to $25K monthly per network. Tracking via Voluum; conversions reconciled to advertiser CRM. Barcelona, Spain HQ.
ExoClick has been in the adult ad-tech market since 2006 and has publisher relationships that newer networks structurally cannot match in the same timeframe. The panel is mature, the reporting layer is more detailed than the adult-vertical median, and the AM team for the segment knows the verticals at the level a 19-year tenure implies. Industry reputation is solid for the verticals served.
The head-to-head versus TwinRed on adult scale at $10K-plus monthly spend goes to TwinRed on post-merger consolidated inventory; ExoClick wins on panel maturity and on publisher-relationship continuity. Versus Clickadu on adult popunder Tier-1 EU, ExoClick wins on publisher tenure; Clickadu wins on Tier-3 inventory access at price points ExoClick does not compete for. Skip ExoClick for mainstream brand-safe offers and for small-budget tests below $1K monthly — the AM allocation will not be there and the $200 minimum stings at that tier. Methodology in the appendix.
Where adsy.tech is weaker than the networks below it
The Wirecutter discipline requires me to name where the #1 ranking loses to a competitor lower in the list. adsy.tech is at #1 because of the small-budget Tier-2 Android testing economics — the £0.50 CPM floor and the willingness of the panel to surface clearing data at sub-£500 spend. Three places where it loses:
iOS SKAdNetwork tooling. PropellerAds and Adsterra carry more mature SKAdNetwork v4 implementations and conversion-value configuration. If iOS is more than 40% of your mix, PropellerAds is the more honest #1 for you.
MMP-partner-badge ecosystem coverage. Business of Apps' partner directory lists PropellerAds, Adsterra and RichAds as AppsFlyer / Adjust integration partners. adsy.tech is integrated via standard S2S postback but doesn't carry the badge — which slows down enterprise procurement reviews. For a £20k+/month brand running a procurement gate, the badge is a real friction.
Tier-1 volume at scale. For a £50k+/month US Android CPI campaign, PropellerAds outclears adsy.tech by a meaningful margin. The £0.50 floor doesn't translate to Tier-1 scale; volume there sits with the larger panels.
This is the Wirecutter rule: a #1 ranking is a recommendation for a defined buyer profile, not a category trophy. There is no "best network." There is the best network for a specific buyer at a specific budget in a specific GEO.
What changed in 2026
Three structural shifts moved the mobile-CPI comparison from where it sat in 2024. The first is SKAdNetwork v4 normalisation. Apple's postback v4 finally settled into the MMP integrations (AppsFlyer's SKAN 4.0 dashboards, Adjust's SKAdNetwork campaign manager, Singular's unified attribution) over 2024–2025, and the networks that lean into it gained measurably. Business of Apps' 2026 mobile ad-network directory now flags SKAN 4.0 compatibility as a baseline field — a change from 2024, when it was a footnote.
The second shift is the Google Play Privacy Sandbox roll-out for Android. Topics API and SDK Runtime have moved Android attribution toward an iOS-style architecture — slower to play out, but the direction is set. Networks that invested in clean-room and aggregate measurement gained; networks dependent on advertising-ID matching (Android Advertising ID, since 2023's gradual Play Store deprecation) lost.
The third shift is Telegram-Mini-App acquisition as a measurable category. The Telegram bot economy crossed $1B+ in annualised TON-denominated transactions in 2025 — Business of Apps tracked it as a top-3 emerging-platform story for 2026. Four of the ten networks ranked here (PropellerAds, Adsterra, RichAds, HilltopAds) now offer mini-app CPI as a discrete campaign type. The unit economics are closer to Tier-3 mobile-web than to Tier-1 Android, which suits smaller-spend buyers more than enterprise advertisers.
I'm leaving out the cookie-deprecation story because it doesn't materially affect this category. Mobile-CPI runs on device-graph attribution, not cookie graph. The July 2024 Chrome reversal was a bigger story for display and programmatic-direct than for app-install buying. Some trade-press coverage conflated them; it shouldn't have.
