One real dating campaign on EZmob traffic turned a cautious $50-per-day test into 275 conversions at a 154.7% ROI over three months — and the network’s own case study published the spend discipline, targeting changes and tracking setup behind it. This page rebuilds that campaign into a playbook for media buyers running dating offers on push traffic in 2026: what was launched, what went wrong in the first weeks, the five optimizations that turned it around, and the exact order to copy them in.
The campaign at a glance
| Field | Detail |
|---|---|
| Vertical | Dating affiliate offer (registrations and deposits) |
| GEO | France |
| Format | Floating push |
| Entry budget | $50 per day test phase |
| Duration | About 3 months, continuing into February 2026 |
| Conversions | 275 |
| ROI | 154.7% |
| Net profit | Over $500, with month-over-month profit growth and a falling CPA |
A note on the numbers before anything else: every figure here comes from the floating push dating case study published on EZmob’s blog, submitted by the buyer who ran the campaign. It is one campaign, in one GEO, on one offer — read it as proof that the playbook converts on this inventory, not as a guaranteed outcome for your funnel.
Why floating push fits dating offers
Dating offers live or die on repeat exposure. A user rarely registers on the first impression; they come back to the offer after seeing it two, three, five times. Push-based formats carry that advantage structurally: the creative lands on the user’s device and stays visible when they become active, so the same subscriber sees your angle repeatedly without you paying for a fresh impression every time.
Floating push adds a second property that matters specifically for dating: the ad floats over the page the user is reading, so it gets noticed without requiring the user to have subscribed to a notifications channel. That widens the reachable audience on mainstream publishers where a dating creative needs to earn attention fast.
The third property is cost control. Push-family bids are priced per click, and the entry points sit well below display CPMs, which is what let this campaign run a genuine $50/day test instead of a token one. At that daily budget you can collect real conversion data across dozens of publisher feeds in the first week — which is exactly what the optimization phase needed.
How the campaign was set up
The offer and the payout model
The offer was a dating affiliate program where the money is in registrations and deposits. That payout structure dictates your tracking before it dictates your targeting: a registration and a funded deposit are worth very different amounts, and if your tracker treats them identically you will optimize toward cheap, worthless signups.
Tracking: the postback that made optimization possible
The buyer connected the EZmob postback to their tracker before scaling anything. The setup split the two events by value:
- All new registrations were tracked under the conversions column.
- All deposit amounts were passed back as goal values.
That split is the reason the rest of this case study works. When the buyer later reallocated budget toward the top publisher feeds, they could see not just which feeds produced registrations but which produced depositing users. Without the postback, the campaign would have been optimized on conversion count alone — and on dating offers, conversion count is where bad traffic hides.
If you are setting this up yourself, the EZmob-to-Keitaro postback guide walks the same connection for one of the trackers most media buyers already run.
Entry targeting: deliberately broad
The campaign launched with wide targeting across devices, browsers, languages, publisher feeds and SubIDs, at the low daily budget. That was the right instinct executed too loosely — broad entry targeting is how you discover which slices of supply convert, but on this campaign it diluted the $50 across too many low-intent segments before the data existed to prune them.
The first weeks: what went wrong
The published case study is candid about the opening phase: low daily registrations and an underperforming ROI, driven by the overly broad targeting. Four specific problems compounded each other:
- Language spread. An expansive language selection reached audiences with low engagement intent for a France-targeted dating offer, suppressing conversion rate across the board.
- Device and browser spread. Spend leaked into devices and browser versions that historically delivered less value per user.
- Publisher feed spread. Budget was spread across many feeds instead of concentrating on the few that consistently converted.
- No connectivity segmentation. Wi-Fi and cellular traffic were bought as one audience, even though the two behave differently on a format that pushes notifications when the user is active.
None of these are exotic problems — they are the default state of every new push campaign. The difference between a campaign that stalls at this stage and one that compounds is whether the buyer acts on tracker data segment by segment instead of relaunching from scratch.
The five optimizations that turned it around
1. Narrow the languages to the ones that actually engage
Language targeting was refined to prioritize French, German and Arabic audiences. The effect was immediate and twofold: stronger performance on the remaining traffic and more consistent, controllable spend. For a France geo this is the single highest-leverage cut — the buyers your offer can actually convert cluster in a small set of language segments, and every impression bought outside them is a discount on nothing.
2. Concentrate devices and browsers on the proven slice
Device targeting was narrowed to iOS and Android only, lower-performing devices and non-core browsers were excluded, and the campaign concentrated on recent browser versions. Once those settings were in, performance held strong with minimal ongoing optimization — the sign that the remaining traffic was genuinely the campaign’s audience rather than a lucky week.
