Executive Summary
Goal of 6.25% net revenue by H1FY28 compared to baseline of 2.75% net revenue in H2FY25, out of which 5.6% net revenue is targeted from ads (seller + brands) 3Y Vision: To build a fair auction-based advertising marketplace that allows sellers to drive sustainable growth with maximum efficiency and simplicity while delivering highly relevant shopping experience for customers
To drive ads from 2.7% baseline to 5.6%, we have to strengthen the core ads flywheel by working on the below themes: Fair Auction-Based Marketplace for high competition: Ranking purely based on relevance and bid value, ensuring a strong correlation between bid and scale. Clear and actionable feedback to sellers, helping them optimize their bids for better scale and increasing competition on platform Drive Sustainable Growth: Ads will be the primary driver for growing new and high margin & high converting catalogs that are currently not supported by organic ranker (high-quality, high-priced products, combos, and niche categories). Ads will help to drive overall orders (Ads + organic), ad views will remain incremental rather than cannibalizing organic views Maximum Efficiency (Overall ROI): Ads ObyV ratio to improve to 70% by serving views in the highest converting contexts via budget pacing, real time bidding and differential slotting Brands Ads targeted to grow to 7.5% Spend/NMV from 5.4% today, as we solve the scale problem for brands via ads products. Awareness products (both static & video ads) will drive the top of the funnel discovery while keyword bidding & ranking changes in CLA will drive higher market share for brands across all REs Seller services targeted to grow from 0.1% to 0.4%. The Seller Insights Dashboard will evolve into a business intelligence platform, and initiatives like AI-led catalog enrichment, reviews & ratings, and partial refunds for return cancellations will address seller pain points. User monetisation targeted to grow to 0.32%, from quality assurance, faster delivery and insurance offerings to our consumers
Goals for next 3 Years across top metrics
Key Milestones/Phasing
* Most of the seller ad opportunities have been identified in the last 2-3 months and part of H2 of ads , so not tagging it as Net new. Also haven’t done phasing since it is H2 in nature
Details of 10X/Whitespace Opportunities
Seller Ads
Meesho Ads Marketplace Vision:
To build a fair auction-based advertising marketplace that allows sellers to drive sustainable growth with maximum efficiency and simplicity while delivering highly relevant shopping experience for customers
Key principles of the marketplace in line with the vision
https://docs.google.com/presentation/d/1IqUmwmhmOMiPLZLHrIZQFWrgQFlmW6_6kJu82o4Ae1w/edit#slide=id.p
Details of principles / key initiatives
1. Fair auction-based marketplace with high Competition dynamics (Opportunity: 70-100 bps)
Bid / ROI is input by seller unless seller actively chooses Auto Bidding
Bidding options provided by different platforms (Meta, Amazon, Google) are detailed . Basis these benchmarks, we shall have options for CPC bidding and ROI bidding (excluding unscaled catalogs). Sellers will be able to choose Auto Bidding but it will not be on by default (as is done currently). They will be able to choose bids within a range rather than having to choose from 3 options (as is available currently). No assurance of scale or ROI - ranking purely based on relevance and bid; with high correlation between bid and scale (Opportunity: 60-80 bps)
Currently, ROI Assurance campaigns enjoy unfair advantage over other campaign types - Their ranking is at CPC as suggested by auto CPC algorithm but charging is at a discounted CPC to ensure promised ROI (ensures them scale but comes at the cost of platform revenue) Floor CPC is set low to deliver on promised ROI with sellers receiving refunds as compensation if target ROI is not met Going forward, Target ROI campaigns will align ranking CPC with charging CPC to eliminate undue scaling advantages. Floor CPC will be set based on competition and expected ROI at sscat level, with no refunds if targets are not met. If ROI falls significantly due to floor CPC caps, sellers will receive feedback, and campaigns will be scaled down (validated from ). High sensitivity to scale and bid: Achieve greater than 50% importance to Rev/view in ranking
to ensure high sensitivity of bid and scale. Differential slotting algorithm will work on top of this layer, to ensure the ad catalogs are served only when they are within a certain acceptable range of pcvr in comparison to the organic catalogs. The key constraint in achieving min 50% importance is O/U guardrail which we’ll be able to satisfy due to much tighter relevance.
