We messaged 75 Indian brands as a buyer. 31 still haven't replied.

I wanted to buy a sling bag. Simple question: is it in stock, and can I pay cash on delivery in Delhi?

Now multiply that by 75. Between 25 and 30 October 2026, we sent that same kind of question to 75 Indian D2C brands across beauty, fashion, food, jewellery, wellness and home. One real product from each brand's own site. One polite message. No links, no pitch.

By 1 PM IST on 30 October, 31 of the 75 had not replied at all. Of the brands that did reply, a human took a median of 3 hours 48 minutes to write back.

These are early results from our Reply Report study. It closes on 1 November and the final report lands next week. But the pattern is already clear enough to act on, and I would rather you fix your inbox before Diwali than read about it after.

Key takeaways

• 75 brands asked, 44 replied, 31 silent so far. 12 of the silent ones were asked on 30 October, so give them time. The other 19 have had between 1 and 5 days.
• Customer service response time in India is slow when a human writes the reply: median 3 hours 48 minutes across 29 human replies.
• Speed and quality went together. Humans who replied within an hour answered fully in 6 of 8 cases. Slower humans answered fully in 9 of 21.
• Site chats replied within minutes every time we got through. Their answers were often half answers.
• 5 brands asked for a phone number or an order ID before they would answer a question from someone who had not bought yet.

How did we test Indian brands' reply time?

We acted like a normal buyer. For each brand we picked a product that was live on its site, then asked whether it was in stock (in a size, where sizes applied) and whether COD works for Delhi. We used the site chat when it was open to us, and the email address printed on the brand's own site when it was not.

The sample so far is 75 brands, 15 a day on 25, 26, 27, 29 and 30 October. 68 messages went by email and 7 through site chat. We logged the send time in IST, every reply, who seemed to write it, and whether it answered the question.

Two honest limits. We did not test WhatsApp, and plenty of these brands do most of their selling there. And a reply time from email is a different thing from a WhatsApp reply time. But a buyer who writes to your support address is a buyer, and this is how many of them are treated.

How many brands replied, and how many stayed silent?

44 of 75 replied in some form. 31 stayed silent. Here is how the 44 break down.

• 29 replies were written by a person.
• 10 came from a bot or AI chat.
• 2 were ticket auto acknowledgements with nothing else so far.
• 3 we could not call either way.

Leave out the 15 brands we wrote to on 30 October and the picture is still rough. Of the 60 brands asked between 25 and 29 October, 19 have not replied to a pre purchase question after 1 to 5 days.

Category made a difference. Jewellery was the worst group in our sample. Of the 8 jewellery brands asked before 30 October, 4 had not replied. That is a category where one order can be worth thousands of rupees.

How long did Indian brands take to reply?

When a human wrote the reply, the median time was 228 minutes, so 3 hours 48 minutes. The fastest human replied in 4 minutes. The slowest took a little over 2 days.

The spread is wide. 8 of the 29 human replies came within an hour. 9 came in 1 to 4 hours, 5 in 4 to 12 hours, and 7 took more than a day.

Site chats were a different story. All 7 we got through to replied within 8 minutes, and 6 of them, all bots, within 3 minutes.

Speed is the part most founders measure. But look at what the speed was worth.

Did the replies actually answer the question?

Not often enough. Of the 40 replies with real content, 19 answered both parts (stock and COD), 10 answered one part, and 11 did not answer at all.

The humans who replied fast were also the ones who answered well. Within an hour, 6 of 8 human replies answered fully. Slower than an hour, 9 of 21 did. I do not think that is an accident. A team that replies fast usually has the answer in front of them: stock data, COD rules, the product page. A team that replies the next day is often reading from a script.

Bots answered fully 4 times out of 10. The usual gap was COD. They knew the product, they had a checkout link ready, and then they went quiet on whether cash on delivery works for a Delhi PIN code, which is the one thing a lot of new Indian buyers need before they pay.

What kinds of replies lost the sale?

Some replies were worse than silence. A few we saw:

• The wrong stock answer. Three replies told us a product was out of stock or not in the catalogue while the brand's own store data showed it available. One came from a beauty brand's agent within 27 seconds. One from a snack brand's named agent in 4 minutes. One from a coffee brand's AI chat, while the page itself showed Add to Cart. I wrote up what that AI chat did, step by step.
• The phone number wall. 5 brands asked for a phone number or an order ID before answering. One skincare brand's ticket asked for our Order ID and registered contact details. We had not ordered anything. That was the point. Several site chats also refused to start without a phone number, so we used email for those.
• The template. Three skincare brands under one parent company sent the identical ticket mail promising a reply in 2 to 3 days.
• The redirect. One fashion brand's auto reply said support had moved entirely to WhatsApp. On 13 brand sites in our sample, the chat button simply opened WhatsApp.

None of this needs a big budget to fix. It needs someone to own the inbox, and an answer sheet next to them.

What is a good customer service response time for an Indian D2C brand?

For a pre purchase question, aim for under an hour during working hours, with a real answer. In our data, the brands that replied within an hour also answered properly most of the time, which is what turns a question into an order.

If you are on WhatsApp, the bar is higher, because buyers there expect a chat. I covered the WhatsApp side in how fast to reply on WhatsApp, with data from 30 brands, the earlier cut of this same study, and the money side in what a missed lead costs.

Should you put a bot on it?

A bot fixes speed. It does not fix knowledge. Our bots replied in a minute and still missed the COD question most of the time, and one told a buyer a product was out of stock while it was on sale.

If you add a bot, give it your real stock data, your COD and delivery rules by PIN code, and a clear hand off to a person for anything it cannot answer. Then read its chats every week. I wrote more on this in the piece about Meta's new AI agent on WhatsApp.

Reply audit: do this today

1. Send your own brand the same question we sent: "Is [product] in stock, and do you offer COD to Delhi?" Use email, site chat and WhatsApp.
2. Time each reply in minutes. Write it down.
3. Check if each reply answers both parts. Stock and COD.
4. Remove any step that asks a pre purchase buyer for an order ID or phone number before answering.
5. Make a one page answer sheet: stock check, COD rules, delivery time by region, return policy.
6. Pick one person who owns the inbox from 10 AM to 8 PM, with a backup.
7. Set a target: a real answer within 60 minutes in working hours.
8. If you run a bot, read 20 of its chats this week and fix every wrong answer at the source.

What happens next with the Reply Report?

The study closes on 1 November with a sample of about 85 brands, a little short of the 100 we planned, because some brands had no way to reach them without a login, or offered a button that went straight to WhatsApp. The final report, with the full numbers and the method, lands next week. Every number in it will come from the log, the same as the numbers here.

Want a second pair of eyes before your next buyer writes in?

Most of the leaks in this study are fixable in a week: an owner for the inbox, an answer sheet, and no phone number wall before a buyer can ask a question.

Start with the free Leak Check. It takes 2 minutes and shows where your funnel loses buyers between the ad, the page and the reply. If you would rather talk it through, message me on WhatsApp and I will reply myself.