What Happens When a Complaint Reaches the Consumer Helpline

A complaint that leaves your channels does not disappear. It changes address. And the address it most often changes to in India is the National Consumer Helpline.

Our own coverage of helpline data found roughly 5.1 lakh complaints filed against ecommerce and quick commerce platforms in 2025, up nearly 14 percent from about 4.4 lakh in 2024. The largest categories were not exotic. Non delivery, wrong product delivered, replacement failures and refund issues. In other words, ordinary operational misses that were not closed at the first attempt.

That is the useful frame. The helpline is not primarily a legal problem. It is a mirror held up to your fulfilment and refund operations, and the reflection arrives with a response deadline attached.

The path a complaint takes once it leaves your inbox

A consumer who has given up on your support usually registers on the helpline through the portal, the app, the toll free line or a message. The complaint is logged with a docket number and routed by category to the company or platform named in it. Many large ecommerce and quick commerce companies participate as convergence partners, which means the complaint lands directly in a company facing queue rather than in a general pile.

Your side receives it as a case with the consumer’s account of what happened, the order reference if they supplied one, and a request for the company’s response. You respond in that same system. The consumer sees the response and either accepts it and the docket closes, or does not, in which case the docket stays open and the consumer has a documented record of having tried.

That record is the part sellers underestimate. An unresolved helpline docket is not a dead end for the consumer. It is the evidence they attach when they file formally before a consumer commission, and filing is largely online now. It is also one of the inputs that shapes how the Central Consumer Protection Authority reads a company, because the authority looks at patterns across many complaints rather than at any single one.

Why response speed is a commercial question, not just a compliance one

There are prescribed response expectations, and they are set by rules that have changed and that read differently depending on how your entity is classified. Do not lift a number from a blog. Confirm the current position for your category and turnover with your own counsel.

The commercial case does not depend on that number. A complaint answered quickly, with the order reconstructed and a concrete remedy, closes at a cost you already knew about, which is usually the refund or the replacement. The same complaint answered slowly or generically costs the refund anyway, plus the operational time of a second and third round, plus a docket that stays open in a public dataset, plus the chance that the consumer escalates formally and the cost stops being predictable. Speed is the cheapest lever you have and it is entirely within your control.

Prepare the response file before you write the response

Most weak responses are weak because the person writing them could not find anything. Build the file first. For the disputed order you want the order identifier and timestamp, the listing and price exactly as displayed on the purchase date, the payment gateway reference, the courier docket with the complete scan history rather than a summary status, proof of delivery including any image or delivery confirmation the courier captured, every message exchanged with the customer across chat, email and phone notes, the refund or replacement reference with its own date and settlement reference, and the version of your returns and refund policy that was live when the customer bought.

Then write in a specific shape. State what happened factually and in order. State what you have already done. State what you are doing now, with a date. Give a reference the consumer can check, such as a refund reference or a fresh docket. Avoid restating policy at a person who is complaining precisely because they think your policy was applied badly. Policy language reads as a refusal even when it is not one.

Assign one owner for this queue. Helpline responses handled by whoever is free produce inconsistent answers to identical fact patterns, and inconsistency across similar complaints is exactly the shape a regulator finds interesting.

Read the complaint mix as an operations signal

Now the part that pays for the whole function. Tag every docket by root cause and read the distribution monthly.

Non delivery clustered on one courier partner or one pincode band is a lane problem, not a consumer problem. Wrong product delivered clustered on one warehouse or one set of similar looking SKUs is a picking and barcoding problem. Replacement failures usually mean reverse pickup is not being attempted the number of times you think it is. Refund complaints often mean the money moved on your ledger but the customer never got a message saying so, which is a communication failure being reported as a financial one.

Compare that distribution against your internal ticket mix. The two rarely match, and the difference is the interesting bit. Categories that are over represented at the helpline relative to your own inbox are the categories where your first line said no and the customer did not accept it. That is a policy calibration problem sitting one level above the agent.

The brands that get their helpline volume down do not do it by writing better responses. They do it by reading the dockets as a defect log and fixing the three lanes that generate most of them. The response quality then matters less, because there is less to respond to.