User acquisition platforms for mobile games: what they do and when you need one
Search for "user acquisition platform" and you get five different kinds of product, sold under one name. A network's own buying tool, a DSP, a measurement partner, a creative tool and a cross-network platform can all call themselves a UA platform. They solve different problems.
This guide sorts them, explains what AI agents in a UA stack do today and where they stop, and covers the cases where no platform is the right answer. For the practice of user acquisition itself (strategy by lifecycle stage, costs, the checklist), read our mobile game user acquisition playbook. This piece is about the tools.
What a UA platform is
A user acquisition platform is software that helps you buy, measure or manage paid installs for an app. In mobile games the term usually means one of five things.
1. A network's own buying platform
Every large ad network has a self-serve tool where you set budgets, bids and creatives for its own inventory: Google's App campaigns, Meta's app campaigns, AppLovin, Unity, Mintegral, TikTok and others. These are where most spend is placed. Each one optimizes inside its own walls and reports in its own terms, so a studio on six networks runs six dashboards, six attribution views and six sets of quirks.
2. A demand-side platform (DSP)
A DSP buys programmatically across ad exchanges rather than selling its own inventory. Moloco is the best-known example in mobile games: it is a DSP and owns no inventory. A DSP widens your reach beyond the big networks and brings its own bidding models. It is still one more channel to manage, with its own reporting.
3. A mobile measurement partner (MMP)
AppsFlyer, Adjust, Kochava and Singular attribute installs and in-app events to the campaigns that drove them, which gives you one view of performance across networks. Measurement is their core, and several now ship campaign tools too. Adjust and Kochava both offer campaign execution features. If a vendor tells you MMPs "only measure", check that claim against the MMP you actually use.
4. A creative tool
Creative platforms produce, tag, test or analyze ad creatives. Some generate variants with AI, some break creatives into elements and track which ones took spend. They matter but they do not place budgets. Our guide to creative testing and scaling covers the method.
5. A cross-network platform
The fifth kind sits above the networks and works across all of them. It comes in two shapes, and the difference is who decides.
- A workbench. One console for every network. Your team builds campaigns in bulk, reads one report and writes the rules (cut this budget if ROAS falls below target, pause that ad if CPI spikes), and the tool applies them. Every decision is one your team wrote down in advance.
- An execution layer. It pulls spend, installs and revenue from every network into one data model, decides where the next dollar should go and acts on it across networks: moving budget, adjusting bids, pausing what is not paying back. You set the targets and the limits it works inside.
Both are sold as "a UA platform", so it is worth asking which one you are looking at. Either way, acting across networks safely needs clean data and controls on spend.
What a cross-network platform should do
If you are evaluating the fifth kind, these are the capabilities that separate a working product from a dashboard with a recommendation panel.
- One data model. Spend, installs, revenue and retention from every network and your MMP, reconciled into one source of truth. Without it, every decision downstream is made on numbers that disagree with each other.
- Decisions tied to payback. Budget moves judged on ROAS or payback against your targets, by cohort, not on CPI alone. A cheap install that never pays back is not a win.
- Execution, not only advice. The platform can place the change on the network, not just suggest it in a report someone reads on Monday.
- Guardrails. Spend limits, approval workflows and a full audit trail. Any tool that can move money needs all three.
- Data ownership. Know where your data sits and who can take it with them. A platform that keeps your history in its own warehouse makes leaving expensive.
- Neutrality. A network's own tools optimize toward that network's inventory. A cross-network layer should have no stake in which network wins.
On that last point, this is how we put it for our own product: "AXON closes the loop only on its own supply, optimizing its own revenue. We close it network-neutral, on your data, on any supply."
AI in user acquisition: what agents do today and where they stop
Every vendor in this space now says AI. It helps to separate three things the word covers.
Bidding models inside the networks. Every large network already uses machine learning to decide which user sees your ad and what to bid. This has been true for years, and you cannot switch it off. It optimizes for that network.
Assistants that read your data. Chat tools that answer questions about your campaigns. Useful for analysis, but they do not act, and their answers are only as good as the data underneath. Point an LLM at raw tables and it hallucinates. Point it at a validated semantic layer and it reasons.
Agents that act. Software that detects a change, decides what to do and does it across networks: shifting budget between campaigns and networks, adjusting bids, flagging anomalies, catching tracking breaks, producing the daily report. These are the routine operations that eat a UA manager's week. SuperAI, our agent layer, is an autonomous AI workforce executing 80–90% of publishing ops inside guardrails, 24/7.
Where agents stop today:
- Strategy. Which markets to enter, how hard to push a soft launch, when a game is ready to scale. Agents inform these calls; people make them.
- Creative concepts. Agents can compose and test variants of a concept. Finding the next concept that changes a game's CPI is still creative work.
- Commercial relationships. Network deals, rate terms and new-channel onboarding are negotiated by people.
- Situations without history. A new genre, a new network, a market shock. Agents are good at the patterns they have seen and need human review on the ones they have not.
The practical question for any AI claim in a UA tool: what can it change on its own, inside which limits, and where is the log of what it did?
When a platform is the wrong answer
A UA platform is worth its cost when you spend across several networks and have people to run it. It is the wrong buy in these cases.
- You spend on one or two networks. The networks' own tools and your MMP cover it. A cross-network layer earns its keep when there are networks to coordinate.
- You have nobody to operate it. Software that makes decisions still needs someone to set targets, review what it did and handle the exceptions. If UA is a part-time job at your studio, SuperScale Media, our managed UA service powered by the SuperScale Stack, fits better than a license.
- Your tracking is broken. Fix attribution and event mapping first. Any platform, AI or not, will optimize confidently toward bad numbers.
- The game is not ready. If retention and monetization do not support paid acquisition yet, no tool will make the payback work. Our guides to player LTV and CPI cover the numbers to check first.
Questions to ask a UA platform vendor
- Which networks can it act on today, and which only report?
- What can it change without a human approving it, and how do you set those limits?
- Is there an audit trail of every change it made, with the reason?
- Where does our data live, and can we take it with us?
- What does onboarding involve: an SDK in the game, or integrations on the data side?
- How is it priced: per seat, per game, or as a share of spend?
Where SuperScale fits
SuperScale is one platform: licensed or managed. License the SuperScale Stack for your own team, or hand your UA to SuperScale Media and our team runs it on the same stack.
- The SuperScale Stack is the licensed product: the SuperVYZR app, the SuperAI agents and the SuperPlatform semantic layer, which brings 30+ no-SDK integrations into one gaming-native source of truth deployed in your own BigQuery. You own the data. Your team operates it for €7k per game per month, on a one-year agreement. Multi-title publishers license annually, priced per portfolio. See how the SuperScale Stack works.
- SuperScale Media is a standalone managed UA service powered by the SuperScale Stack, with no stack license required. Your spend runs through SuperScale's own network agreements — one IO, 40+ networks. It carries an 8% management fee on managed spend. You pay the 8%, not a license plus the 8%.
Both are on the pricing page. If you are not sure which side of the "when a platform is the wrong answer" list you are on, talk to us and we will tell you.