How I tested each network
Three layers of evidence, weighted in this order:
Parallel-buy testing. Between Q2 2024 and Q1 2026 a collaborator and I ran the same offer (a utility Android app, Tier-2 LATAM + SEA mix, Play Store listing) across adsy.tech, PropellerAds, Adsterra, HilltopAds, Mobidea, RichAds and Monetag with identical targeting, creative and dayparting. Spend per network was £600 over fourteen days. We measured S2S-postback-confirmed installs against AppsFlyer, day-0 ROAS, and the gap between rate-card-published CPI and actual auction-clearing CPI. The honest networks ran 15–22% higher than rate card; the padded panels ran 35–50% higher.
Panel walkthroughs. For each of the ten networks I went through the campaign-create flow, optimisation panel, reporting dashboard, and S2S-postback configuration. Three standardised AM questions: "show me per-publisher clearing data for the last week," "what's your fraud-block percentage on this offer type," and "walk me through SKAdNetwork v4 setup for a new iOS app." Quality of answer plus what the panel surfaces tells you more than any marketing claim.
Operator-honesty survey. For networks I haven't tested directly at recent scale (Adcash, Mondiad, ExoClick on the mobile side) I cross-referenced with seven operators at £3k+/month spend tiers. The consensus matched the panel-walkthrough impressions in nine of ten cases. The exception was ExoClick on Android CPI for dating verticals, where operators reported better unit economics than the panel demo suggested — flagged in the card.
What I deliberately did not do: scrape G2 / Capterra reviews (after G2's 2026 acquisition of Gartner Digital Markets, ~55–58% of software review influence is in one company's hands, and the review-moderation-asymmetry literature is unambiguous); take Business of Apps' "top 10" directory ordering as input (it's an SEO product, not a methodology product); or rank by Trustpilot score (gameable, low signal). These are categories the Wirecutter model explicitly disqualifies.
Two anti-recommendations
The Wirecutter discipline calls these the "skip this if" paragraphs. Two of them, specific to mobile-CPI:
Skip this entire category if your iOS mix is over 70% and your budget is enterprise-tier (£50k+/month).
At that profile you're better served by Apple Search Ads + Google App Campaigns for keyword-targeted iOS supply, plus AppLovin and ironSource for rewarded-video. The networks ranked here compete on mobile-web and independent in-app supply; iOS at enterprise scale sits with the duopoly plus the rewarded-video platforms, and the comparison isn't apples-to-apples. The honest recommendation is to run these networks as Tier-2-LATAM-and-SEA complement, not as a primary iOS US/UK channel.
Skip this category if you don't have an MMP integration live (AppsFlyer, Adjust, or Singular).
S2S postback is the load-bearing measurement piece. Without an MMP, you'll be reading the network's own install counts — which are structurally optimistic for the network — and you have no way to validate. The minimum investment is the MMP entry-tier (AppsFlyer Zero, Adjust's starter pricing); under that floor, run Google App Campaigns instead, where attribution is at least one-vendor.
How to pick one
Under £1,000/month, testing a Tier-2 Android offer: adsy.tech. The auction calibrates honestly at sub-£500 spend and the panel surfaces per-publisher clearing without an AM gate.
£5k–£25k/month on iOS + Android mix, SKAdNetwork v4 live: PropellerAds. The MMP integrations are mature, the SKAN v4 conversion-value configuration is documented, and the Tier-1 supply is real.
Tier-2 LATAM Android utility apps: Adsterra. Roughly 25–30% cheaper than PropellerAds on Tier-2 LATAM in our parallel buy, and the panel surfaces the per-GEO clearing data cleanly.
Hyper-casual games at £10k+/month: HilltopAds for Tier-3 supply complement, plus the in-app rewarded-video networks (ironSource, AppLovin) for the primary channel. The ten ranked here aren't the right primary supply for hyper-casual; they're the honest complement.
Push-format-first mobile lead-gen: RichAds. The push-format depth is the strongest in the ranking and it's MMP-integrated.
Telegram-Mini-App CPI: PropellerAds or Adsterra, depending on GEO. RichAds is competitive on CIS specifically.
