Every post on Performance Marketing — notes from running campaigns, not theory.
27 posts
Dashboard metrics in Google Ads represent platform opinion rather than actual profit. Minh explains why business data dashboards and full-funnel thinking must guide optimizations.
Scaling Meta Ads during peak sale days easily fails if you only look at Blended ROAS. Minh explains why intraday scaling must track New Customer ROAS.
Chasing cheap leads with broad discounts often attracts churn. Minh explains why spending more to isolate competitor audiences yields far higher lifetime value.
Weekly marketing reports often go unread. Minh outlines the four boundaries required to let AI agents monitor anomalies instead of cluttering your dashboards.
Stacking ad groups inside a single Demand Gen campaign causes Google to dump budget into cheap audiences. Minh explains why splitting campaigns is essential to scale.
Should you launch Black Friday sales on November 1st or wait? Minh breaks down why the decision hinges on your customer acquisition mix and catalog depth.
Meta removed hard gender targeting at the ad set level. Here is how a 90% bid decrease through value rules restores control, and who actually needs it.
Should a premium agency add a cheap offer to fill the funnel? Minh explains why low-ticket only works with an audience, and staying high-ticket wins before that.
Meta One sells organic reach, search ranking and human support. Minh separates the fairness question from the budget question, and explains why Advanced is the tier worth buying.
Nick Theriot rates GPT-6's ad copy at 80%. The bigger win sits in the coordination layer, and Minh explains why that is where the hours actually go.
Many Facebook Ads accounts do not fail from too few tests but from creative that never shows proof. Minh breaks down a case where CPA dropped 75% after switching to video.
When cost per purchase stops falling, Minh explains why the ceiling sits in the business model, not in the ad account, and what to raise instead.
Meta optimizes for the event you declare, not the outcome you want. Minh breaks down how to redefine what counts as a lead and what an order is worth before touching targeting.
Turning off Meta's creative enhancements once does not keep them off. Minh breaks down why the setting resets and what to re-check before the platform edits the brand.
An advertiser who spent over 300 million dollars on Meta Ads still cannot predict a winner. Minh argues the real edge is a kill threshold set before launch, not intuition.
Google Ads is more than Search and Shopping. Learn how to test creatives using CPA curves, decay rate, and audience portability.
Meta now separates click-through, engage-through, and view-through attribution. Here is how to read performance correctly when reported conversions drop.
When scaling Google Ads past $10k, CPC spikes while conversions flatline. Solve this scaling trap using Vertical and Horizontal strategy.
Threads ads are now global with new formats like carousel, catalog, and app ads. A strong moment for early testing.
Relying on your website, blending in, and over-narrow targeting create sales friction. Minh shares a 4-pillar framework to fix it.
Search Four 2026 in practice: stronger data, value-based bidding, AI-powered reach, and flexible budgets to capture high-value demand.
When lead volume rises but close rate stays low, the issue is often optimization quality—not budget. This article shows how Minh fixes conversion signals and lead quality before scaling ad spend.
Smart Bidding is only as powerful as the conversion data behind it — lessons from running Google Ads at Kyna English, ELSA Speak, and 51Talk.
The hide-sponsored-results feature is less about removing ads and more about forcing marketers to improve relevance and user experience quality.
The 95/5 rule shows that only 5% of B2B buyers are in-market at any time. This post explains how to balance demand capture and demand creation for both short-term and long-term growth.
If you set a target and Google “misses”, it’s usually not a bug. Targets primarily control bidding aggressiveness and how much room the system has to explore and scale.
In 2026 creative testing, Minh favors quality first, quantity second. A practical way to avoid spray-and-pray testing and scale with intention.