Get Resource

The line no checkout page wants to quote is paragraph 4.4 of the CMA's price transparency guidance (CMA209, 18 November 2025): "administration fees, however described, such as booking or processing fees" belong in the price you are first shown. That is the rulebook on drip pricing in one clause. If you cannot buy without the fee, the seller does not get to reveal it on the payment screen.

What Is Drip Pricing and Why Does It Matter?

Drip pricing is showing a low headline price, then adding mandatory fees later in the purchase; since 6 April 2025 the DMCC Act treats leaving those fees out of the first price as an unfair practice. It is not one of the 32 banned practices in Schedule 20. The CMA's guidance (CMA207, 18 November 2025) files it as an omission of material information under sections 225(4)(b) and 230, so cite those and CMA209 in a complaint. Schedule 20 is the wrong list.

The enforcement is real. On 15 April 2026 the CMA fined the AA and BSM driving schools £4.2 million over a mandatory £3 booking fee charged from April to December 2025: £7 million, less 40% for settling early. The ceiling is 10% of global turnover or £300,000 if higher, so settling looks rational. By our arithmetic that is 5.5 times the £760,000-plus ordered refunded, or roughly £9.50 for each of the 80,000-plus learners (both are lower bounds).

Two neighbouring rules: the UK subscription rules from January 2027 add a 14-day cooling-off on renewals, and EU right to repair, live since 31 July 2026, is another rule sellers would rather you did not quote.

StubHub UK's case, announced on 23 June 2026, shows the pattern at ticket scale. Delivery and service fees surfaced only at the final checkout stage, from 6 April to 7 December 2025. The CMA fined it £889,200 and ordered £590,000-plus back to 51,350 customers, £10.33 each on average: close to the AA learners' figure, so a fee worth a tenner can draw a fine. Four numbers decide whether chasing yours is worth it.

What matters at your kitchen table is the route the money took. StubHub's buyers filed nothing: the CMA says affected customers need take no action, and refunds went back to the card used. Treat that as the exception. Without an order, nobody sends your fee back until you ask.

Is It Illegal to Add a Booking Fee at Checkout?

Yes, if the fee is mandatory: under the DMCC Act a booking fee you cannot avoid must be in the headline price, so adding it at checkout is the unfair practice the CMA is enforcing. Find your situation below and read across.

Two rows settle most disputes: could you have bought without the fee, and is it a card surcharge. On a small fare the sting is the percentage, not the pounds (or not only the pounds).

Worked example: take a £20 coach fare with a £1.50 booking fee, booked 8 times a year, with 19 months since the rules began on 6 April 2025. These are assumptions, not findings. Step one: £20 + £1.50 = £21.50 all-in. Step two: £1.50 ÷ £20 = 7.5% on top of the headline, so divide your own fee by your own first price. Step three: £1.50 × 8 = £12.00 a year. Step four: £12.00 × 19 ÷ 12 = £19.00 paid since 6 April 2025. Put your numbers in below.

Can You Get a Refund for a Hidden Booking Fee?

Sometimes: where the CMA has made an order, as with StubHub, refunds come automatically to your card; otherwise you must ask the seller, cite CMA209 paragraph 4.4 and keep proof of the first price you saw. Our call: ask in writing the week you notice, rather than wait for a regulator that can only take on a few firms. The limit: if you could have bought without the fee, it was optional under paragraph 4.2 and the claim collapses, so check that first against the CMA's price transparency guidance.

The rules stop short in places. The total-price rule covers how a price is marketed, not how or when you pay, so instalments are a separate question. Blaming a third party gets a seller no cover: paragraph 4.5 puts unavoidable supplier costs in the headline. Outside a CMA case the chase is yours alone, as with a closed Meta ad account balance.

• The CMA's investigations into Trainline, Virgin Atlantic and RED Driving School (19 August 2026) cover rail fees of £0.59 to £2.79, resort fees and £7-plus booking fees; no conclusions yet, so they show scope, not guilt.
• Each CMA case repays only its own period, so a fee outside it needs its own request.
• A per-person or per-kilo price can be lawful if the unit price is shown prominently and the total appears once you enter a quantity.

