Get live statistics and analysis of Kanish's profile on X / Twitter

After selling sandals for ₹2500+ to both men and women purely with ads, with no discounts, here are 4 learnings on ecommerce marketing that we refer back to again and again 1/ What you talk about >> How you talk about? We first justified higher price with better fabric, insole…. but overtime this ad angle plateaued. Then shifted to design based messaging and scaling has been much smoother ever since Design is something people inherently value more as it has both appeal + bragging rights, while fabric just gives them comfort with no bragging rights 2/ Men and women convert to different styles of creative Men prefer less loud, more text and want to know about highly functional aspects of product to convert Women want to know more about aesthetics, overall appeal, trust social proof (reviews, testimonials) more. and how product fits into their existing collection/routine (last point is critical) 3/ Create creatives for niche identity Tested this first in August. These creatives speak to the identity of audience of a specific segment like, people who think of themselves as travel freaks, foodie…. Placing the product as part of their identity connects better 4/ First purchase is for design, Second purchase is for quality Quality of fabric cannot be felt from a screen, so people justify first purchase looking at design. Assuming a good design means better fabric also Don’t discount design for functionality

Crossed the first 2L/day ever for this brand, it's always a good feeling For a brand that launched just 5 months ago, this is a🤌🏻 The secret -> Started with just 2 products, and avoided pushing all products to ads. First build learning in two then expand this is way faster -> No iterations on ads. For new brands, they shouldn't iterate initially, goal must be to test as wide a variety of ads one can test. We tested over 19 angles per creative, and still testing -> Creatives focussed on lifestyle scenarios of where all products can be useful and how it saves time. Instead of detailing about product features, mentioned what does it do for customer

There are 10,000 Indian ecom brands under ₹10cr revenue (don't recall where I read this) In 5 years? Many of them will blow up into big businesses and then they'll need tools like A/B testing Post-purchase upsells. Quizzes. Reviews. Surveys. But most tools are expensive with USD price Here, a SAAS vendor, can provide 80% of the features 20% of the price (many are doing this in other areas) This would be an interesting GTM play

Was talking to a apparel brand founder the other day, his numbers were Revenue of 35L per month with a ROAS OF 3X His concern was “As I increase the ad budget, my ROAS starts to tank, and becomes unprofitable” With scale ROAS does dip, but entirely falling of a cliff says something isn’t right Now, they had 3 to 4 workhorse ads in their ad account, getting the most ROI and spend So, the natural inclination of the brand was to scale, produce more of such assets Here’s where it doesn’t go as planned → Creating more of the same creatives pushes the algorithm into a eco-chamber trying to find just similar audience as before. Eventually, this eco-chamber runs out of gas and performance dips → For every ad creative, there is a messaging that is governing the creative, and sometimes the messaging isn’t scalable and hence as one increases spend it falters The approach to take here is to start testing more diverse creatives on a smaller scale Creative diversity ensures that new audiences are reached, and existing ones do not fatigue from seeing the same creative over and over again If the majority of the winning creatives talk about comfort in apparel, next set of creatives can show social proof, introduce other categories. This attracts audiences with different needs, and also convinces audiences in separate stages of the buying journey

This is a hard and depressing read It shows the reality of e-commerce in India and pure grit of founders to march ahead This phrase "Customers expect international quality but want fast fashion pricing" does describe a lot There are bad actors in everything, but here it is quite down under Returns are profit crushers indeed, in fashion this zooms even more While we discuss on reducing that CAC or increasing ROAS, this is just 1/3rd off most business There is 2/3rd left unoptimized and returns are a big factor where people tend to return even if product quality is good Marketplaces have started to shift their policies and stopped being lenient, D2C brands (even new ones) must also set hard guardrails on returns, exchanges Here are a few things you can do On Product pages ✅ Detailed product pages to reduce wrong orders High-quality images from multiple angles, zoom-ins on fabric texture, and 360-degree views. Precise size guides with actual garment dimensions + a "Compare with your existing item" feature. ✅Offer free returns only for prepaid orders. COD orders must pay a small return insurance fee (e.g., ₹49-99) if they want a return option. ✅Don't hide your return and exchange policy Do not link main details of returns to another page, instead convey upfront in the product page - How many days are returns accepted? What is the requirement for returns? In General ✅Use a payment gateway that flags high-risk orders (e.g., Cashfree, Razorpay, or specialized fraud detection tools). Avoid COD only on high-risk orders (e.g., repeat customers). ✅Offer a 5% discount for products that are final sale / no return ✅For products which are sold on sale, make all orders as final sale ✅Accept returns only if the original seal is intact ✅Adjust return window & Shorter return windows ✅Encourage size exchanges instead of full refunds. Incentivize store credit with bonus discounts ✅Flag customers and pincodes who return frequently and limit their access to COD Then there are also good consumers, here it is the responsibility of the business to not go over-board with marketing and setting unrealistic expectations from product Underpromise and overdeliver should be the approach

