Remittance apps vs banks vs padala centers: how the costs differ

Apps, banks and padala centers do not differ mostly on the headline fee. They differ on where the cost is taken and how far the channel physically reaches. Comparing them on the upfront fee alone is how the cheapest sticker turns out to deliver the fewest pesos. The fair comparison is one number: pesos received for the exact amount sent. Everything below is why that one number, not the fee, is the thing that actually separates the three, and how to read each channel so the comparison is real.

Why total cost is the only fair basis

Every channel charges in two places: a fee you can see, and an exchange-rate gap you usually cannot. Both halves are posted, and the providers say so in their own words. Wise states it uses “the live mid-market rate, and a small, upfront fee to cover our costs,” and contrasts that with providers that advertise a free transfer while “hiding extra fees in the exchange rate” (wise.com/us/pricing/send-money, checked 2026-08-08). Western Union states on its Philippines page that it “makes money from currency exchange” and that “fees and rates subject to change” (checked 2026-08-08). The fee alone is not a comparable number. A transfer with a visible fee on a tight rate can deliver more pesos than a zero-fee transfer on a wide one, and there is no way to tell which from the fee.

The conversion gap is the half that does the hiding. It is the difference between the mid-market rate and the rate a channel actually pays out at, taken inside the conversion rather than billed as a line. This is why the channel ranking can invert with the amount: a fixed fee is a large share of a small transfer and a trivial share of a large one, while a percentage margin scales with the amount. The cheapest channel for $200 is not necessarily the cheapest for $2,000. Not because the providers changed, but because the two cost components scale differently against the amount.

What each channel is actually built for

The three are not three prices for one product. Each solves a different access problem, and the cost shape follows from what each is built to do. Read this way, “which is best” stops being a single answer and becomes a question about the specific transfer.

Apps / digital services. The cost is a stated fee on a conversion, and both halves are shown before sending. Remitly’s own GCash payout page puts the structure plainly: “the transfer fee depends on how much you send, your payment method, and the delivery speed you choose. You’ll see the fee and exchange rate before you send” (checked 2026-08-08). The trade-off is built into the product rather than hidden: a slower delivery option is priced lower than a faster one on the same transfer. The failure mode to know is separate from price. The cheapest digital quote assumes the recipient can actually receive a bank, card or wallet credit. If they cannot, the low price is for a delivery that does not reach them.

Bank transfers. A fee plus a conversion gap, and the path itself can add cost: an international bank transfer often routes through a correspondent bank between the sending and receiving institutions, and that intermediary can deduct its own charge that neither end quotes upfront. This is the cost hardest to see in advance, because it is not the sending bank’s fee and not the conversion gap. It is a third deduction inside the rails, and it is the usual reason the pesos that land are less than the sending screen implied. On posted speed, Western Union states a bank transfer to the Philippines “may take up to five days” (checked 2026-08-08), against “in minutes” for a mobile wallet on the same page.

Padala / agent centers. The payout form is cash, handed over at a counter against an ID, and that is what separates the channel from the other two: it requires no account, no card and no phone on the receiving side. Western Union posts that its cash pickup to the Philippines “can be available same day” (checked 2026-08-08). A bank transfer pays out as a balance in an account, reachable only through a branch or ATM the recipient can get to; a digital service pays out as a bank, card or wallet credit on the same terms. Where the recipient holds none of those, the agent counter is the one payout form of the three they can present an ID for and collect.

The axis the comparison usually misses

Most comparisons stop at “app vs bank vs padala,” but two providers post that the payout method, not the channel name, is what sets the fee. WorldRemit states it on its Philippines page: “The transfer fee depends on the receive method and does not vary with the amount” (worldremit.com/en/philippines, checked 2026-08-08). Remitly publishes the same relationship as a table. Its posted USD→PHP transfer fee on 2026-08-08 was 0.00 USD for bank deposit, debit card deposit and mobile wallet, and 4.99 USD for cash pickup and home delivery (remitly.com, checked 2026-08-08).

