Opinion · India · Public-sector banking · Customer experience · AIHow India quietly rebuilt customer experience in state-run banking
₹85,371 cr (US$13.2 bn1) loss in FY18. ₹1.78 lakh cr (US$21.3 bn1) profit in FY25.
Eight editions of one reform programme. Seven years.
In January 2018, the Ministry of Finance and the Indian Banks' Association launched a reform programme few outside the sector noticed. It was called EASE — Enhanced Access and Service Excellence — and at its first measurement, the twelve public sector banks it covered were collectively in the worst shape they had been in for a decade. Aggregate loss for FY2017-18 was ₹85,371 crore (≈US$13.2 billion1). Non-performing assets sat at record highs. Customer experience scores trailed the private banks by margins that were measured rather than guessed. The programme set out, with quiet ambition, to change all three.
Seven years and eight versions later, the picture is different. In FY2024-25, every public sector bank was profitable. The aggregate net profit for the group reached ₹1.78 lakh crore (≈US$21.3 billion1) — the highest in their history. The average EASE Index score has roughly doubled from its FY20 baseline. More importantly for the customer, three things that the index measures most heavily — digital adoption, customer responsiveness, and the maturity of personalised banking — have shifted decisively. EASE 8.0, now branded EASERise around four R.I.S.E. themes (Risk & Resilience, Innovation, Socio-economic impact, Excellence), has placed Generative AI and Agentic AI explicitly on the customer-experience agenda. Seven PSBs have already deployed thirty-two GenAI use cases in production under EASE 8.0.
The customer is the centre of the agenda
Through every edition of EASE, the most heavily weighted themes are the ones the customer experiences directly: digital customer experience, customer responsiveness, and (from EASE 5.0 onwards) the maturity of analytics-driven personalisation. The substantive shift inside PSBs over the seven-year period is not that they have got better at banking in the abstract — it is that they have got materially better at the touchpoints. App ratings, NPS, self-service adoption, first-contact resolution, and the share of journeys that complete digitally have all moved. The leading PSBs have closed a meaningful share of the digital-and-CX gap with the private sector. The gap remains, and the framing of EASE 8.0 acknowledges it; but the curve is in the right direction, and the rate of closure has accelerated with each edition.
The trajectory of the programme tracks the industry: cleanup, then digital, then personalisation, now AI. Each edition has built on the last and pushed the standard.
| Edition | Year | Customer / digital focus | What it pushed PSBs to do |
|---|---|---|---|
| EASE 1.0 | FY18–19 | Foundational baseline. Six themes; measurement framework established. | NPA cleanup, governance reset, core digital plumbing put in place. |
| EASE 2.0 | FY19–20 | 27 action points across six themes; doorstep banking launched. | Grievance redressal SLAs, senior-citizen services, first wave of mobile banking adoption. |
| EASE 3.0 | FY20–21 | "Tech-enabled, simplified and collaborative banking." Mobile-first. | Paperless onboarding, dial-a-loan, integrated digital lending journeys. |
| EASE 4.0 | FY21–22 | Analytics maturity; co-lending with NBFCs; smart lending. | Cross-channel campaign analytics, behavioural segmentation, MSME credit auto-decisioning. |
| EASE 5.0 | FY22–23 | Digital customer experience as the centre of the agenda. Cloud and APIs. | Account Aggregator integration, open APIs, real-time decisioning at the app surface. |
| EASE 6.0 | FY23–24 | "Future-ready, customer-centric, digitally-enabled" banking. Hyper-personalisation. | Next-best-action engines, contextual offers, propensity-scored journeys at scale. |
| EASE 7.0 | FY24–25 | Smart automation. Queue management, MSME credit deepening, agentic operations. | Branch-level automation, AI-assisted relationship banking, personalised in-app credit offers. |
The leaderboard tells a story of consistency
Within the aggregate improvement, the rankings have been competitive. The same handful of banks recur near the top of the Top Performing Banks category, and a different cohort of mid-pack banks regularly take the Top Improvers awards — exactly the differentiation the programme was designed to surface.