Had a 2 hour call with a brand (does about 1.5 crore per month) and a major part of the discussion was what are the next set of hires for 5 crore per month target The longest time spent on - Video editor They have become the dark horse of scaling that no one talks about, but they hold the keys to your brand's success A good video editor will cost you 1.5x that what you generally plan for, but a good one is worth even 2x that We have also started to think about incentivising a performance pay for video editors based on results... to align incentives even better

One of the easiest way to save marketing budget in Meta is checking view through conversions In the ads dashboard, under the columns section, head to attribution settings and tick all boxes If there is more than 15% to 20% of purchases from the view columns, it means ads are over-indexing on brand aware (mostly existing customers) Because conversions are happening with just view of the ad and not a click, this means impressions of the ad are mostly towards brand aware who are then going to Google and converting This is a wasted spend, as in most cases these audiences would have purchased even without the ad, ad had limited impact but took all the credit This is common in Advantage+ campaigns, so we try to set existing customer cap to either 0 or less than 5%

Scaling to 10 Lakh per day profitably just from ads

Meta ads is truly amazing, works for a brand doing 10 orders a day and also for brands doing 100,000 orders a day - The best distribution tool ever built

20 L crossed in first 10 days of the month ROAS - 3.28 ✅

Just like UPI, govt should build UCI - United Checkout Interface for ecommerce stores This will auto-populate checkout data, remove COD option for risky customers and reduce friction to buy Conversion rate 🚀🚀

Spending 1L per month in ad account or 3L per day, our creative testing framework doesn't change much Try an ABO with multiple ad sets grouped by concept - Test for 3 to 4 weeks Is it able to get most spend to winners? Great, continue If not, try same setup but with CBO If both don't work Fix your creatives, not your creative setup Creative testing setup doesn't really make a difference, it's not science You should instead dedicate more time to making new creatives

And finally it's 10L in a single day✅ First ever Working with this brand since day 1 Starting at 0 revenue

In the early stages, you should try to sell to nerds in your category Like selling protein to gym freaks They are nerds and they signal status by being on top of the trend, so they are early adopters But Sell features, not benefits They know all the benefits already Also, nerds have high LTV and big word of mouth if your product lives up to their expected standard

We were stuck at 2.1 to 2.3 ROAS for 4 months with an ad account Finally broke the ROAS ceiling and scaled to 3.1 ROAS 5 shifts in ad creatives that helped improve ROAS ✅ Women Convert Better Based on Identity These creatives speak to the identity of audience of a specific segment like, people who think of themselves as travel freaks, foodie…. Hook - "If you are a foodie like me" Placing the product as part of their identity connects better ✅ Oddly satisfying and uncomfortable statics We lacked in statics and over reliant on videos. To solve this we started making statics that actually stood out just with visuals. The biggest inspiration for visuals was Reddit, in that subreddits of product category ✅ Claim-based headlines This product was “India’s 1st” and ad with this hook ran through the entire account and lifted the ROAS of the whole account by over 20% . Think of any genuine claims about your product and then use it as ad hook ✅ Trust = Reviews + Visual Proof + Transparency Customers don't just want to know what you're selling but who is buying it. There were multiple review ads each with it’s own ad angle. Each ad angle = Each feature product, this helped reach wider audience as each customer values different functions in the same product ✅Retarget Aggressively with Contextual Ads Use cart abandonment and browsing data to retarget prospects with highly personalised ads. Examples For undecided users: “Still thinking? Here’s why 200+ people love it.” For cart abandoners: “Complete your purchase and get free shipping.”