Neither provider states how that spread compares to the spread between providers, but the corridor survey does, and it points the other way. In the World Bank’s US→Philippines pricing for a USD 200 transfer, the gap between the payout methods’ median total cost is about 2.6 percentage points: 1.83% for a mobile wallet across 6 products, 3.46% for a bank deposit across 15, and 4.46% for cash pickup across 36. The spread within a single payout method is larger than that. Cash-pickup products alone run from 1.10% to 9.99%, a spread of nearly 8.9 points among firms all delivering the same payout form (read via the Internet Archive capture of 2026-07-11, data collected Aug 04–21, 2025).

So the payout method is not the bigger lever. Which firm you quote, for the payout you have already fixed, is where the wider variation sits, and that is an argument for quoting rather than for a brand. What this supports is a procedure, not a ranking: first fix the payout the recipient can actually use, then compare only the channels that offer it. A channel’s quote is a price for the payout form that channel delivers in, so where the recipient cannot reach that form, the quote covers a payout that does not complete. This is also why wallet payout has reshaped the question for many families. It collapses the old cheap-but-bank-only versus reachable-but-dear trade for a recipient who has a wallet, and changes nothing for one who does not.

The habit that survives every price change

Because posted prices move and rankings invert with the amount and the payout, no page can name “the cheapest channel” and stay true for long. Any page that tries is stale within a day. What does not go stale is the method, and it is short:

1. Establish which payout the recipient can actually use: bank, card, wallet, or cash.
2. Quote the exact amount, to that payout, on two or three channels at the same time.
3. Compare one number only: pesos received. Not the fee, not the advertised rate, not last month’s quote.

Done at the moment of sending, this beats any remembered ranking, because it prices the actual transfer rather than a generalisation about the channels. This is the same discipline as the cheapest way to send money; this page is the structural half of it (why the channels differ) and that page is the procedure.

How to read this

This is information about how the channels are built, not a verdict on which to use. That decision is set by the amount, the recipient’s access and the day’s rate, none of which a page can know. The cost mechanics described here divide into two kinds of source. What each channel says about its own fee, rate and speed is quoted from that provider’s own posted page on 2026-08-08. The corridor-wide figures, meaning the total-cost percentages and the per-payout medians, come from the World Bank’s Remittance Prices Worldwide survey for this corridor and are dated to its collection period rather than to today. No specific exchange rate is stated anywhere, because those move daily.

Two withdrawals belong on the record, and both were withdrawn because the source contradicts them rather than because it went missing. Earlier versions of this page attributed two structural rankings to the World Bank Remittance Prices Worldwide US→Philippines corridor: that digital services tend to win total cost for a bank deposit, and that the payout-method spread exceeds the spread between providers. The corridor survey supports neither. On bank-deposit payouts priced at USD 200, the 10 internet-access products carry a median total cost of 3.46% against 3.00% for the two agent-access products, so digital does not hold the lower median; and as set out above, the spread among firms inside one payout method is roughly three times the gap between the payout methods’ medians. Both claims are gone, and the figures that replaced them are the survey’s own.

The corridor page itself needs a caveat on how it was read. Direct requests to remittanceprices.worldbank.org returned HTTP 403 on 2026-08-08 at every path tried, including the corridor page, its /en/ variant, the methodology page and the data-download page. That is a block on automated requests, not a dead site. The figures above were therefore read from the Internet Archive’s capture of 2026-07-11, which carries the corridor’s Q3 2025 survey with its collection period stated as Aug 04–21, 2025, covering 14 firms and 58 priced products. Those are 2025 survey figures for a USD 200 transfer, not live quotes, and they are cited as a structural pattern rather than as today’s price. Separately, the World Bank Data Catalog entry loads directly and records the dataset’s scope: 377 country corridors, from 48 sending to 111 receiving countries, compiled quarterly (checked 2026-08-08).

For the speed side of the same choice, see how long a remittance takes; and note that money never travels in a balikbayan box — currency is prohibited and uninsured, which is the whole reason a remittance channel is the question at all.