Clear actionable feedback to seller on bid options that can help attain higher scale (Opportunity: 10-20 bps)
Currently, only ~10% of fully budget utilized campaigns increase their budget within 30 days. Going forward, if a campaign utilizes its budget and meets target ROI, feedback will highlight “Limited by budget”, showing potential incremental views and sales at a higher budget. If a campaign fails to utilize its budget due to high ROI or low CPC, feedback will indicate “Limited by high ROI / low CPC”, suggesting bid adjustments for increased visibility and sales. Supporting and are available. 2. Drive sustainable growth for all use cases (High Margin, New catalogs , Low ASP) (Opportunity: 120-150 bps)
Ads should be primary lever for growing new catalogs (Opportunity: 40-50bps)
Key problems in new catalogs’ ads flywheel (): Poor ads conversion - Low obyv ratio (~20% vs 45% plat average) due to overexposure in irrelevant contexts. Need to distribute views in higher converting contexts Organic order degrowth - majority views come via ads & explore which have a lower obyv vs exploit, leading to degradation of organic profile. Tracked separately as part of Need for ads - Limited seller need as 10% views are currently reserved for explore, plus there are inefficiencies leading to 40% explore orders going to duplicate catalogs. Need to ensure that explore only helps build early profile of new catalogs and ads help fast track them to 80 OC Ads should be the primary lever for flywheels other than price (we will target these use cases - higher priced, better quality/rating, combos, niche categories) (Opportunity: 40-60bps)
All these usecases have higher margins (validated by seller LODs) but lower organic scale vs platform. These products also have a significantly lower ObyV ratio (high priced catalogs: 39% vs 44% platform average), leading to poor ROI and low adoption. Our goal is to improve conversion ratio by serving these products in high-converting contexts potentially via a native ads ranker and kicking in the flywheel in ads. As next steps, we will identify CPDs (e.g., high-priced, better quality duplicates with low scale) and boost their discovery, incrementally improving the ads ranker Ads should drive growth in organic orders while maintaining user relevance. Removing de-biasing and profile degradation (Opportunity: 40-50bps) For unscaled & low OC catalogs, ads boost UGC, ratings, and conversion rates, driving organic growth. However, for top 50 OC catalogs, ads currently cause a ~15% drop in organic views and orders due to a weaker platform level conversion features in ranker score. Corrections are in progress across REs, aiming for a 10-15% organic order increase when ads are run. Early search experiments show a +5.3% increase per catalog:
Aside from splitting platform level features in organic and ads, we can look to implement sales velocity tracking (used by Amazon) to help organic rankers detect trends & seasonality faster, improving organic order momentum when ads are first run. Platforms like PDD have also resorted to direct boost in organic visibility based on ads spend and capping on organic orders to make ads a necessity (source - expert call, details ). 3. Maximum efficiency (Improve Obyv ratio to > 70% and overall ROI for sellers) (Opportunity: 100 - 120 bps)
Primary goal here is to achieve ObyV ratio of 70% which enables much higher revenue at lesser views (higher CTR, higher ObyC enabling higher CPC therefore higher overall rev/view). This can be achieved by -
Improving good catalog penetration / adoption via ROI bidding (explained in fair auction) Real time ad slotting based on tighter relevance guardrails for serving ads Campaign level budget pacing via real time bidding (CPC values linked to conversion probability of a context) or relevance threshold based throttling of views Improving true order incrementality for sellers when run on ads by targeting a 10-15% increase in organic orders across OC buckets when ads is run Identifying niche high converting contexts for a catalog to improve ObyV ratio e.g. retargeting
Increasing Relevance of Ads and Real time Ads slotting based on relevance: (Opportunity: 70-80 bps)
Currently, ~52% of the ad views are getting served in much lower relevance, with correspondingly lower ObyV ratios. In essence we can achieve the same revenue via much lesser views keeping a higher relevance bar. Currently relevance thresholds are at a global level reducing their effectiveness and initial results of deploying tighter relevance thresholds at SSCAT level have shown (KR 1.2).
In the end state via real time slotting - ads will be served only when delta in predicted conversion between organic and ad catalog at every individual slot level is within a threshold after accounting for revenue /view from that slot
2. Campaign level budget pacing over the day to serve ads in highest converting contexts (Opportunity: 20-30 bps)
Ad rev % is at 5%+ at the start of the day and at under 1% by end of the day. This is because we currently optimize for utilizing available budget as soon as possible while being within ObyU guardrail. This leads to lower conversions and ROI for sellers.
