Nielsen looked at ~500 campaigns: targeting drove 9% of the sales from advertising. The creative drove 47%. Perfect aim at a bad ad is still a bad ad. Make the ad worth catching. Then aim.
Your customers aren't NPCs. Targeting 18 to 65 is shouting at the city. Upload your buyers (Meta needs at least 100) and a lookalike finds the people most like them. Start from who already bought.
Grocery store test: 24 jams drew 60% of shoppers to stop, but only 3% of them bought. 6 jams: 40% stopped, 30% bought. More choices draw a crowd. Fewer choices make the sale.
Your ad account has a five-star wanted level. Meta's algorithm needs ~50 conversions a week per ad set to learn, and every big edit restarts the count. Fix the pixel, run one ad set, fresh ad, then hands off.
Deloitte tracked mobile sites of 37 brands: a 0.1-second speed-up lifted retail conversions 8.4% and travel conversions 10.1%. Before you buy more clicks, make the door faster.
Expedia had an optional "Company" field. Customers typed their bank there, then the bank's address, and the card check failed. Deleting that one box was worth $12M a year in profit. Every box is a reason to leave.
Researchers sent test leads to 2,241 companies. Average callback: 42 hours. 23% never called back. Calling within the hour was nearly 7x likelier to qualify the lead than an hour later. Ads bring the lead. Speed keeps it.
Same eight car washes. One loyalty card started empty, the other had two free stamps out of ten. 19% vs 34% finished. Don't hand people an empty card. Hand them a head start.
"Save the planet" got 37% of hotel guests to reuse their towel. "Most guests in this room reused theirs" got 49%. Tell people what people like them do.
A perfect 5.0 rating should sell more than a 4.5. It doesn't. Purchase likelihood peaks around four and a half stars and drops near five, because perfect looks fake.
He searched for the store by name and was standing at the door. The store still paid for the click. eBay turned those ads off and people just took the free link right below.
Ten thousand clicks, zero sales. You asked the ad for clicks, so it found the best clickers in town. Ask for purchases and it goes looking for people who buy.
Monday, 9 a.m., blank calendar?
B4KER: one brief → an editorial proposal (title, format, creative direction per post) → approve → images and videos created and on your calendar. Autopilot schedules them too.
https://t.co/JxR3YYoFiM
Dropbox paid in storage, not cash.
2009: search ads cost $233–388 per customer for a $99 product.
Then: invite a friend, you both get space. Sign-ups +60%, permanently. 2.8M invites a month by April 2010.
Pay users in what they already want.
Local businesses: what if your ad opened a chat instead of a website?
B4KER: goal "WhatsApp conversations" → the tap opens a WhatsApp chat. Reply freely for 24h; after that, approved templates to opted-in contacts only.
https://t.co/JxR3YYoFiM
People pay more for what they built.
Norton, Mochon & Ariely (2012): people who assembled an IKEA box bid 63% more than for the same box pre-built. Unfinished builds lost the effect.
SaaS onboarding: let users finish one small thing fast.