That indicates her marketing messaging and promotional stuff is not aligned with what he is delivering

Exclude the pincodes which have highest returns, call and verify the details of the orders when placed on the site.

1. How fast does he deliver? A product like this needs instant delivery before the customer changes his mind.

2. Is the product similar to the way it was promoted?

3. Do you think the delivery execs or delivery partners have something to do with it? (Like for instance, I used to get returns because the delivery partner I chose back weren't professional.)

4. Does the product offer value for the money? Also, is it easily available in the local market at any store?

5. Are you sure that those fake orders aren't by your competitors (just to increase your spend.) Find out the IP addresses and technical details of all the cancelled orders.:)

Call and confirm before dispatching cod orders. If possible completely discontinue cod orders. Add sampling tea bag to each dispatch. If the customer dislikes it and uses just that tea bag while leaving the rest intact then issue full refund on the return

 Why do they return?

1. Use whatsapp api for a notification and an embeded url to be clicked, showing all info filled in previous step to reconfirm.
2. Use a whatsapp bot to reconfirm.

This metric can be used for retargeting too

Have you done any analysis what's causing the return 40% is a huge number do some basic RCA and filter out using a Pareto principle,
Take consumer feedback about the product
Is shipment is an issue , products are being undelivered etc ?

Apart from what people suggested, you can push the online payment orders. The one way will be charging the COD option.If Facebook ads is the primary source of revenue.Make Lookalike audience for delivered orders as well as online payment customers. Ihave many clients where 70-80% orders are prepaid.It takes time, but it can be achieved in some months time by pushing the prepaid. you can also do aggressive BOF campaigns to reinforce the branding and push some discount on prepaid orders. 40% return is significant and getting majority of customers for prepaid orders is the way to go. The LTV value of customers is good in these niches so focus on existing customers as well.Those are the easy ones to target for online payment for repeat orders.

agreeing with many of the above comments,
a communication with the customer before delivery can be a key solution to your problem,

you can inform the customer at the time of order placement that someone from the CSE team will call for confirmation or feedback

what i can think of is
Nominal Charges on Returns considering the market (India) your customer is from. (5/10% of invoice amt)

To me 40% return is an indicator for optimising ur funnel.

Check the messaging, collect data like if from why're they returning.

Then analyse is there a gap between what you say nd what delivered.

Maybe a discount on online payment could help.

You need to collect data, analyze and keep on trying various things till 40% becomes less than 20%.

Once again optimization at all level is the key

 In my experience in the tea segment 40% return is too high. Shouldn't be more than 15.. Whats the ad target? You need to check if your targeting is proper and suspend cod temporarily.

Thirdwatch from Razorpay lets you flag risky transactions, impulse purchases, fake addresses, fight COD fraud and reduce RTO losses by upto 30%. There's also a 1 month trial to experience the platform at 0 cost with one click plugin install for top e-commerce platforms
https://razorpay.com/thirdwatch/

So I have a shopify store - Dohful Cookies (dohful.com). The reason I'm mentioning is because when I was starting I took a conscious decision of not having COD even if that meant loosing a few customers. Because as you may see on the website out cookies are upwards of Rs. 299 for a pack of 4 cookies. So obviously not everyone is a customer. So we thought anyone who's ready to spend on cookies is probably well versed with online shopping, usage of credit cards etc. And is comfortable taking a risk of Rs. 300 - Rs. 500. Such a customer is looking for the most genuine and amazing cookie experience and not COD. That's why we don't have COD. And truthfully speaking we've never lost any genuine customer to this. So yes, what I'm trying to say is that maybe your client if dealing in lifestyle products should clearly identify their customers and whether COD even classifies as a value proposition for their customers.