Thoughts on marketing automation
From ad measurement to multilingual site setup, what we learned building our own system.
AI Brand Voice: Why AI Text Sounds Robotic and How to Fix It
AI text sounds robotic mostly because of a vague brief. This guide shows how to define your brand voice with five concrete criteria and where a human editor fits.
Seasonal campaign timing: a three-week countdown plan
Start a seasonal campaign three weeks before sale day. The weekly plan, last order date and ad launch timing are here.
Reading Competitor Ads in the Meta Ad Library: Four Signals
The Meta Ad Library shows what a competitor is running, not what it spends. This guide covers four signals worth reading and where each one stops being reliable.
WhatsApp Response Time: How Fast Should You Reply to Ad Leads?
The speed of your first reply to a WhatsApp ad message is the first gate to a sale. Here is how Meta's 24-hour window and an automated first reply keep it fast.
Attribution Window: 1-Day vs 7-Day Explained
An attribution window is the period in which a conversion can still be credited to an ad. This guide compares 1-day and 7-day windows and explains why dashboards disagree.
High ROAS, Low Profit: Why It Happens and How to Fix It
ROAS divides conversion value by ad spend and nothing else, so a high ROAS can sit on top of a thin profit. Here is how break-even ROAS shows the real picture.
When Google indexes the wrong language, check these three settings
If your brand search shows the English version, three settings are the usual cause. Diagnosis and fix steps are here.
Ad funnel basics: why measuring an appointment business with an e-commerce funnel fixes the wrong thing
An ad funnel differs by business. Measure an appointment business with an e-commerce funnel and you fix the wrong thing.
Your conversion count may be tripled: the alias event trap
Meta returns the same conversion under several names. Sum them in your reporting and one sale becomes three, with revenue inflated to match. An anatomy of a bug that survives because it is silent.
Editing a campaign during the learning phase restarts the test
A new ad set's early cost is not its real cost. Budget or targeting changes made in this window reset the counter and make you pay for the expensive period twice.
Meta's click count makes your ads look better than they are
The clicks column counts profile visits, likes and comments alongside actual visits to your site. The gap between the two is wide enough to point your diagnosis at the wrong step entirely.