How Learnard tightened paid search before back-to-school and improved premium signup CPA with myr

At a glance

Overview

Learnard helps students and professionals turn source material into an interactive study workflow: upload PDFs, lectures, or links and get structured notes, flashcards, quizzes, fill-in-the-blanks, AI podcasts, and cited chat.

The product fits a clear use case: exam prep, lecture capture, and research digestion. Google Ads should capture people actively looking for study tools, not everyone curious about "AI homework."

Learnard runs paid search year-round with spikes before exam periods and back-to-school. The challenge is classic for edtech B2C: broad keywords attract cheating-adjacent queries, free-tool hunters, and clicks that never upgrade to Premium.

They needed monitoring that understood seasonality and signup quality, not just daily spend.

The business

Learnard serves:

• University and high school students preparing for exams
• Professionals digesting reports and meeting recordings
• Teachers creating quizzes and study guides from curriculum materials
• Researchers organizing papers and source libraries

Revenue comes from free tier → Premium (higher limits, podcast features, chime-in, etc.). Google Ads must bring users who upload materials and hit paywalls with real value, not one-session tire-kickers.

Target queries include AI study notes, flashcard generator from PDF, lecture recording to notes, quiz maker from textbook, and alternatives to generic AI chat for studying.

The challenge

Before myr, Learnard's paid search pain points were seasonal and structural:

Intent pollution
Search terms routinely included:

• "AI do my homework" / "essay writer free"
• "Chegg alternative free" / "answers site"
• Generic "ChatGPT for students" with no study-tool intent
• Download piracy adjacent queries

Seasonal whiplash
CPA looked acceptable in September, spiked in October when competition increased, and the team often reacted with blunt bid cuts instead of surgical negatives.

Conversion complexity
Multiple events (signup, first upload, premium checkout begin) made it easy to optimize toward the wrong micro-conversion unless someone audited tracking monthly.

Limited growth bandwidth
A small team could not manually review search terms across campaigns every day during peak season.

What changed

Learnard connected Google Ads to myr and ran a pre-semester deep analysis. Agents mapped spend to Premium checkout begins as primary, with signup as secondary diagnostic only.

Three shifts:

1. Pre-season negative keyword packs approved before budget ramp
2. Daily agent alerts on new high-spend junk queries during peak weeks
3. Tracking verification checklist triggered automatically when conversion rate diverged from click trend

They treated Google Ads like infrastructure: monitor continuously, change deliberately.

How they run it with myr

Pre-season (2 weeks before ramp)
Export myr deep analysis → approve baseline negatives → set CPA guardrails per campaign tier (brand, category, competitor).

Peak season (daily)
Agents surface:

• New search terms over $25 spend with no Premium intent signals
• Campaigns hitting budget cap before 4pm (lost impression share on winners)
• Conversion rate drops correlated with landing page or tag issues

Weekly
Growth reviews myr ranked list: what to negate, what to isolate into SKAG-style exact campaigns, what to leave alone during learning phase.

Approval workflow
All negatives and pauses approved in myr until team enabled autopilot for known junk patterns (homework solver, essay generator, etc.).

Results

Illustrative outcomes from Learnard's first full semester with myr agents:

Learnard entered back-to-school with cleaner campaigns going in, not cleaner hindsight going out.

Why it works

Edtech paid search breaks when teams chase cheap signups. Learnard used myr to align spend with Premium economics and to run seasonal scale with daily hygiene.

Critical factors:

• Intent filters for academically dishonest and free-only queries
• Seasonal playbooks enforced by agents, not memory
• Conversion sanity checks before bid panic
• Read-only connection + approve changes = trust for a privacy-sensitive product

What's next

Learnard plans to:

• Segment campaigns by persona (student vs professional vs educator) with separate agent guardrails
• Use anomaly detection during finals weeks when CPC spikes are expected
• Tie myr weekly reports into creative testing decisions

The takeaway

Learnard's Google Ads win was not "more keywords." It was protecting seasonality from predictable waste and scaling only when campaigns earned it.

myr gave a small team enterprise-grade monitoring: daily triage, plain-English fixes, and confidence to spend when exams matter most.

Business type: B2C SaaS · Edtech / study tools
Best for: Seasonal B2C with signup → paid upgrade funnels and strict intent requirements