Qui Pus Info

Up to ₹21,300 over five years. That is where RO purifier maintenance cost lands on this site's model, with top-of-market parts and a technician twice a year, all spent after the box is paid for. The cheaper buy, if your tap water tests below 500 mg/L of dissolved solids, is a UV+UF purifier, which has no membrane to replace.

What makes up RO purifier maintenance cost?

Four lines make up the bill: filters, the membrane, the UV lamp and the service visits, and on a five-year model built from June 2026 Indian part prices, the visits outweigh any single part.

Most advice fixes on the membrane, the dearest part. Wrong fear, in our view: on our assumptions it is changed twice in five years, while the doorbell rings ten times. The prices below are reported ranges from RO Care India's 4 June 2026 Kolkata price guide; it is a service network, not a brand.

Worked example: an RO+UV purifier on hard city water, top-of-range parts, a service visit every six months. Our assumptions, not the guide's: filters yearly, the membrane twice in five years, the UV lamp once.

• Sediment filter (₹150-400) plus carbon filter (₹200-500), one set a year: ₹900 x 5 = ₹4,500.
• RO membrane (₹1,200-3,500), twice: ₹3,500 x 2 = ₹7,000.
• UV lamp (₹800-1,800 with installation), once: ₹1,800.
• General service visit (₹400-800), twice a year: ₹800 x 10 = ₹8,000.

That totals ₹21,300, about ₹355 a month. At the bottom of every range it falls to ₹8,950, and since the guide ties membrane prices to quality and brand compatibility, much of that gap is the part you accept. Your own version takes under a minute: last service bill times ten, plus five years of filter sets at your technician's price. That is your floor before any membrane.

It is the same ownership sum as the payback for 3-star and 5-star refrigerators, and it belongs before the festive sale, when a list price dressed up as a discount flatters the box.

Four more numbers decide whether your bill stays near that floor. Two come from the same guide, the others from BIS's IS 10500:2012 standard and the National Green Tribunal's May 2019 order, which the environment ministry reported to the Lok Sabha on 13 March 2020.

The overhaul is the figure to plan around. Skip visits to save money and the saving can return as that one bill, bigger at either end of the price range than a year of routine upkeep.

RO or UV: which costs less to keep?

UV+UF costs less to keep because it has no RO membrane to replace, though the saving only counts if your tap water already sits under 500 mg/L of dissolved solids, BIS's acceptable limit.

Find your row, then read across at your technician's price. The UV+UF row drops the membrane but omits the UF cartridge, which the guide does not price, so read it as a floor. The last row is an RO with four years of parts and service included, leaving only year five to pay.

The tinted cell is the worked example. Midpoint prices take ₹6,175 off it; UV+UF at the same prices removes the ₹7,000 membrane line.

That last row describes a real product. Eureka Forbes's product page listed the Aquaguard Aspire Glow 4X, an RO+UV model, at ₹18,999 as of 22 October 2026, promising "zero filter cost, zero spare cost, and zero service cost for full 4 years". If both membrane changes and the lamp fall inside those years, the cover absorbs ₹18,800 of our high-end bill, about the purifier's price. That is the company's claim, not a tested result. Our call: where an RO is needed, an all-in warranty beats a separate AMC, if the card agrees.

What TDS level actually needs an RO?

Above 500 mg/L, the acceptable limit in BIS's IS 10500:2012 drinking water standard, an RO earns its upkeep; below it, a meter reading alone cannot clear the water, because it measures dissolved solids, not germs or trace arsenic.

Regulators have drawn the same line. The National Green Tribunal's 20 May 2019 order directed the environment ministry to prohibit RO below 500 mg/L, the ministry told the Lok Sabha. The Supreme Court stayed the key directions of a tougher NGT order of 1 December 2021 "until further orders" on 1 March 2022, Verdictum reported, and we found no report of the stay lifting by October 2026. No ban binds you.