This week, pull up your last three booking confirmations, find any fee that first appeared on the payment screen, and email each seller asking for a refund, quoting CMA209 paragraph 4.4 and attaching your screenshot of the first price. The decision is whether the fee was optional: if you could have bought without it, drop the claim; if not, send the email. General information, not legal advice.

The line Meta would rather not quote is the one its own support assistant gave: close the ad account, and the balance refunds automatically to the original payment method. In the first-hand case behind this piece, the account was closed on exactly that advice. Months later the closed Meta ad account balance is still sitting there, and nothing has come back.

Where does a closed Meta ad account balance go?

Nobody outside Meta can say from public documents we could read, but in the case behind this piece it went nowhere: the money stayed in the closed account for months, with no refund and no reinstate option.

Here is the sequence, as the account's owner gave it to this site: first-hand, not independently documented, one advertiser's record rather than a statistic. A $75 top-up meant for the active account landed in a second one. The ticket asked Meta to move it or refund it. The assistant answered with the closure advice. Every attempt to reach a person met the assistant again, so the account was closed. A fresh ticket later drew the opposite answer. That is not a refund route. It is a one-way door with a refund sign painted on it.

Other providers write their exit terms where customers can cite them: UK rules give subscribers 14 days to walk away. Meta's version, on this record, was a sentence from a chatbot. When a company withdraws what it promised, as in a pledge with no service behind it, the customer holds the paper. These four counts are one advertiser's case file, not Meta's rates.

The people count is the one that should worry you most. An irreversible step with no reviewer in front of it is how a fixable misdirected payment becomes a permanent one, and the assistant that advised closing is the same voice that later cited the payment mode.

Can you get a refund from a closed Facebook ad account?

Not on anything we could confirm: Meta's help pages block automated reading, so the rule is unverified here, and the one account on record shows a refund promised before closure and declined afterwards.

Start with the one sum you can do yourself, whatever the pages say.

Worked example: assume a $75 top-up lands in the wrong account, your intended account spends $15 a day, you have waited 3 months (90 days), and your card adds a 3% foreign-transaction fee on the re-top-up. These are illustrative assumptions, not Meta rates. Step 1, your ten-second check: $75 divided by $15 a day is 5 days of ads you are not running. Step 2: re-funding the right account costs $75 times 1.03, which is $77.25. Step 3: if the stranded balance never returns, you have paid $75 plus $77.25, which is $152.25 to get $75 of ads running. Step 4: 90 days of waiting against 5 days of runway is 18 times. Swap in your own figures below.

Which move fits depends on where you are in the sequence. The fourth row is our own rule of thumb, not a Meta threshold.

Every row turns on one hinge: who made the decision, and whether you can read it later. A chat reply the next chat reply can overwrite is not a decision.

Are Meta ads manual payments refundable?

We cannot confirm it: Meta's refund and payment-mode pages could not be read for this article, and the only evidence is one ticket that cited a non-refundable payment mode months after the closure advice.

Each friction favours the provider. The advice came before closure and the contradiction after it, so the customer holds the screenshot and the provider holds the money. Nothing was appealable either: no person reviewed the matter. Where regulators have stopped makers using stock excuses, as with the refusals EU makers can no longer use, customers get something to cite. Here we could cite nothing, and that gap is the argument.

Our view, and it is an opinion: a company that offers closure as the fix owes the customer one human check before the door locks. Watch for these:

• An assistant's reply may be the only record you hold, so save it with its date.
• In the case, a later answer turned on payment mode, so look up which mode your account uses before you top up.
• Your card issuer sets its own dispute window and the clock may already be running, so ask it today.

This advice is wrong in one case: if Meta's own pages state a refund that covers your payment mode. Read them first, at the links in the method note below.

This week, before you close any Meta ad account to recover money, ask for a person's written decision on your payment mode and keep it; if the account is still open, our reasoning (not a Meta instruction) is to spend the stranded balance there on ads you would have run anyway.

“Returns do not take into account the load and taxes, if any.” That line sits under Baroda BNP Paribas Mutual Fund's SIP table, as on 31 December 2024, and it is the rule your fund house would rather you never quoted. For a Rs 2,000 monthly investor, SIP returns in rupees are the only figure that pays a bill, and the page does not print them.

Why do SIP returns in rupees tell you more than XIRR?

