TL;DR

• Mobile traffic arbitrage is buying ad impressions on mobile placements and reselling the resulting actions to an advertiser under CPI, CPA, or RevShare at a margin.
• Profit is the spread between traffic cost and payout, which makes ROI — not click volume — the number that decides whether a funnel lives or dies.
• The minimum starting kit is an ad network account, an offer from an affiliate network, a tracker, and one funnel combining creative, geo, prelander, and destination.
• AdsGram runs campaigns across Telegram Mini Apps, channels, and bots with CPM, tCPC, and tCPA models; there is no minimum deposit, and the documentation recommends 100–250 USD₮ for a meaningful first test.
• Attribution runs on the {click_id}, {campaign_id}, {banner_id}, {publisher_id}, and {record_data} macros, plus a conversion confirmation API or integrations with AppsFlyer, Adjust, AppMetrica, Keitaro, and Kochava.

Mobile is still where advertiser money and user attention meet, and it is getting more expensive to buy into. Sensor Tower's State of Mobile 2026 puts global in-app purchase revenue at $167 billion for 2025, up 10.6% year over year, while average cost per install in Q1 2026 reached $5.84 on iOS and $1.92 on Android. Rising acquisition costs reward the people who can read unit economics and find underpriced inventory — which is the entire discipline behind mobile traffic arbitrage.

This article covers what the model is, how a funnel is assembled, which payout structures exist, how to track conversions when platform signal is limited, and how to launch a first campaign step by step. Start with the definition, because a lot of beginner losses come from confusing arbitrage with regular media buying.

What mobile traffic arbitrage is

Mobile traffic arbitrage is a model in which you buy ad impressions on mobile placements and sell the resulting target actions to an advertiser for more than the traffic cost you. The arbitrage is the price gap: one price for an impression, another for a result.

The vocabulary you need to know:

• Offer — the advertiser proposition you get paid for: an install, a registration, a first deposit, a purchase.
• Affiliate network (CPA network) — the marketplace that aggregates offers and pays commissions.
• Funnel — the working combination of traffic source, geo, creative, prelander, and offer. You test and scale funnels, not individual creatives.
• Media buyer — the person who purchases traffic and manages funnels.
• Hold — the delay between a conversion firing and the network releasing your payout.

The difference from standard performance marketing is who carries the risk. In performance marketing you spend the product owner's budget; in arbitrage you spend your own, and every failed test comes out of your pocket. That single fact should shape how you size tests.

How a mobile arbitrage funnel works

A funnel is a five-link chain, and a break in any link zeroes out the profit. Understanding the chain matters more than memorising dashboard settings.

1. Traffic source. You buy impressions on CPM or clicks on CPC inside an ad network.
2. Creative. A banner, video, or native post that filters out non-target users before they cost you a click.
3. Prelander. An intermediate page, channel, or bot that warms the user up and sets expectations before the offer.
4. Destination. The offer landing page, an App Store or Google Play listing, or a Telegram Mini App.
5. Postback. The server-side signal confirming the conversion was credited to you.

The operating principle is that every link is measured separately. High CTR with weak install conversion points to a creative-to-landing mismatch, not a bad source. Installs with no payouts point to an attribution problem or a quality rejection, not a creative problem.

Payout models in mobile traffic arbitrage

The payout model defines what you pay for on the way in and what you earn on the way out, and profit exists only when entry cost stays reliably below payout.

Start on CPI or CPL. The conversion cycle is short, data accumulates fast, and you can kill a losing funnel within days. RevShare demands a cash cushion because revenue arrives over months, and a beginner rarely has the runway to wait for it.

The AdsGram platform offers three optimisation models — CPM, tCPC, and tCPA. One nuance worth internalising before your first launch: when tCPC is enabled, billing still happens per 1,000 impressions. tCPC sets the target click price the system bids toward, not the unit you are charged for. tCPA behaves the same way — you set the desired cost per conversion, the system optimises delivery toward it, payment stays on CPM, and conversion tracking via postback has to be configured first.

Where to buy mobile traffic

Mobile traffic sources differ in entry cost, moderation strictness, and audience intent, and those three factors decide which verticals can survive on each source. There is no universally best channel, only the right fit for a specific offer.

For a first funnel, in-app and Telegram inventory are the practical entry points: the test budget is small, moderation is workable, and you get placement-level data fast enough to learn something within a week.

Telegram as a mobile traffic source

Telegram gives mobile buyers a rare combination — billion-scale reach, native ad formats, and destinations that live inside the app itself, so the user never has to bounce through a browser. Telegram reported 1 billion monthly active users as of March 2025, and the ecosystem now includes Mini Apps that behave like full products.

Three things this changes operationally:

• A shorter path to conversion. Click → bot or Mini App → action. There is no browser-to-store-to-app gap, which is exactly where most mobile funnels leak.
• Native formats. Post, video, and task mechanics read as part of the interface rather than as interruptive banners, which holds CTR up longer before creative fatigue.
• Intent-adjacent targeting. Telegram Premium status and TON wallet ownership let you filter toward users who have already demonstrated willingness to pay, at the buying stage rather than after the install.