Here is the grey area, and it is opinion, not rule. Water under the limit that tastes salty passes BIS and can still taste poor; an RO to fix that is defensible, just not a safety purchase. Most homes need the test, not the machine. Or rather, most homes on a municipal line do.

• Borewell or tanker water: a meter cannot see arsenic or bacteria, so get a lab test against the IS 10500 drinking water limits first.
• The MRP gap: that Eureka Forbes page shows a ₹31,000 MRP; read what India's 2026 e-commerce rules make sellers show about prior prices before calling the gap a saving.
• Reject water: on low-TDS water, an RO drains water for no safety gain.
• AMC quotes: price any contract against two single visits and one filter set a year.

This week, buy or borrow a TDS meter and test the kitchen tap. Under 500 mg/L on a municipal line, buy a UV+UF and keep the membrane money. Over it, buy the RO, and sign no AMC until you have set it against the ₹21,300 high case.

Free apps never send you a bill, so you pay in permissions. A flashlight app that wants your location, a game that wants your microphone, a shopping app that wants to draw over everything else: each request looks harmless when you tap Allow, and most of us never look again. An app permissions cleanup is the fastest way to stop paying, and most of it takes minutes.

Do Apps Ask for More Than They Need?

In July 2025 a UK consumer test of 20 popular smartphone apps flagged the permissions it considers risky, and precise location, microphone and draw-over-apps stood out as the three to question first. The full study is on Which?'s own site.

Precise location matters most because it is more than a city. Once an app holds it, the app can tell which shop you walked into and which car park you left from. Most of the tested apps asked for it, which tells you the request is routine, not rare. Shopping apps play the same game on prices, as our fake festive discounts piece shows.

Draw-over-apps is the sleeper. It lets an app sit above the screen you are using, and nothing visible happens when you grant it, so nobody notices. Revoke it from any app that has no floating tool job, such as a chat head or a screen magnifier you actually use.

Microphone access needs the least thought. Or, more precisely, it needs the least argument. A recipe app has no reason to listen, yet it can ask. Ask one question per app: would it miss the microphone if it went quiet tomorrow? If the answer is no, revoke it. Pick your five most used apps and run that check on each, which takes under a minute apiece.

These four numbers come from Which?'s July 2025 study of 20 popular smartphone apps, and each one maps to a setting you can check yourself.

The first card is the one to act on. Precise location turns a grant into a running record of where you go, and one revoke ends it without uninstalling anything. Keep it only where a feature stops working without it, like turn-by-turn directions. A ride-booking app needs it while you book, not all day.

Which Grant Do You Cut in an App Permissions Cleanup?

Revoke a grant when the app's main job never needs it, and keep it only where a feature you actually use stops working without it, since the table below sorts the common cases so you can decide each one in seconds.

Read the table left to right: find your situation, then take the action in the middle. The right-hand column tells you what you give up, which is usually nothing you were using.

Decide which grant to check first on your own phone. The bars show how many of the 20 apps Which? tested asked for each grant in July 2025.

Can Revoking a Grant Break an App?

Revoking a grant does not uninstall the app, but it can stop one feature you rely on, such as maps or call alerts, so test before you move on. Change one grant at a time, open the feature, and only then revoke the next.

Some apps degrade quietly. A map might show a location dot that never moves, and a messaging app might refuse to send photos without storage access. Those failures look like bugs, so people blame the app and re-grant everything. The fix is patience, not an all-or-nothing switch. If you want the background to the consent prompts many apps now show, our piece on India's consent manager deadline explains where those prompts come from. The policy side of AI-made content is in our piece on India's AI labelling rules, which shows how slowly the rules move from paper to practice.

• Map and ride apps: keep location on While using the app.
• Call and voice apps: keep the microphone for the days you call through them.
• Shopping and game apps: revoke camera, microphone and draw-over-apps first.

This week, open your permissions list and revoke location and microphone from any app you have not used in the last month. Then check the list again on Sunday to confirm the changes stuck.

