From the Rakuten playbook: How telcos can leverage real-time data and ecosystems to supercharge growth
Mobile networks face a paradox: While the connectivity they provide has never been more essential, with data consumption reaching all-time highs, soaring usage has failed to translate into meaningful revenue growth. A recent PwC global telecom industry report forecasted 2025-2029 compound annual growth rate (CAGR) to drop to just 2.8%, with monthly average revenue per user (ARPU) projecting to slide over 12%, despite some of the highest data growth ever recorded.
For operators around the globe, this is not a technology problem, but a business model problem – one that threatens to prove existential for telcos unable to embrace effective AI-driven approaches.
Rakuten’s four strategic imperatives for telecom growth
At DTW Ignite 2026, one of Europe’s premier telecom industry events, Vasanth Raju, Rakuten Head of AI Product, took the stage to share lessons the company has learned over the past year building and deploying AI agents across the Rakuten Ecosystem to mobile leaders in attendance.
As the industry rushes to accelerate AI adoption, fragmented pilots and chatbots have proliferated. But if AI projects are merely increasing token usage without moving the needle on ARPU or operational margins, costs will continue to rise with minimal returns.
“Simply put, you cannot build a competitive moat by buying a third-party API, particularly in an era where compute and Large Language Models (LLMs) are commoditized. The true advantage lies in what you feed the model,” Raju said.
Rakuten identified data architecture as a true bottleneck. To transform a utility provider into a growth engine, the company proposed four strategic imperatives.
1. Unify the data estate
Rakuten sees AI as an amplifier: It makes the input louder. If you feed it fragmented data, you get fragmented output. The company made a strategic bet in building a deep learning foundation that unifies data across the entire Rakuten Ecosystem, spanning 70 businesses and 10 million offline touchpoints.
By enforcing a single Rakuten ID, the company has turned 45.9 million monthly active users into a high-fidelity profile. “Unlike generic AI agents that rely on inferred ‘shadow’ preferences, our AI reasons over deterministic, first-party transaction data – actual purchases, payments and redemptions,” Raju emphasized. “Companies cannot buy this asset after the fact. They earn it over years by genuinely serving customers.”
2. Solve real user problems
Rakuten prioritizes AI for high-impact utility, focusing on AI agents that plan, execute and complete transactions. For example, users can now search and book a hotel room or a golf course entirely through conversation with Rakuten AI.
“Our focus remains on augmenting human intelligence and care. A prime example is RAPTOR, our Gen AI-powered customer service agent,” Raju said. “By operating on a unified data foundation, RAPTOR goes beyond simple advice to resolve issues and anticipate needs, freeing human agents to service more customers.” By boosting auto-resolution rates across more than 10 business units, RAPTOR is expected to deliver 4.3 billion yen in annualized cost savings in 2026.
3. Partner with the best, build the best
To scale at the speed of the market, Rakuten embraces a dual-track strategy. The company partners with frontier model providers for rapid deployment and general intelligence, while building domain-specific small language models (SLMs) via model distillation. This ensures cost-efficiency, ultra-low latency and total control over proprietary data.
Rakuten currently has 11 AI agents live across services such as Rakuten Ichiba, Travel and Mobile, with more than 50 services planned. “Because we own the full stack, our agents don’t just advise – they transact, creating a ‘first mile’ of personalization and a ‘last mile’ of execution that competitors cannot replicate,” he stated.
4. Close the loyalty feedback loop
The backbone of Rakuten’s growth is the company’s AI growth flywheel, which moves through three sectors: engage, expand and differentiate. Loyalty programs, Rakuten Points and SPU, are not just rewards – they are the connective tissue of the Rakuten Ecosystem. In 2025, the company found that users who engage with the loyalty program use 4.6 times more services than those who don’t. The revenue impact is exponential: customers using three segments (Mobile, E-commerce, FinTech) generate 13.5 times the annual revenue of a single-segment user.
Most importantly, loyalty acts as a massive churn-reduction tool. “If we index the churn of a single-service user at 100, those using four or more services have a churn rate of just 1. We have successfully turned loyalty into a self-reinforcing cycle of revenue,” Raju explained.
From utility to growth
Looking ahead, Rakuten sees the future of telecom as not just the pipes, but the ecosystem built on top. Mobile operators hold the industry’s most valuable real-time context: location, app usage and service patterns. This is the “ultimate sensor.”
To the mobile industry, the company offers a clear roadmap for transformation:
- Leverage the network as the ultimate sensor. Treat mobile context as a proprietary data asset.
- Move beyond the utility trap. Build a data-centric architecture before asking AI to do anything with it.
- Use AI to control the first mile and the last mile. Capture intent early and complete transactions end-to-end.
Business first. Data second. Technology third. The mobile industry must leverage AI today to activate their flywheel for future growth.
The four strategic imperatives outlined above represent one vision for telecom’s future. For a broader look at the mobile industry’s changing economics, and the paths available to operators, read the Rakuten Symphony Industry Growth Report 2026.