3. Reallocate budget to the top six publisher feeds
Publisher feed targeting was streamlined to the top six performers — the feeds delivering consistent conversions and engagement. This is the move most self-serve buyers skip, because it feels like shrinking the campaign. It is the opposite: the feeds that convert at your CPA are a supply constraint, and the fastest way to scale a dating campaign is to feed them the budget that weak feeds were absorbing.
4. Split the campaign by connection type
The buyer’s account manager recommended splitting the campaign into Wi-Fi and cellular versions. The two connection types carry different user contexts — Wi-Fi users browse from stable environments, cellular users are often on the move — and splitting them allowed tailored bidding and delivery strategies for each. The published case study credits this segmentation with improved efficiency, increased conversions and maximized overall performance.
This is also the cheapest optimization on the list: it costs nothing but a campaign duplicate, and it converts a blended average into two optimizable audiences.
5. Make micro bid adjustments the ongoing habit
Strategic micro bid adjustments let the campaign compete in high-value segments — particular feeds, device and language combinations — without repricing the whole campaign. On push traffic, where a few cents of bid decides which placements you win, the buyer who adjusts bids at segment level consistently outbids for the traffic that converts and stops overpaying for the traffic that does not.
The results
After three months the campaign closed with 275 conversions and a 154.7% ROI, netting over $500 in profit for the buyer. Month-over-month analysis showed profit increasing while CPA stayed consistently lower — the pattern you want from a dating campaign that has found its supply: efficiency improving as volume grows, not degrading.
Into February 2026 the campaign opened the month strong, and the buyer launched additional campaigns both to grow overall profit and to test the same playbook in new international markets. France remains one of the stronger dating markets on push inventory, and the case study’s own conclusion is that the targeting discipline — not a creative miracle — is what unlocked it.
The playbook, in order
Strip the campaign down and the sequence is reproducible:
- Set up the postback first. Registrations as conversions, deposit amounts as goal values. No optimization before tracking is split by value.
- Launch broad at a small daily budget. Wide targeting on purpose — you are buying a map of what converts, at $50/day scale, not at scale.
- Cut languages before anything else. Keep the two or three language segments your offer actually monetizes.
- Concentrate devices, browsers and feeds. iOS and Android, recent browsers, top-performing publisher feeds only.
- Duplicate the campaign by connection type and bid each version on its own data.
- Run micro bid adjustments weekly at segment level — feeds, devices, languages — rather than moving the whole campaign’s bid.
Running it on EZmob in 2026
The formats and targeting controls this campaign used — floating push, language and device targeting, publisher feed selection, connection-type segmentation — are all self-serve on EZmob’s self-serve advertising platform. You can review the formats, bid types and GEO coverage before spending anything, and register an advertiser account without a sales call. New campaigns go through creative moderation before they go live, so submit your dating creatives early and keep the claims on the landing page consistent with the ad.
Two warnings worth carrying into your own launch. First, do not skip the postback to save an hour — every optimization in this case study depended on knowing which traffic deposited, not just which traffic registered. Second, resist raising the daily budget before the language and feed cuts are done; more budget on unpruned targeting buys more of the problem, not more of the profit.
How this compares with the other dating content on this blog
If you are still choosing between networks for dating traffic, the best ad networks for dating offers guide ranks the field by format coverage and creative eligibility. If you already have a network picked and need the landing path right, the landing pages guide for pop and push traffic covers pre-lander structure for the same audience. This post sits between the two: the network is chosen, the offer is chosen — the question is how a real campaign on this traffic was actually optimized to a 154.7% ROI.
FAQ
Is floating push good traffic for dating offers?
It was in this case: a France-targeted dating offer converted 275 registrations and deposits over three months at a 154.7% ROI, starting from a $50/day test budget. Floating push reaches users without requiring a prior notification subscription, which widens the audience on mainstream publishers.
How long before a dating push campaign shows its real numbers?
This campaign ran roughly three months from first launch to the published result, with the first weeks underperforming. The turnaround came after the targeting cuts landed — performance jumped within a day of the refinement round. Give the test phase enough budget to produce segment-level data before judging the offer.
What should I track on a dating campaign?
Split registrations and deposits at the tracker level: registrations as conversions, deposit amounts as goal values. Optimizing on registration count alone rewards cheap, non-depositing traffic, which is the most common way dating campaigns look profitable for two weeks and then collapse.
How much budget do I need to test dating offers on push?
The published campaign started at $50 per day. That is enough to gather conversion data across multiple publisher feeds within the first week, which is what the language, device and feed optimizations were built on. Scale the budget after the pruning cuts, not before.
Related guides
- Floating push case study: dating offers in France
- Best ad networks for dating offers: 2026 verdict
- Connect EZmob to Keitaro for postback tracking
- Landing pages for pop and push traffic
Start advertising on EZmob — launch a floating push dating campaign self-serve, with new campaigns going live after creative moderation.