Going forward, we will pace budget over the day to utilize it in most converting contexts Budget pacing will be implemented in a different manner for different campaign types: No bid/ROI campaigns: Real-Time Bidding will limit visibility in poor-converting contexts based on CPC-linked conversion probability. For Fixed CPC campaigns, we can do this in either of the following ways: CPC based Pacing: Adjust CPC dynamically by ±x% (outside in view of Amazon) with seller consent to optimize spend Relevance throttling: Control budget pacing by tightening/relaxing relevance thresholds for catalog visibility Budget pacing and Real Time Bidding can happen only if we can calibrate our models and make pctr & pcvr more accurate
3. Ad views should purely be incremental to organic view (Opportunity: 10 - 15 bps)
Currently 13% of organic views are either cannibalized by ads (6%, same page dedup) or are repeat views (7%, no cross page dedup) leading a to 10% drop in organic orders when a catalog is ad-active. These repeat views come at an extremely poor ObyV ratio as the catalog has already been viewed once on the higher ad slot. Ad view to be shown only if difference between organic slot and ads slot is above a certain threshold
4. Identify high converting contexts e.g. Retargeting to improve ObyV ratio (Opportunity: 10bps)
We need to improve our ability to identify high converting contexts. Ex - resurfacing wishlisted catalogs in Search and Reco feed leads to ~3x higher ObyV v/s overall feed of same users, however are surfaced to only ~20% of the users in relevant contexts
In order to achieve the above capabilities we will need to build a few capabilities - Current ad tech system design and revamped design for the same are articulated . Brand Ads
Meesho Brands Ads Vision:
To build a fair auction-based advertising marketplace that allows Brands to drive sustainable growth and achieve scale with maximum efficiency and simplicity while delivering highly relevant shopping experience for customers.
Key Principles from Meesho Ads Marketplace Vision remain same
From our CPD discussions we have learnt, Brands are more interested in getting higher scale of orders than only optimising for ROI (”We would prefer getting 2L in GMV with a ROI of 3 compared to getting 1L GMV with a 5 ROI”, “ROI is not a key metric that we track, we are more interested in getting more users to land on our product pages since our long term brand strategy is to acquire a large user base which will be loyal to the brand and will have my brand on top of their mind”)
We spoke with the Anchor & Focus brands across PCW and WMF namely Himalaya, Mamaearth, Joy, P&G, Lotus Herbals, Swiss Beauty, Wottagirl, Blue Heaven, Paree, Baidyanath, Bata, Relaxo
https://docs.google.com/presentation/d/1SxYO_jmJgXLLYyxzGdG02-8pwvxQFAkTymWV__7OLnk/edit#slide=id.g33085d0b126_0_0
Increase Brands Spends/NMV from 5.4% today to 7.5% in Mar’28
Outside in benchmarks Ads spend/NMV
Tmall: ~6%
Amazon India: ~7.5%
Drive scale for brands via targeting capabilities across formats & real estates
As we are onboarding anchor & focus brands, we believe providing different ads capabilities in terms of keyword bidding and targeting becomes supremely valuable.
PCW and WMF brands prioritised increasing order scale over ROI using CLA Brands seek more control over the current CLA product to help them invest in targeted use cases: specific contexts (competitor brands, category, real estate etc.) While replicating all the improvements discussed in CLA for brands, we are facilitating targeted ads also
We will enable these capabilities across all Real Estates and ad formats, to enable our big anchor & focus brands to drive scale using the suite of products available to them for different use cases
Brands have different use cases to target the users
Target new to brand, new to SSCAT, new to mall users Re-target existing users who have not purchased but interacted with the brand (click, views) Target users when they are considering a competitor product Target users when they are considering buying their category Different ads products like video ads, display ads & cla ads are used by brands to target consumers across the user journey and drive maximum scale for their products.
User First Perspective
“We know 70% of the views on amazon is driven by search, we simply bid the highest on certain keywords where we want to scale the orders, roi is lower as we expand to relatively broader keywords, but we want to maintain a certain share of voice, ROI is the last thing I check, after looking at CTRs, conversions etc”- Himalaya
“Capturing the share of voice is very important to us, we want to capture a large share of voice in our category of products, wherever there is a shortfall we want ads to cover for that”- P&G
“we want to drive scale of non-baby care categories, hence acquiring new users in these sub categories is paramount importance to us, we use a mix of video ads and banner ads to drive awareness of our brand and products on amazon, but its missing today on Meesho”- Himalaya
“We have Shahrukh Khan as our brand ambassador, and we want to use his brand appeal to build our brand using the video ads and display ads to reach users in tier 2+ cities” Vi-John
Outside in Perspective
Blinkit Keyword bidding + Bid Booster
Amazon keyword bidding guidelines
Screenshot 2024-09-12 at 12.14.53 AM.png
Catalog Video Ads Display Video Ads Search Feed Widget Ads
. Screenshot_2024-09-11-13-16-25-21_fd1e8ef594b195c55a3bba4818d0ce35.jpg
Video Ads in PDP Reco Display Ads (Sponsored Brand) in PDP Reco
Amazon provides sponsored brands functionality, where brands can re-direct traffic to brand store while running an ad on a particular SKU, through this 40% of brands order actually happen on non-ads catalogs as well, providing huge incentive to brands to continue spending on these ads.