AdsGram operates both as its own ad network inside Telegram and as an official Telegram Ads reseller, so network buying and official Telegram Ads buying both run from one account. If you are still mapping the landscape of placements, our overview of Telegram ad platforms sets out how the options differ.

Mobile attribution under ATT and Privacy Sandbox

Deterministic click-level attribution is no longer guaranteed on mobile, which changes where an arbitrage funnel should send its traffic. On iOS, App Tracking Transparency and SKAdNetwork mean install-level data comes back aggregated and delayed, with limited conversion values. Android's Privacy Sandbox is moving in a comparable direction.

Two practical consequences for a media buyer:

• Store-destination funnels need an MMP. AppsFlyer or Adjust are doing the reconciliation work that raw click data can no longer do on its own. Budget for that from day one, not after the first campaign.
• In-ecosystem destinations preserve signal. A Telegram Mini App or bot as the destination keeps a deterministic click_id all the way through to the conversion event, because the whole path stays inside a single environment you control. For beginner funnels, that difference in measurement quality often matters more than a slightly cheaper CPI elsewhere.

How to start mobile traffic arbitrage

Starting in mobile traffic arbitrage takes six sequential steps, and skipping any of them turns a test into a write-off.

1. Pick one vertical and one or two geos. Narrow focus reaches statistical significance faster and keeps your reporting readable.
2. Source the offer. Confirm with the network: allowed traffic sources, hold period, quality requirements, and average CR on the offer.
3. Connect a tracker. Without one you cannot tell which placement converts. AppsFlyer, Adjust, AppMetrica, Keitaro, and Kochava are the standard mobile stack.
4. Build three to five creatives and one prelander. Test hypotheses, not button shades: different hooks, different pain points, different formats.
5. Set a test budget. Per AdsGram documentation there is no minimum deposit, and 100–250 USD₮ is the recommended range for evaluating network performance meaningfully.
6. Launch, collect data, cut the waste. The goal of a test is not profit. It is knowing which placements and creatives convert.

Pre-launch checklist:

• Offer confirmed and your traffic source approved by the network
• Tracker configured, with a test click traced end to end
• Postback fired and verified on a test conversion
• Creatives cleared moderation
• Stop-loss defined: the exact spend at which the funnel gets switched off

How to buy mobile traffic in AdsGram

Launching in AdsGram runs through eight steps in the advertiser dashboard and takes roughly five minutes once your creatives are ready. The field names below are the real ones, so you are not hunting blind.

1. Register. Create an advertiser account. If you are already connected as a publisher, sign in with the same login — that way AdsGram will not serve your own ads inside your own app.
2. Campaign type. Choose the format: Post (an image or GIF with text styled as a Telegram message), Video, or Task (a native block in a task list).
3. Legal information. This section applies only to campaigns targeting users in Russia. ERID support is enabled through AdsGram support, and the service neither generates nor validates the token.
4. Targeting. Available parameters are interface language, region with country include or exclude, VPN settings, platform (iOS/Android), Telegram Premium status, and wallet owners. You can also block placements by Platform ID — up to 50 per campaign, which is your blacklist mechanism.
5. Audience. Segments can be included or excluded, with a limit of 10 segments per campaign.
6. Strategy. Once targeting and the daily budget are set, AdsGram suggests a recommended CPM, and a recommended tCPC if you toggle that model. Recommendations depend on the targeting you filled in. The Set max RF value button caps impressions per user over a 7-day window — essential for not burning a narrow audience.
7. Banners. A campaign holds up to 10 banners. Treat that as a built-in A/B test and upload several hypotheses at once rather than one creative at a time.
8. Launch. After moderation the status changes to Ready to start, and delivery begins once the balance is positive.

On iteration speed, one rule saves days: editing a creative sends the campaign back into moderation, while strategy and targeting changes do not require re-moderation. So adjust bids and geo freely, and batch your creative changes instead of trickling them in.

Tracking and postbacks for mobile traffic

Without correct attribution, arbitrage is a lottery: you see spend but not which placement pays it back. AdsGram tracking runs on macros the platform substitutes into your link automatically.

Available macros: {campaign_id}, {banner_id}, {publisher_id}, {click_id}, {record_data}.

An example campaign URL:

https://your-tracker.com/?campaign={campaign_id}&banner={banner_id}&pub={publisher_id}&clickid={click_id}&record={record_data}

What not to miss:

• {record_data} is required for web landing links, Keitaro, and Kochava.
• {campaign_id} is required for Telegram ecosystem landing links.
• If a required macro is missing, conversion tracking fails outright — it does not degrade quietly.
• {publisher_id} is exposed only as a unique identifier, without the placement's name. Enough to build a blacklist, not enough to approach publishers directly.

