Week 5 - Google Ads Round 2: One Small Change That Already Boosted Tool Usage

Ads Are Live Again — And Early Signals Look Promising

We restarted the Google Ads campaign with the same conservative budget and slightly sharper copy. The real question: would the recent site changes actually move user behavior?

Just a few hours in (campaign kicked off around midnight), the first Analytics numbers already told a clear story.

The Win: Removing One Extra Click

Previously, users had to click a prominent "Try it out" button on the landing page before reaching the actual grammar tool.

That extra step — which felt minor to me — was quietly killing conversions.

Now the tool loads directly (or much closer to it). Result:

• Significantly more people are pasting text and running grammar checks
• First-time engagement jumped noticeably

One tiny friction removal → measurable uptick in core feature usage.

That’s classic funnel optimization: the fewer hurdles between ad click and value moment, the better.

The Problem: Nobody Is Clicking 'Learn'

By the way, this is exactly the same pattern we saw a week earlier in the initial Google Ads data—feel free to check out the detailed initial analysis: First Google Ads Results: What CHF 100 Taught Us About Early Traffic & User Behavior.

Here’s the bigger red flag so far:

Almost zero clicks on the "Learn" button — the one that leads to our interactive quiz and detailed explanations of each grammar rule.

This matters because the whole philosophy behind AI Grammar Mentor is learning from mistakes, not just blind auto-corrections.

If users reach the tool, see corrections, accept some… but never click to understand why something was wrong, we’re only delivering half the promise.

Current data (8-hour snapshot):

• Tool usage: up
• "Learn" button clicks: near zero

Why the Button Is Failing (My Best Guess)

A few quick self-reflections:

• The book icon + "Learn" label feels too academic / school-like
• Most adults associate school with obligation, boredom, or bad memories
• People want fast, low-effort improvement — not "homework"
• The button sits quietly in the UI instead of feeling exciting or rewarding

If it reminds me of hated school textbooks, it’s probably repelling others too.

Quick Experiments We’re Running Next

No big redesign yet — just targeted tweaks:

• Rename to something action-oriented: "Start Quiz", "See Why", "Master This Rule"
• Swap the boring book icon for something more dynamic (lightbulb? brain spark? question mark with arrow?)
• Test placement / contrast so it stands out without being pushy
• Add subtle social proof or teaser text: "Most users improve fastest after 2–3 quizzes"

Goal isn’t to judge the quiz quality yet — it’s to get people to see the quiz at all.

If the new trigger works → more quiz views → we can then measure quiz completion, repeat visits, time spent, etc.

If people click but still bounce from the quiz page → then the problem shifts to quiz UX, not the entry point.

Step-by-step iteration.

Analytics Blind Spot

One more embarrassing oversight: not every page has Google Analytics event tags yet.

That means I’m still partially blind to movement across blog posts, pricing, or other non-tool areas.

Fixing that today so future data tells the full story.

The Bigger Picture

Every change — even the small ones — is about closing the gap between "someone clicked the ad" and "someone came back tomorrow because they got better at writing."

We’re still very early, but the pattern is already visible: reduce friction → more usage → more chances to show real value → higher likelihood of retention (and eventually payment).

If you’ve ever wanted grammar help that doesn’t just fix your text but actually helps you stop making the same mistakes — without feeling like school — that’s what we’re building.

Paste any sentence or paragraph right now. Get corrections + explanations that actually make sense. No extra clicks, no paywall ambush.

Try it yourself and see if the difference clicks for you.

The chat bubble is live if anything’s unclear — drop me a line.