By Ashutosh Singh, Founder, Social Ashya
A little over two years ago, Social Ashya was a small team out of Lucknow doing what most young agencies do — a bit of content, a bit of social media management, a client here and there who needed “someone to handle Instagram.” Today, our paid media desk alone has helped clients generate more than $1 million in trackable revenue through Meta Ads and Google Ads.
I’m not writing this to brag. I’m writing it because when we started running performance campaigns seriously, I couldn’t find honest breakdowns of how agencies actually get there — everyone talks in vanity metrics, nobody talks in process. So this is the process.
The Problem We Kept Running Into
Every founder we onboarded had the same story: they’d run ads before, spent money, got some clicks, maybe a few leads — and then quietly stopped, unsure if it had “worked.” Nobody could tell them their actual cost per acquisition, let alone their return on ad spend by campaign.
That gap — between spending on ads and knowing what the ads actually returned — is where most budgets die. So before we touched a single ad account, we fixed measurement first. No campaign launched without proper conversion tracking, whether that meant the Meta Conversions API, Google Ads enhanced conversions, or a clean GA4 setup tied back to actual sales data. It sounds unglamorous, but it’s the single highest-leverage thing we do for a new client.
Meta Ads: Where We Built the Volume
Meta was where most of our early wins came from, and the approach that worked wasn’t complicated — it was disciplined.
Creative testing, relentlessly. We treat ad creative like a live experiment, not a one-and-done design task. Every campaign launches with multiple creative concepts — UGC-style videos, static carousels, founder-led testimonials — running against each other simultaneously. We kill underperformers within 48–72 hours based on hook rate and cost per result, not gut feeling.
Layered audience strategy. Rather than betting everything on one audience type, we typically run a mix of:
- Broad, interest-based audiences to feed Meta’s algorithm signal
- Lookalike audiences built from actual purchasers, not just page engagers
- Retargeting segments split by funnel stage — website visitors, add-to-cart, past customers
Catalog and dynamic ads for e-commerce clients. For clients with product catalogs, dynamic retargeting consistently outperformed static remarketing, often by a wide margin, because the ad automatically shows the exact product someone browsed.
Across our e-commerce and D2C accounts, this approach has taken average ROAS from roughly [insert your real baseline] to a sustained [insert your real current average — many of our accounts sit in the 3.5x–5x range].
Google Ads: Where We Captured Intent
If Meta is about creating demand, Google is about capturing it — people who are already searching for a solution. Our approach here is more surgical.
Search campaigns built around buyer intent, not just volume. We segment keywords by funnel stage — someone searching “best [product] for [use case]” is closer to buying than someone searching “[category] tips.” Budgets follow intent, not impressions.
Performance Max, used deliberately. PMax can be a black box if you let it run unchecked. We feed it clean first-party audience signals and high-quality creative assets, then monitor placement and asset-level reports closely rather than “setting and forgetting.”
Remarketing lists for search ads (RLSA). Bidding more aggressively on searchers who’ve already visited the site consistently lowers cost per acquisition, because you’re not paying full price to re-earn attention you already had.
The combination of tight negative keyword management and intent-based segmentation typically cut client cost-per-lead by [insert your real % — commonly 30–50% within the first 60–90 days for our accounts].
Why Running Both Together Mattered More Than Either Alone
The real unlock wasn’t Meta or Google individually — it was making them work as one funnel instead of two disconnected ad accounts.
Meta built awareness and warmed up cold audiences with content that felt native to the platform. Google then caught the demand that awareness created — the person who saw a Meta ad three days ago and is now Googling the brand name or the product category directly. When we started reporting on this combined journey instead of siloed platform numbers, clients finally understood why “top of funnel” spend mattered even when it wasn’t converting on day one.
What Actually Moved the Needle
If I had to compress two years of testing into the handful of things that mattered most, it’s this:
- Tracking before scaling. No amount of ad spend fixes broken measurement.
- Creative volume, not creative perfection. More tested variations beat one “perfect” ad every time.
- Funnel-stage budgeting. Treating cold, warm, and hot audiences as different jobs with different KPIs.
- Weekly, not monthly, optimization cycles. Ad platforms change fast; a monthly check-in is too slow to catch what’s working.
- Reporting in revenue, not reach. Clients don’t spend money to get impressions — every report we send ties back to leads, sales, or revenue.
What’s Next
Crossing $1 million in generated revenue isn’t a finish line for us — it’s proof the system works, and now it’s about doing it faster and for more brands, with the same discipline around tracking and testing that got us here in the first place.
If you’re a founder who’s tried ads before and walked away without clear answers on what worked, that’s usually a measurement problem before it’s a strategy problem. Fix that first, and the rest gets a lot easier to figure out.
Ashutosh Singh is the founder of Social Ashya, a branding, content, and performance marketing agency based in Lucknow, India.