Our products · Marketing · Ad creative and testing
HookVerdict
Your next 30 Meta ads, written and tested in the languages your customers speak.
An AI ad creative and testing engine for Indian D2C brands. It drafts a brand brief from your website, writes ad variants with a clear hypothesis each, runs tests inside budget limits you set, and judges winners on real unit economics.
Live at hookverdict.com · free founding access
Who it is for
The buyer and the problem
Founders and marketers at Indian D2C brands spending ₹1–20 lakh a month on Meta and Google ads.
- Creative fatigue: a handful of near-identical ads a month, mostly in English only.
- Tests run without a hypothesis, so nobody learns why an ad won.
- Platform ROAS ignores cash on delivery, returns and GST, so "winning" ads can lose money.
How it works
HookVerdict, step by step
- 01
Brief
Paste your website URL and HookVerdict drafts a brand brief for you to edit.
- 02
Research
Optionally analyse competitors' ads to find hooks nobody is using.
- 03
Generate
Ad variants in five formats and ten languages, each with the hypothesis it tests.
- 04
Approve
Nothing goes live until a person approves it in the review screen.
- 05
Test
Tests run within the monthly cap, per-test limit and kill switch you set.
- 06
Verdict
Each ad is judged winner, loser, weak hook or still learning, on unit economics, and summarised weekly.
Features
What it does
- Five formats: Meta copy, UGC scripts, carousels, static images, Google search ads
- Ten languages for Indian audiences
- Templates in 1:1, 4:5 and 9:16
- Ad-policy and ASCI compliance checker
- Budget guardrails: monthly cap, per-test limit, kill switch
- Break-even ROAS that accounts for COD, returns and GST
- Free tools: break-even ROAS calculator, fatigue checker, swipe file
Integrations
Works with
- Meta Marketing API (read)In pilot
- Google Ads (copy)Available
- ShopifyIn pilot
- ShiprocketIn pilot
- RazorpayIn pilot
Under the hood
Where AI is used, and where it is not
The model
Model chosen per task by evaluation: overnight English batches run on the open-source Qwen model on Surge's server; Indian-language copy uses a foundation model because it scored clearly better on our tests.
Rules and code
Test decisions, unit economics, budget guardrails and compliance checks are deterministic code.
People in the loop
Every ad is approved by a person before it can go live.
Use cases
How businesses use HookVerdict
Regional-language ads
A skincare brand turns its product page into Hindi and Tamil Meta ads in one sitting.
Finding new hooks
Competitor analysis shows which angles are overused and which are open.
Profit, not ROAS
A "3x ROAS" ad is checked against returns and COD costs before more budget goes in.
Weekly creative pipeline
Thirty fresh, hypothesis-led variants a week from overnight batches.
What it shows about Surge
The capability behind the product
Model choice by evaluation
We measure each task and pick the model that passes, open-source by default. This is the stance we offer clients.
Generative AI with guardrails
Compliance checks, budget limits and human approval before anything is published.
Marketing and content AI
The same skills behind our AEM content and campaign accelerator.
Related accelerators: Content & campaign agent for CMS (AEM) · Catalogue enrichment + shopping assistant
Book a working session
A working session with our architects on one of your use cases. You bring a process and sample documents; we leave you with a scoped plan: model, hosting, integrations, timeline and cost range.