Because XIRR is a rate, not a result: it says how fast each rupee grew while it was invested, and says nothing about how many rupees there were or how long they stayed. Value Research is right that CAGR is the wrong yardstick for a SIP. Call the complaint a quibble about presentation. Actually, no: when only the flattering number is printed, presentation is the whole problem.

In August 2025 Value Research reported a seven-year SIP investor on Reddit who put in Rs 11 lakh, saw barely Rs 16 lakh, and called it a 7-8 per cent CAGR. Value Research recomputed an XIRR of about 10.97 per cent, close to the Nifty 50's 11.13. One anonymous post, not a statistic: same money, two verdicts, and no fund page says which to trust.

Who sets the presentation? Not SEBI, going by its circular of 22 October 2018, which I read as reproduced by TaxGuru, not from SEBI's own PDF. It asks for scheme returns against the Total Return Index as CAGR over fixed periods. SIP returns and XIRR do not appear, so the SIP percentage is house style, and house style can omit rupees. We saw the same loud pledge and quiet fine print in Ford's 2021 India service pledge against its shrinking network; on household money more widely, see India's tax trap.

The footnotes are blunter:

• Axis Mutual Fund (as on 31 October 2025) applies XIRR to “say Rs. 10,000” a month, five times your Rs 2,000.
• Baroda BNP Paribas (as on 31 December 2024) assumes a regular plan, so plan cost is inside the percentage and load and tax are not.
• It adds that past performance “should not be used as a basis of comparison with other investments”.

These four numbers come from Rs 2,000 on the first working day of each month for ten years at a steady 12% XIRR, our arithmetic, set out below.

Early instalments have years to compound, so stopping halfway does not give half the result: five years reach roughly a third of what ten do, on our arithmetic.

How much will a Rs 2,000 monthly SIP become in 10 years?

At a steady 12% XIRR it becomes about Rs 4.5 lakh from Rs 2.4 lakh deposited, while 8% gives roughly Rs 3.6 lakh and 15% about Rs 5.3 lakh, before load and tax.

Worked example: assume Rs 2,000 on the first working day of each month, as in the Axis and Baroda footnotes, for 120 months at a constant 12% XIRR. Monthly rate: 1.12^(1/12) − 1 = 0.9489%. Value: 2,000 × ((1.009489^120 − 1) ÷ 0.009489) × 1.009489 = Rs 4,48,072. Deposits: 2,000 × 120 = Rs 2,40,000, so the gain is Rs 2,08,072. As a day-one lump sum, those deposits need 6.4% a year to reach that pile. Real XIRR uses actual dates, so expect small differences.

Your turn, in one minute. Divide your app's current value by the money put in (1.87 above), raise it to the power of one over your years with the x^y key, and subtract 1. Expect a rate well below your XIRR: the average instalment works for only half the term.

Find your years down the left and read across at your quoted XIRR. Cells are value in lakh (1 L = Rs 1,00,000), before load and tax.

The shaded cell is the worked example; one step right adds about Rs 0.8 lakh, one step left removes about Rs 0.85 lakh. Going from 15 to 20 years adds more than the first fifteen years' whole gain.

What takes the rupees away after the fund page stops counting?

Load, tax and plan cost all come off after the table ends: the fund page excludes the first two outright, and the third is the regular-plan cost its numbers assume.

The CBDT's FAQs on the 2024 changes, published by PIB, set long-term gains on equity-oriented funds at 12.5% above Rs 1.25 lakh a year and short-term gains at 20%, from 23 July 2024; Business Standard reported Budget 2026 left long-term rates unchanged for FY27. Selling the worked example in one year owes the tax in the grid, on our estimate, slightly more as the final year's units are short-term. Confirm current rates with a chartered accountant first.

Then the quieter leak: your statement shows whether your folio is Regular or Direct, so check before comparing percentages. Our call: a recurring debit deserves a yearly statement of rupees in, rupees now and rate on outlay. The UK is heading that way (our reading of its subscription rules).

• Ask your platform for amount invested and current value, not only XIRR.
• Compare funds over identical start and end dates.
• Spread sales across financial years only if you do not need all the money at once; each year has its own Rs 1.25 lakh exemption.

Our call: keep XIRR, but never accept it alone. This week, open your latest statement, note the money put in and its value, and run the one-minute sum above to decide whether the rupee gain justifies your plan and monthly amount. If your folio is already Direct, go straight to the tax check.