Created & launched 90 ad creatives in first 2 weeks of brand launch You have to win the creative game with volume We have built a system that helps us produce this creative volume from the get-go, and we start building on it from day 0 let me know if you are interested to know more about the process, will consider a thread if there is enough interest

There was a discussion around debt and funding last week, so created a calculator over the weekend that gives a rough estimate of max possible growth rate according to your revenue and margins

And finally it's 10L in a single day✅ First ever Working with this brand since day 1 Starting at 0 revenue

There are 10,000 Indian ecom brands under ₹10cr revenue (don't recall where I read this) In 5 years? Many of them will blow up into big businesses and then they'll need tools like A/B testing Post-purchase upsells. Quizzes. Reviews. Surveys. But most tools are expensive with USD price Here, a SAAS vendor, can provide 80% of the features 20% of the price (many are doing this in other areas) This would be an interesting GTM play

After selling sandals for ₹2500+ to both men and women purely with ads, with no discounts, here are 4 learnings on ecommerce marketing that we refer back to again and again 1/ What you talk about >> How you talk about? We first justified higher price with better fabric, insole…. but overtime this ad angle plateaued. Then shifted to design based messaging and scaling has been much smoother ever since Design is something people inherently value more as it has both appeal + bragging rights, while fabric just gives them comfort with no bragging rights 2/ Men and women convert to different styles of creative Men prefer less loud, more text and want to know about highly functional aspects of product to convert Women want to know more about aesthetics, overall appeal, trust social proof (reviews, testimonials) more. and how product fits into their existing collection/routine (last point is critical) 3/ Create creatives for niche identity Tested this first in August. These creatives speak to the identity of audience of a specific segment like, people who think of themselves as travel freaks, foodie…. Placing the product as part of their identity connects better 4/ First purchase is for design, Second purchase is for quality Quality of fabric cannot be felt from a screen, so people justify first purchase looking at design. Assuming a good design means better fabric also Don’t discount design for functionality

Have been into helping ecommerce brands with marketing for last few years Ask Anything, will try to share the best of my knowledge and experience

If most DTC brands in India sold on product quality rather than marketing ability, we would have less than half the brands

20 L crossed in first 10 days of the month ROAS - 3.28 ✅

Fashion brand founders, Which of the two size or color do you ask customers to select first? I am on team size first, but would love to hear team colors and their thinking behind it

80L+ sales for high-ticket jewellery brand, all online only About 2 months back, started working on performance marketing ads with a new brand that manufactures and sells lab-grown diamonds. Some tidbits for such high-price purchase -> Offline presence One of the first questions customers ask is “Do you have an offline store?”, though they will never visit the store. Just knowing there is one increases the trust of the customer with brand from 0 to about 70. Offline store shows seriousness in business and a validation of not a scam -> On creatives The best-performing creative is a 3d video, with no text, no price. Just logo and product. Even the quality of leads are from such creatives. The classic KISS principle -> Multiple funnels are required In such high-priced products, buyer behaviour is complex. You will find leads talk to you for 7 days straight, then ghost for 2 months and then come back and buy in 2 days Need more than one purchase conversion funnel along with brand awareness activities -> Audience is aware Over the last few months, we have actually seen the audience go from questioning “What is lab-grown diamond” to being fully aware and ready to buy. Indians are accepting this, so good growth years ahead -> Men vs women buying behaviour Female leads are more responsive but the conversion percentage is higher with male leads. So, the plan is to engage female with direct buy ads and male with gifting ads

This is a hard and depressing read It shows the reality of e-commerce in India and pure grit of founders to march ahead This phrase "Customers expect international quality but want fast fashion pricing" does describe a lot There are bad actors in everything, but here it is quite down under Returns are profit crushers indeed, in fashion this zooms even more While we discuss on reducing that CAC or increasing ROAS, this is just 1/3rd off most business There is 2/3rd left unoptimized and returns are a big factor where people tend to return even if product quality is good Marketplaces have started to shift their policies and stopped being lenient, D2C brands (even new ones) must also set hard guardrails on returns, exchanges Here are a few things you can do On Product pages ✅ Detailed product pages to reduce wrong orders High-quality images from multiple angles, zoom-ins on fabric texture, and 360-degree views. Precise size guides with actual garment dimensions + a "Compare with your existing item" feature. ✅Offer free returns only for prepaid orders. COD orders must pay a small return insurance fee (e.g., ₹49-99) if they want a return option. ✅Don't hide your return and exchange policy Do not link main details of returns to another page, instead convey upfront in the product page - How many days are returns accepted? What is the requirement for returns? In General ✅Use a payment gateway that flags high-risk orders (e.g., Cashfree, Razorpay, or specialized fraud detection tools). Avoid COD only on high-risk orders (e.g., repeat customers). ✅Offer a 5% discount for products that are final sale / no return ✅For products which are sold on sale, make all orders as final sale ✅Accept returns only if the original seal is intact ✅Adjust return window & Shorter return windows ✅Encourage size exchanges instead of full refunds. Incentivize store credit with bonus discounts ✅Flag customers and pincodes who return frequently and limit their access to COD Then there are also good consumers, here it is the responsibility of the business to not go over-board with marketing and setting unrealistic expectations from product Underpromise and overdeliver should be the approach