retargeting-ads-on-amazon.jpg
Rewards and Coupons
Screenshot_2024-09-11-11-00-25-70_fd1e8ef594b195c55a3bba4818d0ce35.jpg
Coupons distribution for new users acquisition:
Products to be built
Keyword Bidding in Meesho:
We will take the following inputs from the brands - keyword and the bid value In the backend, we are envisioning a system where the keyword is not restricted only to search, but across all REs wherever the keyword is relevant Eg, if a brand is bidding on “facewash”, we will rank catalog with higher CPC in all contexts related to “facewash” The reco feed where the parent catalog is facewash If the search query has facewash If the CLP collection is created for facewash sscat etc. New Ads formats- video ads & banner/widget ads
To enable effective brand awareness for users on the platform, we will invest in building more awareness constructs, including video based ads
Seller Services
Seller Services Vision
To develop a comprehensive suite of seller services and programs that enable sellers to scale efficiently while optimizing their costs, ultimately driving sustainable growth on Meesho’s platform
Key Challenges & Proposed Solutions Aligned with Focus Areas
Opportunity Size
Details of key initiatives
Ideas Under Exploration
A few potential 10x ideas(B2B Market place, Virtual Design Lab, etc) are still under exploration.
User Monetisation
Opportunity: Add 32 bps of revenue via Quality Assurance, Faster Delivery to our consumers
Mapping the consumer needs with the core meesho ecosystem, we have identified 2 big opportunities for user Monetisation
Quality Assurance- solves for increasing trust regarding product quality (Opp Size: 16 bps) Faster Delivery- drives more value to users in terms of lower delivery times (Opp Size: 10 bps) We believe the above 2 opportunities will help us core meesho experience problems for our users and also provide us an opportunity to drive incremental revenues.
Apart from the above 2, we will explore other opportunities like sachet insurance and OTT subscriptions as well.
User first perspective
Quality Assurance:
The WhatsApp pilot with QA (priced at INR 20 covering fading and shrinkage for apparels with refund of 60%) has validated a strong user pull 0.6% adoption (transactors as % of message viewers), and claim rate of ~1.5% “Color Fading aur Shrinkage Protection Plan jo aapne introduce kiya hai ye kaafi sahi plan laga mujhko..kyuki Meesho se humesha ye complain rahi hai ki kapde ki quality aachi nhi hoti hai aur jab kapde ko wash karte hai toh uska color bhi fade ho jata hai”
Faster Delivery: Users have need for faster delivery for various use cases and they are willing to pay Rs20-30 for the particular order
" Maine summers ke liye kapde order kare hai apne liye aur app par bahut time dikha raha tha, isliye maine pay kara taki mujhe jaldi mil jaye”
" Eid ane vali hai aur mujhe ghar jana hai , 20 pay karke agar mujhe jaldi mil jayega to meri tension kam ho jayegi ki ayega time se ki nahi, isliye maine pay kar diya”
Sachet Insurance: High insurance premiums is the key problem, we believe offering them sachet insurances in the price range of Rs.20-30 will help drive adoption
“Kids education vala insurance kafi acha hai, yeh market me kahin dikhta nahi hai, Mujhe agar yeh Meesho par dikhe to main apne husband se bat karke unhe bolungi khareedne ko”
Experiment Results: We did an experiment to offer PA & VBD insurance at RS.29 with 1 year coverage, landing to pay stands at 0.24% which is better than the industry benchmark of ~0.1%
Outside in perspective
Quality Assurance: Shopee- premium 3% of cart value, 9% adoption rate
Faster Delivery: users can also opt for faster shipping methods (e.g., Shopee Xpress) at an additional cost
Sachet Insurance:
PhonePe’s total business (calculated as total policies sold * average premium) has grown from 3Cr in Q2’2020 to 176 Cr in Q2’2024
Opportunities Proposed for Inorganic Growth
Gen AI-led catalog enrichment
Problem: Lack of high quality listing/ merch images on Meesho
Idea: A Generative AI tool integrated into the catalog upload journey, allowing sellers to enhance images instantly. Eg: Increasing Resolution, Model images generation for fashion
1. Challenges that might arise in solving this problem statement organically (Time to launch MVP, time to build new capability etc):
Lack of inhouse expertise Time to build and validate the MVP 2. Why inorganic method will be suitable for this problem statement (Shortening time to market, external capabilities, new category/area etc)
A lot of external players already exists with whom we can quickly partner and try this out 3. If there is some level of conviction, describe the type of solution needed (M&A, strategic investment or partnership)
We want a partnership with the external players who are providing the Gen AI led catalog enrichments. We can integrate their capabilities within the existing catalog upload process to test out the pull and UE for this offering