- Affiliate Programs
- Affiliate Networks
- Ad Networks
- Articles
- Tools
- Social Media Accounts
- Events
- 💵 Cards from $2
Push traffic has been developing unevenly over the past few years: competition is growing, the quality of placements is constantly changing, and manual campaign optimization is taking more and more time. Against this backdrop, ad platforms are gradually moving to automated payment models, where the webmaster no longer sets the click price, but the target action — and the algorithm itself tries to keep the desired CPA.
With this approach in mind, in 2025 Push.House introduced CPA Goal — a new payment model that optimizes campaigns for a specified lead cost. Before the official release, several teams got early access. Below are the results of their tests and a detailed breakdown of how CPA Goal performs in real media buying.
CPA Goal is an automated bidding model that has only recently entered the market. Unlike traditional formats where you set a price per click (CPC) or per thousand impressions (CPM), with CPA Goal you simply specify how much you are willing to pay per conversion, and the system uses historical data to calculate effective bids for each placement.
Its main objective is to automate campaign optimization, help partners achieve more stable results even in challenging verticals, and spend ad budgets more efficiently in line with specific marketing goals.
How it works:
At the same time, on promising placements the cost per click increases, but is automatically calculated in such a way that the final lead price does not exceed the target CPA. In other words, the model minimizes unnecessary testing and helps find working placements faster, especially in web push and in-page formats.
A team that has been working with Push.House for several years tested CPA Goal on Zeydoo survey offers. The formats used were web push and in-page.
The test logic was simple: to see how effectively the model can learn under a short conversion funnel.
It’s no secret that in 2025 audiences react worse than ever to overloaded pages. The Push.House partner team confirmed this in practice: in their tests, complex and long landings for this offer consistently showed poorer results. According to the team:
«On landings with many sections or a large amount of text, the LP CTR dropped, but the conversion rate did not grow or only increased slightly. A simple landing delivers a higher LP CTR and, as a result, a higher conversion rate. Most likely, the less a user has to think in this case, the higher the chance they will complete the required action».
The team explains this by the specifics of the format and user behavior:
«In other niches, things can work differently — sometimes deep warming-up is really needed. But in push and in-page, the logic is the opposite: the shorter the path, the higher the final conversion rate».
Dating and sweeps are the main niches the team works with, and over time they have developed stable approaches.
For dating: natural photos without explicit content, no «perfect» bodies or models, maximum simplicity in the headline and copy, and a strong CTA:
For sweepstakes: the key is to clearly communicate the potential benefit of the offer to the end user without violating moderation rules.
Neutral but motivating headlines work best:
Here are several examples of working creatives for in-page and push:
All creatives are made in Figma. The logic is simple:
«There is no point in ordering expensive banners from a designer, especially if you’re a beginner or need to run quick tests. A couple of YouTube tutorials — and you’ll be able to handle it yourself».
Important! Don’t forget to localize both your creatives and landings for the selected GEO. Otherwise, you can say goodbye to your advertising budget from the very start.
To clearly show the results the team achieved over 30 days, below are screenshots from the Push.House dashboard with GEOs, bids, spend, and final results.
What’s important here:
This is how the campaigns performed in the tracker:
At this stage, we can see:
This clearly shows that the algorithm does not try to «push» the campaign at any cost, but rather limits traffic to the placements that fit the target CPA.
In terms of overall statistics, during the test period the team achieved the following results:
Despite the fact that some campaigns performed worse, the overall result was positive and showed strong scaling potential.
If after reading this case study you decide to test CPA Goal yourself, here are some practical tips from the team that will help you save budget and reach profitability faster:
1. Set a goal below your actual payout
For example, if the payout is 0.2 — set 0.15; if it’s 0.1 — set 0.07.
After initial tests, you can gradually decrease the target. This will help you attract more cheap but still high-quality traffic. However, don’t set it too low.
2. Use a per-placement spending limit
In Push.House, you can set a test budget for each placement. We usually set it at x2 of the payout. For example, if the payout is 0.15 — the limit is 0.3.
3. Set traffic loss percentage in your tracker
Experienced media buyers know exactly what this is about, but beginners often underestimate its impact on final profit. We use:
This helps properly match statistics and avoid discrepancies in reporting.
Another team working with Push.House also tested CPA Goal. Their stats confirmed the findings of the first team: the model learns faster and maintains a stable CPA when the target is set correctly.
The team summarized all key insights that will help you understand how the model behaves in real conditions and decide whether to test it yourself:
1. The algorithm is stable, learns fast, and reaches working metrics quickly.
On average, CPA Goal needs less time to identify effective placements, redistribute traffic, and maintain the target conversion cost. The better the funnel quality and the more accurate the target, the faster you will move into profit.
2. Simple landings and clear creatives perform best.
Test results show a clear pattern: in-page and push audiences respond better to short, uncluttered landings and creatives without exaggerated promises or complex messaging.
3. Using limits increases stability.
Campaigns with pre-set daily limits and placement restrictions achieved more accurate and predictable results.
4. CPA Goal reduces testing costs.
Compared to CPC campaigns, initial tests with CPA Goal are significantly cheaper. The model filters out weak placements faster and saves budget already during the first day.
Even during early testing, CPA Goal demonstrates stability, overall budget savings, and a noticeable performance uplift. Partners confirm that the model works, distributes traffic correctly, and allows successful campaigns to be scaled without unnecessary manual work.
And if you want to replicate this case and test the CPA Goal model — you can already do it right now with Push.House.
