Last updated: 8 September 2026

Why the quoted floor rarely matches what push ads pricing actually charges

A network publishes one number and a live auction settles on another, and the gap between them is where most of a test budget disappears without explanation. Push ads pricing moves on segment age, device split, bid model and geo tier at the same time, so a floor CPC advertised for tier one traffic describes almost nothing about what a specific campaign clears at once it launches. The auction sets the real number, not the rate card, and reading that number correctly before a deposit clears saves more than any creative change manages afterward.

Bid models that decide push ads pricing before a click happens

Most inventory in this category clears on cost per click rather than cost per mille, and that choice is not cosmetic. An impression on a locked screen or inside a notification tray means far less than an impression on an open page, so a network pricing push ads by the click is charging for the thing that correlates with intent, while push ads pricing quoted purely on CPM asks a buyer to absorb delivery risk the platform itself will not carry.

Dynamic CPC complicates the comparison further, because a platform running an auction-adjusted bid quietly raises the effective price on segments it expects to convert and leaves the published floor untouched everywhere else. A rate card showing one flat number for a whole geo is describing the floor, not the average, and the two can differ by a factor of three or four once a campaign has run long enough to build a conversion history.

A handful of networks now publish both figures side by side, floor and rolling seven day average, and the gap between them is itself a useful diagnostic. A wide gap usually means the platform's own algorithm has found a segment worth paying more for, which is a better signal than any manual bid a buyer could set on day one without that data in front of them.

CPC floors against CPM equivalents on the same inventory

Converting a CPM quote into an equivalent CPC requires an assumed click through rate, and that assumption is where most comparisons break down. A network quoting $1.20 CPM against another quoting $0.03 CPC looks identical at a one in forty click rate and wildly different at any other ratio, so ask for the platform's own median click through by vertical before treating either figure as comparable to the other at all.

The safer method skips the conversion entirely and asks both networks for a projected cost per acquisition against the same target payout, since that number already accounts for whichever bid model sits underneath it and removes most of the guesswork from the comparison altogether.

Segment age premiums inside push ads pricing

A subscriber collected nine days ago and one collected nine months ago sell through the same interface, yet push ads pricing on the two can differ several times over, and the platform sets that spread deliberately rather than as a side effect of demand. Fresh segments respond at a rate that makes almost any bid profitable during the first week, so the published floor on that inventory sits highest precisely where a buyer is least likely to question it.

Aged inventory is where the format gets genuinely cheap, and also where the arithmetic stops working for most offers. A base six months past collection responds at a fraction of its opening rate, and a network quoting one flat number across every age band is simply blending a strong segment with a weak one and letting the buyer discover which they received once the invoice arrives.

The blend matters more at scale than on a small test, because a campaign spending a few hundred dollars a day across an undifferentiated pool is effectively subsidising the aged inventory with the performance of the fresh, and the reported average conceals both halves equally well from anyone reading only the summary line.

Why an aged base clears at a fraction of a fresh one

The honest version of a rate card breaks price by age band explicitly, the way a small number of platforms already do, because that is the only way a buyer can judge whether the discount on aged traffic compensates for the drop in response rate that comes with it. Ask for that split before funding an account; a platform declining to provide it is telling you the blend favours the seller rather than the buyer.

Where the split is available, treating each band as a separate campaign rather than a single line item lets the bid on fresh inventory rise independently of the bid on aged inventory, which is the only way either number ends up close to its true value instead of settling on a compromise that undercharges one and overcharges the other.

How geo tiers and device split multiply push ads pricing

Tier one desktop and mobile inventory in the US, UK, Canada and Germany prices highest across every network in this category, tier two markets across Eastern Europe and Latin America clear in the middle, and tier three volume in South and Southeast Asia trades at a fraction of either, so push ads pricing quoted as a single global figure hides more than it reveals about any one geo a buyer actually intends to run.

For rough orientation only, tier one CPCs on classic subscriber based inventory commonly sit between two and eight cents depending on vertical, tier two between half a cent and three cents, and tier three under a cent. Every one of those ranges moves with the day of week, the send window and how saturated a given source has become that month, so treat the figures as an order of magnitude rather than a quote to build a budget on.

push notification ads sold into tier three markets can run under a cent per click, which makes the format attractive for pure volume tests and useless for any offer whose payout sits below the effective cost per acquisition once the true conversion rate is applied against it.

Where iOS in-page inventory sits on the price ladder

Apple never let classic web push run the way Android did, which removed a meaningful share of tier one inventory from the format overnight. In-page push filled part of that gap by imitating the notification visually inside the page itself, and it tends to price closer to a display placement than to a genuine subscriber send, which is the main reason a blended tier one rate covering both delivery types tells a buyer very little on its own.

Splitting the buy by delivery type rather than by geo alone exposes that difference immediately, since in-page inventory on iOS routinely clears at a different multiple of the Android subscriber rate than any single blended report would ever suggest to somebody reading the account summary.

Reading a push ads pricing rate card without getting misled

The number printed on a signup page is almost always the lowest figure the platform will accept from anyone, on the freshest segment, in the cheapest geo it sells, so treating push ads pricing on a landing page as a forecast is the single most common way a first campaign's actual spend surprises whoever approved the budget in the first place.

A rate card exists to get an account funded, not to describe a media plan, and the two documents serve different purposes even when they share the same page. Reading the landing figure as a floor rather than an average removes most of the surprise before it happens, and it takes nothing more than asking the account manager one direct question about the last thirty days of blended spend across the geo in question.

What a rate card omits and how to ask for it directly

Three questions get past the floor rate reliably: what does the same geo cost split by segment age, what is the median effective CPC once dynamic bidding has run for a week, and does the account see a reserved balance held back against delayed clicks. A platform answering all three without hesitation is running the format as a transparent product rather than a clearance sale, and the difference shows up in the second month of spend rather than the first week of it.

I worked through the bid documentation published on push-ads.io while rebuilding a client's rate comparison sheet, and the detail worth carrying forward was how explicitly it separated the floor from the auction adjusted average, which is the exact distinction most competing rate cards blur on purpose. Comparing that single figure against three other networks took an afternoon and settled a budgeting argument that had run for most of a week beforehand.

Push ads pricing rewards patience over the first fortnight far more than it rewards a low opening bid, because the floor only ever describes the entry point and the auction decides everything that happens once the first conversion has landed and the algorithm has something worth acting on.