Here is how we are using it ✅Creative team 1/ Research Here we input all the details about the brand and any other details we have and then ask a bunch of questions to understand What kind of audience we can target? What kind of creatives we can make? Ways to position a particular product? This is a tool, so final say is always an expert on this skill only 2/ Custom GPTs for product/brand You can create a custom GPT for your own brand and input all the product and brand files into it, then use it a lot for brainstorming. Into it you can also include ad scripts and then brainstorm further 3/ Bridging Creative Strategy & Design One big pain: The strategist thinks in ideas - the designer needs visual direction. We use GPT to convert ideas into visual briefs 4/ Visual Ideation via Image GPTs Now that GPT can create visuals, we use it to: → Generate first-draft ad layouts → Explore different ad styles: meme, UGC, emotional, infographic → Get visual inspiration without needing a designer for round-on 5/ Our copywriters use GPT for: → Rewriting product benefits in simpler/more punchy ways → Headline testing (e.g. “make this more urgent/emotional/playful”) → Finding better analogies and metaphors → A/B variations for ads, banners, landing pages It is not great at copy yet, but we can always pick words that are better suited ✅Graphic designer They use the image generation tool as a base layer on which they can work on, it helps show product in different scenarios. Usually text is not completely generated it's mostly without text but visually rich ✅Media buyer cum Growth strategist 1/ Ad Iteration & Creative Planning After identifying ads that work/not work then start ask it to generate more variations or better versions of the ad from script to visual and more 2/ Data dump → insights Paste in campaign data, analyze what’s working, what’s not. Spot creative fatigue, ad fatigue, or audience drop-off. 3/ Ad Script Evaluation We input full video ad scripts and ask GPT to: → Identify where the flow slows down → Point out weak transitions or unclear messages → Suggest tighter hook + CTA combos → Check for emotional resonance and clarity This turns GPT into a script editor with brutal honesty. 4/ Audience & Growth Strategy Mapping We brainstorm growth questions like: → “What micro-niches haven’t we explored yet?” → “Where else can we reach this audience — influencers? podcasts? creator collabs?” → “How would this brand scale in the US vs. India?” Helps us zoom out and plan next-stage growth from a fresh perspective. Founder Deep Market Research 1/ If you're entering a new category, GPT helps: → Analyze the competitive landscape → Spot whitespace in consumer pain points → Find pricing models, product bundles, or positioning that works → Study what Reddit, Amazon reviews, and blogs are saying You can even ask it to “Act like a brand consultant and help me validate this product idea.” 2/ Storytelling & Positioning Strategy GPT helps us: → Turn founder vision into narrative → Test 10 different taglines → Explain your product like you would to a 5-year-old, or a VC → Convert product features into benefits people care about 3/ Internal enablement (Super underrated) (Currently not doing this) → Summarize internal brainstorming notes into briefs → Write SOPs from scratch → Draft onboarding docs for new team members → Convert Slack rants into Google Doc clarity Bonus: More use cases ✳️ Weekly Brand Insights Memos Drop in survey data, CX chats, or product reviews. GPT can turn it into a digest for your team: what’s working, what’s breaking, what needs fixing. ✳️ AI-Assisted Influencer Briefs Paste in influencer bios → GPT helps tailor custom brand briefs that align with their tone/audience. Would love to learn from other folks on how they are using it?

Just like UPI, govt should build UCI - United Checkout Interface for ecommerce stores This will auto-populate checkout data, remove COD option for risky customers and reduce friction to buy Conversion rate 🚀🚀

Update - Crossed 10L yesterday for the brand at 5.2+ ROAS

Crossed the first 2L/day ever for this brand, it's always a good feeling For a brand that launched just 5 months ago, this is a🤌🏻 The secret -> Started with just 2 products, and avoided pushing all products to ads. First build learning in two then expand this is way faster -> No iterations on ads. For new brands, they shouldn't iterate initially, goal must be to test as wide a variety of ads one can test. We tested over 19 angles per creative, and still testing -> Creatives focussed on lifestyle scenarios of where all products can be useful and how it saves time. Instead of detailing about product features, mentioned what does it do for customer