Addictive Learning Technology Ltd Q2 FY2026 Concall Summary

Key Takeaways

• Sales scaling failure drove a revenue miss: The hoped-for ₹50-60 Cr H1 did not materialize because repeated attempts to scale the sales team failed, each time teaching a critical lesson but never closing the gap.
• A new AI-driven sales process is the fix: A proprietary AI engine now converts discovery-call transcripts into personalized roadmaps, letting freshers close at 1:4 to 1:5 rates and removing the need for experienced closers.
• Community economics are overtaking bootcamps: Community ROAS runs at 2x bootcamp ROAS, so the company is deliberately replacing bootcamps with communities, a shift that promises better stability and predictability.
• The "Holy Trinity" is the long-term moat: Management is building three interlocking assets: a job platform for leverage, an uncopiable AI-driven delivery system, and a US University certification, after which it expects to scale at an incredible pace.
• What to watch next quarter: Whether the new AI sales organization actually delivers the promised 50-100% new-sales growth, whether the US University gets final board approval in January or February, and whether the auto-funnel model shows any sign of life.

The Sales Scaling Problem & the AI Answer

• Repeated failures, accumulated lessons: Earlier scaling attempts were highly training-dependent because experienced salespeople struggled to learn complex products like California Bar, Enrolled Agent, and Independent Director programs. Each failed attempt taught critical lessons that fed into the current redesign.
• The new model replaces closers with process: The company built a ~70-person AI-enabled sales organization from scratch, hiring freshers and college students. A two-call system runs discovery (30-40 minutes), produces a JSON transcript, and feeds it into a proprietary AI engine that generates a personalized 20-page PDF roadmap. The caller studies and explains the roadmap to the lead, achieving 1:4 to 1:5 conversion.
• Auto-funnel remains unsolved: The auto-funnel model is, in management's own words, "very bad" and not yet cracked, leaving the new AI sales process as the primary growth lever for now.
• CAC target is aggressive: Management aims to cut customer acquisition cost by 30% in the next three months, possibly down to 25% or less, a target that depends entirely on the new AI process delivering at scale.

Ramanuj Mukherjee: "We basically hire freshers or even college students and we train them with an AI-enabled process. It has kind of removed the requirement of closers, removed the requirement of experienced sales people. It's very process-driven sales."

Channel Shift: Communities Over Bootcamps

• Bootcamp saturation is real but contained: Only bootcamp sales are saturating; new channels like communities and AI sales calls are expected to unlock growth. The company plans to reduce bootcamps from 8 per month to 4-6 as communities scale.
• Community ROAS is 2x: Communities deliver twice the return on ad spend of bootcamps and offer better stability and predictability, so the company is systematically replacing bootcamps with communities over time.
• Course creation outpaced distribution: Many new categories were launched in the last year, but distribution did not keep pace, so results are not yet visible. The new sales organization is meant to close that gap.

The Holy Trinity & Long-Term Moat

• Three interlocking assets: Management's stated endgame is to control the Job Platform (leverage), build an AI-Driven Delivery Mechanism that cannot be copied, and secure US University Certification. Once all three are in place, the company expects to scale at an incredible pace.
• Job platform traction: The LawSikho Opportunity Portal is drawing about 1.5 Lakh users per month, and an AI system for matching jobs and applicants has been launched. Skillarbitragejobs.com is also live but not yet scaled.
• AI delivery system under construction: The company is building an AI agent that reviews Upwork and LinkedIn profiles and coaches students, a system management claims no other upskilling company has built. A new course delivery system where AI asks questions, refines roadmaps, and assigns daily tasks is targeted for a March launch.
• US University progress: A pre-application in Arizona was approved with recommendations. The entity can launch an online MBA program but must use "Institute" rather than "University" for now. Final board meetings are scheduled for January and February; approval would give the green light. The company will go slow on selling in the US (which requires US-specific processes, accent, and hiring) but will not go slow on selling the US degree to Indian students as an upsell to increase ARPU.
• IIT Roorkee tie-up adds credentialing: A tie-up with iHub Divyasampark (IIT Roorkee) lets the company offer a top-up IIT Roorkee certification on top-selling programs, including an immersion program, certification, and alumni status.
• Global certificate programs address a regulatory gray area: The company is launching certificate programs globally; US University status helps it avoid the "unregulated bootcamp" label that creates friction in some markets.

Ramanuj Mukherjee: "Control the Job Platform (Leverage). AI-Driven Delivery Mechanism (Cannot be copied). US University Certification. Once I hit this trinity, we will be scaling at an incredible pace."

AI Automation Across Operations

• Support emails now AI-handled: Incoming support emails are processed by AI that sees previous conversations, takes context, and creates replies, reducing human intervention.
• AI call quality tool has B2B potential: The proprietary AI engine that generates personalized roadmaps from call transcripts is being considered for a patent. Management intends to sell it as a B2B business but wants to crack its own sales with it first.
• Student success proof points: For the first time, two students crossed \$200,000 in income from Upwork, a milestone that validates the platform's earning potential for learners.

Financial Position & Capital Allocation

• Cash position is adequate, spending is moderating: The company has enough cash and will be more conservative going forward; major spending on AI and the University setup is mostly done.
• No fundraiser at current prices: Management views the stock price as too low to raise funds and aims to improve the cash position from operating cash flow instead.
• No dividend or buyback yet: Promoters, who are the biggest shareholders, acknowledge it is not the time for capital return; cash is needed for growth and the company must be prioritized.
• Promoter buying signals confidence: Ramanuj bought ₹12 Lakhs and Abhyuday bought ₹10 Lakhs in the last six months, with plans for similar buying in the next six months.
• Main board migration on the radar: The company becomes eligible in 16-17 months and definitely wants to migrate.

People, Governance & Unanswered Questions

• Tech co-founder resigned: One tech co-founder stepped down, approaching retirement age (close to 60), having achieved financial goals and finding the role stressful.
• Management cadre being built: The company is considering recruiting 20 smart people every six months and training them to be deployed across the organization.
• Revenue split was deflected: When asked for a split of revenue between LawSikho and Skill Arbitrage, management answered that LawSikho is doing about 30% and explained the brand strategy, but did not provide the actual split the analyst requested.
• Southeast Asia initiative shelved: A question about a revenue-share model for Southeast Asia products was met with "did not pursue it as I was preoccupied with sales challenges," effectively declining to provide details.
• Judiciary business hit by regulatory change: Judiciary test prep slowed because a Supreme Court order now requires three years of experience for judges; the company is pivoting focus to Supreme Court Clerkship and UGC Net instead.
• Philippines and Malaysia expansion needs local partners: The company is looking at these markets for SQE and California Bar courses but requires local partners to avoid being copied and to handle compliance.

Ramanuj Mukherjee: "Very bad. We haven't cracked it yet."

Guidance & Commitments

Key Metrics