Twelve things I've learned running paid media for founders across India, the US, the UK, France, UAE and Canada. What I'd change if I started over.
Performance marketing rewards patience the same way it punishes panic. Most of what I've learned came from clients spending money on the wrong things and us spending months figuring out which thing was the wrong thing. These twelve lessons are the ones I'd put in a frame for any founder considering a paid budget.
In the classroomMost accounts I audit are leaking 30 to 60% of attribution because the foundation is broken. GTM, GA4, server-side, CAPI for Meta, enhanced conversions for Google. Fix the plumbing first. You can't optimise what you can't measure.
People say "Meta is dead" or "Google is too expensive". Almost always, the issue is creative, not platform. The same campaign on the same platform with better hooks regularly outperforms by 3x. Channels are commoditised. Creative is the moat.
If you change the creative, the audience, and the landing page in the same week, you've learned nothing. Test one variable at a time. Run for at least seven days. Decide based on results, not vibes.
A great ad on a bad landing page is a refund waiting to happen. Most marketers don't own the landing page, they own the ad. That's why so much paid media underperforms. The same headline should appear on the ad and on the landing page within 3 seconds.
If you optimise only for ROAS, you'll only ever scale your retargeting. Real growth comes from acquiring new audiences, which always shows a worse ROAS in the short term. Build a portfolio: 60% prospecting, 30% retargeting, 10% testing.
Yes, you'll rank organically for your brand name. Yes, your competitors are bidding on it. Yes, that 80-cent click protects 15 organic clicks. Run branded search. Always.
Whatever a client signs up for, the first 30 days are spent learning. Pixel data is thin, creative is in test mode, audiences are warming up. Anyone who promises performance in week one is either lying or sandbagging the early creative.
It incentivises agencies to spend more, not better. The agencies I trust most charge a flat monthly fee. We do the same. Our fee is the same whether you spend $5K or $50K. You scale only when it makes sense.
Strategy gets the credit. Operations does the work. Refreshing creative weekly, killing tired audiences, fixing tracking when iOS breaks, replicating winning campaigns into new geographies. The agencies that quietly do operations well are the ones that show up the best on the dashboard.
If your organic posts get 12 likes and your reviews are mixed, paid traffic will land on a vibe that doesn't convert. Fix the organic story first. Then add paid as fuel.
Stop-start paid media is more expensive than steady paid media. The algorithms don't learn in three weeks. Commit to a budget you can sustain. Better to spend $5K a month for six months than $20K for one and then nothing.
Every paid dollar you spend should feed a list. Email, SMS, retargeting pool, follower base. If a campaign doesn't grow an owned audience, you'll need to spend that dollar again. The brands that win in paid media long-term are the ones that compound their owned audiences over time.
Fix tracking first. Treat creative as the channel. Test one variable at a time. Own the landing page. Don't optimise only for ROAS. Run branded search. Treat the first 30 days as setup. Don't pay percentages of spend. Operations beats strategy on the dashboard. Don't run ads to a brand you haven't built. Pick a sustainable budget. Compound owned audiences.
If you'd like a hand running paid media this way, the performance marketing offering is built on these twelve principles.
Free 30-minute audit on a discovery call. We'll tell you what we'd change in week one.
Book a discovery call