| Edition | Top Performing Bank — 1st | Runners-up | Top Improver(s) |
|---|---|---|---|
| EASE 2.0 | Bank of Baroda | State Bank of India · Oriental Bank of Commerce | Bank of Maharashtra · Central Bank of India · Corporation Bank |
| EASE 3.0 | State Bank of India | Bank of Baroda · Union Bank of India | Indian Bank (best improvement from baseline) |
| EASE 4.0 | Bank of Baroda | — | — |
| EASE 5.0 | — | Bank of Baroda (2nd Runner-Up) | — |
| EASE 6.0 | State Bank of India (cited) | — | Punjab & Sind Bank |
| EASE 7.0 | — | Bank of Baroda (1st Runner-Up) | — |
Two patterns matter. First, the leaders are stable: Bank of Baroda and SBI have alternated at the top across multiple editions, with Union Bank of India and Indian Bank consistently in or near the top three. Second, the Top Improver awards rotate — Bank of Maharashtra, Central Bank of India, Punjab & Sind Bank, and others have used the EASE measurement to push themselves forward from a lower base. That is the mechanism the programme was designed to create.
EASE 8.0 — R.I.S.E. — and what the new framework actually measures
EASE 8.0 (FY26) was launched as EASERise, a deliberate brand reset that re-bundles the previous themes into four pillars and makes Gen-AI and Agentic AI an explicit pillar rather than an option. Two of the four pillars relate directly to the customer surface — innovation (how new the customer journey is) and excellence (how good the customer journey is).
| Pillar | What is measured | Customer-experience signal | |
|---|---|---|---|
| R | Risk & Resilience | Credit risk management, cyber resilience, operational risk, asset quality. The hygiene layer below customer trust. | Bank stays up. Money is safe. Compliance does not break the journey. |
| I | Innovation | Gen-AI & Agentic AI adoption, new digital products, open-API banking, AI-assisted underwriting, real-time decisioning. | The app does things it could not last year. Decisions are instant; offers are contextual. |
| S | Socio-economic Impact | MSME credit deepening, financial inclusion, priority-sector lending, agricultural credit, sustainability. | The underserved are now served. The first-time borrower is onboarded digitally. |
| E | Excellence | Customer experience & responsiveness, hyper-personalisation, operational efficiency, employee productivity. | The journey is fast, personal, and resolved on first contact. |
The AI-native partner ecosystem — where capability comes from
The interesting question for FY26 is no longer whether PSBs will use Gen-AI and personalisation engines; the EASE scorecard now requires it. The question is where the capability comes from. Building in-house at the pace EASE expects — across 12 banks, against a moving private-sector benchmark — is implausible. The realistic answer, and the answer the leading PSBs are converging on, is partnership with India's growing AI-native software ecosystem: Startup India-recognised, product-led companies that move at the pace of a SaaS roadmap rather than the cadence of a systems integrator.
The clearest documented example is Union Bank of India's customer-engagement transformation in partnership with Appice — a Startup India-recognised, New Delhi-based AI-native customer engagement platform. The integration combined a Customer Data Platform with ML-Ops orchestration and applied it to a previously dormant customer segment. The published results are operational, not aspirational: 26% offer uptake, 4× lift in click-through, and a 69% open rate against personalised credit offers. Those numbers map directly into three EASE scoring axes — digital customer experience, customer responsiveness, and analytics-driven personalisation — which is how a CX programme becomes a scorecard outcome.
The reform programme set the agenda. The AI-native partners are the mechanism through which the agenda becomes operational reality — and the customer feels it.
The pattern around Union Bank's example is what to watch across the sector. Bank of Baroda, Canara Bank, Punjab National Bank, and Indian Bank have each made public commitments to partner with the domestic AI-native ecosystem to deliver against EASE 8.0 objectives, and a number of smaller PSBs are evaluating similar arrangements. The macro consequence is straightforward: a Startup India ecosystem that has historically sold to private banks and digital-first fintechs is now selling, at scale, to state-run banking. The EASE scorecard is the procurement signal.
What this means
The PSB transformation under EASE is not finished, and the gap with the best private banks is not closed. But the trajectory is real, measurable, and increasingly autonomous of the original political sponsorship — the reform has become an operating system rather than a policy. The next two editions will test whether AI capability can be acquired at the pace EASE expects, and whether the smaller PSBs can keep up with the leaders or fall further behind. The early evidence — thirty-two GenAI use cases live, seven banks already on the EASE 8.0 scoreboard, an AI-native partner ecosystem now selling into the sector — suggests the answer to the first question is yes. The answer to the second is still being written.
A reform programme launched to fix a balance-sheet crisis is now, seven years on, the mechanism through which India's state-run banks compete on customer experience and AI. That is a story worth telling.
Whether the next decade of banking customer experience runs on a substrate the banks own, or one they rent, is a question this series will keep returning to.