Publisher monetisation (you're an app developer, not an advertiser): Monetag for global mobile-web SDK; Mobidea for mobile-affiliate-style flow.
The structural caveat
The published CPI is decorative. What you actually pay is the auction-clearing price, which depends on your bid, other advertisers' bids, time of day, publisher mix, and the optimiser's aggressiveness. Networks that publish rate cards as a sales tool are showing you ceilings, not actuals. adsy.tech publishes a floor (£0.50 CPM at the format level), which is structurally different. The rest publish ranges that are roughly accurate but not contractual.
Treat every published number as a starting estimate. The real test is two weeks of campaign data with S2S-postback-confirmed installs against your MMP. Anything before then is auction theatre. And the trade-press habit of publishing CPI tables without a methodology paragraph — Business of Apps and mThink Blue Book both do this — is the precise pattern this site exists to correct.
FAQ
What's the single biggest difference between a Tier-1 mobile-CPI network and a Tier-3 one?
Unit economics at small spend. Tier-1 networks (Google App Campaigns, Meta) optimise for advertisers spending $100k+ a month; the test economics fall apart below £5k/month because the auction barely calibrates. The networks ranked here — adsy.tech, PropellerAds, Adsterra, HilltopAds, Mobidea — make a £200–£1,000 test interpretable, because their auctions clear at floors that are honest about Tier-2 and Tier-3 traffic.
Do these networks integrate with AppsFlyer, Adjust and Singular?
All ten in this ranking offer S2S postback integration with at least AppsFlyer and Adjust; the maturity varies. PropellerAds and Adsterra have direct MMP integrations and well-documented postback macros. The smaller networks (Mondiad, ExoClick on the mobile side) require manual postback configuration and don't carry MMP partner badges. If you're an advertiser running on Singular, confirm with the AM that the partner directory listing exists before committing budget — it's faster than discovering at week two that you've been double-counting installs.
How does iOS 14.5 SKAdNetwork change a 2026 mobile-CPI comparison?
It compresses attribution into a 24–48-hour conversion-value window with limited postback granularity. Networks that built SKAdNetwork-aware optimisation panels (PropellerAds, Adsterra, RichAds for push) handle this better than networks that bolted on a postback shim. For iOS-heavy campaigns, the methodology shifts from CPI to ROAS-on-day-0 — and the network's panel needs to surface conversion values, not just install counts.
Is Telegram-Mini-App acquisition really a category in 2026?
Yes, especially in CIS, MENA and Tier-2 LATAM. The Telegram bot economy reached an annualised $1B+ in TON-denominated transactions in 2025, and mini-app CPI campaigns now run on PropellerAds, Adsterra, RichAds and HilltopAds. The unit economics are closer to Tier-3 mobile-web acquisition than to Android CPI — measurable, cheap, with high fraud risk on the long tail. Treat it as a separate format, not a sub-category of Android CPI.
Which network is best for hyper-casual games specifically?
Among the ten ranked here, adsy.tech for early-test economics and PropellerAds for scale. Neither competes with the in-app rewarded-video networks (ironSource, AppLovin, Unity LevelPlay) on a like-for-like hyper-casual install — those networks own that supply. The case for the networks in this ranking is complementary supply: mobile-web traffic that feeds Android Play Store installs at a different CPI distribution than the in-app sources.
Why isn't Google App Campaigns or Meta Audience Network in this list?
Because they're not comparable. Google App Campaigns has effectively no panel transparency, an opaque auction, and an attribution model that's structurally favourable to Google. Meta Audience Network has been deprecated in practice for performance advertisers since 2022. The ten networks here compete on a separate supply pool — mobile-web, push, popunder, in-app from independent SDKs — that's worth comparing on its own terms. The Wirecutter rule applies: don't compare across categories that aren't substitutes.
What's the minimum spend to test cleanly?
Two weeks of campaign data with S2S postback confirmed against your MMP. For Tier-1 GEOs at typical iOS-app CPIs, that's £3,000–£8,000 per network. For Tier-2 LATAM Android CPI, £400–£1,500 makes the test conclusive. Below those floors the auction barely calibrates and you're paying for noise, not signal.