₹32,990 buys a Sony 43-inch 4K TV this festive season, wearing a "45% off" badge against a ₹59,900 list price. Buy it at ₹30,240 or less, or wait. PriceHistory's September 2026 listing puts that set's recorded low at ₹28,800, which makes the badge a fake discount in the only sense your wallet cares about: it measures the saving from a price almost nobody paid, while the ₹4,190 gap to the low is what decides whether today is the day.

Why a fake discount survives every festive sale

It survives because the percentage is calculated from a list price the seller chooses, while the figure that decides value, the lowest price buyers recently paid, never appears on the product page at all.

A small badge is no safer. The same PriceHistory page shows an LG 9 kg front-load washer at ₹37,990, a modest 5% off, against a recorded low of ₹33,291: still about 14% above what it has sold for. The iPhone 16e at ₹69,900 carries a 54% premium over its ₹45,338 low. Buying a fridge too? Run this test before the 3-star vs 5-star refrigerator payback question.

The usual advice, compare against MRP, is out of date: MRP is a ceiling on the label, not a price the market sets. India's 2026 e-commerce amendment rules on prior prices push platforms toward showing real pre-sale prices. Until every listing does, a free tracker will.

Our rule, and it is a judgement rather than a citation: buy when the price sits within 5% of the recorded low. For the Sony that ceiling is ₹30,240, missed by ₹2,750. PriceHistory lists a 10% instant discount on SBI cards, which would bring the TV to ₹29,691 and pass, but only if the offer's cap is at least ₹3,299. The cap is not stated.

Four numbers decide how much checking falls on you, from Bar & Bench's 1 September 2026 review of CCPA dark-pattern orders, Storyboard18's June 2025 report on the self-audit advisory and PriceHistory's data from past Great Indian Festival sales.

The fines matter most. Bar & Bench counts seven platforms penalised, at ₹1 lakh to ₹7 lakh each: Zepto and FirstCry for drip pricing, BookMyShow for basket sneaking. For a national festive sale that is a rounding error, so the patterns persist and your own two-minute check is the protection that works.

So the test has to be portable, something you can run on a phone in the minute before you tap Buy.

How do I know if a discount is fake?

Compare the sale price with the item's lowest recorded price on a price history tracker, and treat it as genuine only when it falls within 5% of that low after every card offer and checkout fee is counted.

Read each row as a question to answer before paying.

Two rows carry the decision: the trigger gives a yes or no, and the card cap can turn a near miss into a real deal.

Where do I complain about dark patterns in India?

File with the National Consumer Helpline on 1915 or 1800-11-4000, by WhatsApp or SMS on 8800001915, or at consumerhelpline.gov.in, and attach screenshots, the URL, your cart or order ID, and the date and time.

But the product-page price is not your price until checkout agrees. A LocalCircles survey, reported by the Internet Freedom Foundation in October 2025, found drip pricing the most widespread dark pattern on Indian marketplaces; no sample size is given, so read it as a direction. Look at the final payable amount. Consumer Affairs Minister Pralhad Joshi has called extra cash-on-delivery charges a dark pattern and his department opened a probe, so a late COD fee is reportable.

A pay-later prompt at checkout can add interest to an overpayment; read how credit on UPI is changing what Indians spend before accepting one. Complaints also go through the Jagriti, UMANG or NCH apps, escalating to e-Jagriti.

One grey area remains. Trackers are third-party data with unstated methods, and a recorded low can be a flash price that lasted an hour, or less. Our view: it still beats a list price, but if the low looks like a glitch, use the next-lowest point.

• A countdown timer beside a price the tracker shows as routine.
• A fee, cash-on-delivery included, that appears only on the final screen.
• An item sitting in your cart that you never picked.
• A bank offer whose rupee cap is not shown before you pay.

The decision: this week, before the Diwali sales peak, put every item on your list into a free price tracker and write its 5% ceiling beside it. Buy what dips under it. Let the rest go, whatever the badge says.