For custom analytics, your server calls the https://api.adsgram.ai/confirm_conversion endpoint the moment a conversion event is recorded. Ready-made integrations exist for AppsFlyer, Adjust, AppMetrica, Keitaro, and Kochava.

Metrics and unit economics in mobile arbitrage

Arbitrage is not about counting clicks, it is about how much money each dollar returns. This is the minimum metric set.

One rule governs everything else: a funnel is profitable when EPC stays reliably above CPC. Every other optimisation happens inside that inequality.

The second rule is not to conclude from small samples. A kill decision needs volume — as a working reference, at least 100–200 clicks per creative or 10–15 conversions per funnel before results stop being noise. Deciding after twenty clicks is not discipline, it is guessing with extra steps.

Common mobile arbitrage mistakes

Most wasted budget comes from setup errors rather than bad offers.

• Launching without a tracker. You get spend with no placement-level breakdown and no way to clean traffic.
• No frequency cap. Without Max RF, the same users see your creative repeatedly, CTR decays, and budget drains into an audience that has already decided.
• Over-broad targeting on tests. Ten countries in one campaign is ten incomparable auctions collapsed into a single report row.
• One creative per campaign. You pay for a test and receive one hypothesis instead of ten.
• Editing creatives mid-flight. The campaign re-enters moderation and your statistics fracture at the exact moment you needed them continuous.
• Ignoring placement blacklists. Platform IDs exist precisely so you can switch off zero-conversion sources.
• No stop-loss. A predefined test budget enforces discipline better than any dashboard.

Verticals that work in mobile arbitrage

Your vertical determines both traffic price and moderation requirements, so pick it before you pick a source. AdsGram works across a broad range of verticals — gaming, Web3, finance, e-commerce, lead generation, mobile utilities, software, and others; the list is not exhaustive.

For a first funnel, the most workable starting points are:

• Mobile games and Mini Apps — short conversion cycle, cheap installs, easy creative testing.
• Utilities and software — clear value proposition, stable CR, lighter moderation.
• Finance and fintech — pricier traffic but higher payouts, and creatives need care to clear review.
• E-commerce — performs on warmed audiences and pairs naturally with a channel-based prelander.

If you plan to use a Telegram channel as your prelander, our guide to advertising in Telegram channels covers formats and placement selection in detail. When you are ready to move from theory to a live test, the Telegram advertising platform from AdsGram gets a first campaign running in about five minutes.

FAQ

1. What is mobile traffic arbitrage in simple terms?
   Mobile traffic arbitrage is buying ad impressions on mobile placements and reselling the resulting target actions to an advertiser. You pay for impressions or clicks and receive a fixed payout for an install, registration, or purchase. Profit is the spread between the two. Because earnings exist only at positive ROI, the buyer's real job is testing funnels and cutting the unprofitable ones quickly.
2. How much money do you need to start with mobile traffic arbitrage?
   A starting budget for mobile traffic arbitrage covers traffic plus tooling. AdsGram's documentation recommends 100–250 USD₮ to gather statistically meaningful network data.
3. Which metrics matter most in mobile traffic arbitrage?
   The decisive mobile arbitrage metrics are ROI, EPC, and CPA. ROI shows overall return on a funnel, EPC shows what a single click earns, and CPI or CPA shows what an install or target action costs to produce. A funnel is viable when EPC consistently exceeds CPC. CTR and CPM are diagnostic — they show where a funnel leaks, not whether it profits.
4. Do you need a tracker for mobile traffic arbitrage, or is dashboard data enough?
   A tracker is needed from day one of buying mobile traffic. Dashboard statistics show spend and clicks but cannot map them to affiliate payouts by placement and creative. AdsGram passes the {campaign_id}, {banner_id}, {publisher_id}, {click_id}, and {record_data} macros, supports a conversion confirmation API, and integrates with AppsFlyer, Adjust, AppMetrica, Keitaro, and Kochava.
5. How does tCPC differ from standard CPC when buying mobile traffic?
   tCPC, or Target Cost-Per-Click, is a target click price the system bids toward in the auction. The key difference from classic CPC is that in AdsGram, billing under tCPC still happens per 1,000 impressions rather than per click. The platform proposes a recommended tCPC value based on the targeting settings and daily budget you have already configured.
6. How do you track mobile app conversions in Telegram Ads campaigns?
   Through your tracker only. AdsGram macros are not supported in Telegram Ads campaigns and are not passed into the ad URL, so the conversion confirmation API is unavailable for that campaign type. Mobile traffic sources have to be separated using your own URL parameters and unique links at the level of an individual ad or creative, with reconciliation on your side.
7. How does App Tracking Transparency affect mobile traffic arbitrage?
   App Tracking Transparency limits deterministic install attribution on iOS, so store-destination mobile funnels return aggregated and delayed data through SKAdNetwork. That makes an MMP effectively mandatory for store campaigns, and it raises the relative value of in-ecosystem destinations such as Telegram Mini Apps and bots, where a click identifier survives end to end through your own backend.