CX in Financial Services: Why the Banking Ranking Is the Wrong Benchmark for B2B [2026]

Where does banking rank in CX, and what does that ranking measure?

Banking ranks 4th of 13 sectors in KPMG's Global Customer Experience Excellence 2025-2026 benchmark, with a score of 7.52. The ranking rests on 80,594 consumer interviews about 2,684 brands in 16 countries. It tells you how private customers rate their bank. It does not tell you how a mid-market company rates its lender, its payments partner or its insurer, and that is the question a B2B financial institution actually needs answered.

Source: KPMG, Global Customer Experience Excellence (CEE) 2025-2026, consumer-based.

Look at the gaps before you celebrate. Banking sits 0.02 points below grocery retail and 0.04 above insurance, so ranks three to five are separated by 0.06 points (our own arithmetic on KPMG's scores). KPMG describes banking as performing slightly above the market average. A fourth place decided in hundredths is not a competitive advantage you can put in a pitch.

Two consumer benchmarks, two directions

KPMG reports a modest improvement for banking: the banking score rose 0.7 percent, with the Empathy pillar up 1.4 percent and Expectations up 1.3 percent. Forrester's US Banks Total Experience Score 2025 (published 24 June 2025, 25 US banks) tells a different story. Forrester found that customer experience in US multichannel banking declined for the fourth consecutive year, that direct banks' CX Index scores fell to their lowest level in nearly a decade, and that fewer than 60 percent of customers trust their bank brand.

The two studies differ in geography, method and brands, and both are consumer research. They are not contradictory so much as incompatible. That is the point. If two credible benchmarks disagree on the direction of travel, a sector headline cannot be your target. And neither of them measures B2B.

What does B2B financial services NPS look like?

The most useful B2B reference we found is CustomerGauge's "Financial Services NPS Benchmarks" page (Ian Luck, 2026 edition, no publication date shown). It describes its averages as B2B and self-reported by CX programmes, with publicly disclosed scores excluded. It puts the banking average at 30, down from 37. The same page gives a different banking figure in its FAQ (34) and two financial services averages: 60 in CustomerGauge's 2026 survey and 46 blended across editions in the NPS Benchmark Report: Decade Edition. The page does not state a sample size or fieldwork dates, and the individual bank scores on it come from different years and are not labelled B2B or B2C.

We cannot give you one clean B2B banking NPS, because the sources do not agree with themselves. If you need a refresher on how the score itself is calculated, start with what NPS is. For a deeper look at how to read sector figures, see NPS Benchmarks 2026: What Is a Good NPS Score by Industry?.

Why is empathy the pillar that moved, and what does it mean in B2B?

Empathy was the pillar KPMG singled out for banking. Our reading, and it is a reading rather than a finding, is that empathy in B2B financial services looks less like a friendly tone and more like continuity: the same relationship manager who knows the account's history, calls before the covenant conversation rather than after, and does not need the client to explain the situation twice. KPMG's consumer data cannot confirm this for commercial clients. A buying group at a mid-market company typically spans five to 16 people across up to four functions (Gartner, 7 May 2025, 632 B2B buyers surveyed in August and September 2024), and that figure describes the purchase decision, not life after signing. But it is a fair warning that one relationship manager rarely sees every person who judges the bank.

The practical test is simple. Ask your accounts who they would call if the relationship manager left tomorrow. If the answer is nobody, your empathy score is a personal favour, not an institutional capability. This is where the six pillars of customer experience become useful: not as consumer weights to copy, but as a vocabulary for testing which pillar carries your own accounts, with key driver analysis. The same applies to personalization: in a commercial relationship it is won per role inside the account, not per campaign.

Is AI the answer to banking CX?

KPMG's report profiles AI-agent deployments at eight banks, among them ING and BNP Paribas, across Europe, Latin America and Asia. That is evidence of ambition. It is not evidence that the deployments improved the customer's experience, because the profiles describe what banks built, not what customers noticed. We would be careful about using them as a reason to start with the model.

Our position: in B2B financial services, AI works only where three things already exist. Feedback data that is clean enough to trust, a named owner for every account, and a closed loop that turns a signal into a call. Without them an agent speeds up the wrong process. We have written about the ownership side in AI and Customer Experience in 2026 and about the structure behind it in CX Governance.

What are the common mistakes in financial services CX?

Three mistakes recur in the programmes we see in the Nordics. These are practitioner observations, not survey results.

The first is benchmarking against the wrong population. A commercial lender compares its relationship NPS with a consumer bank's app score and either panics or relaxes for no good reason. The second is counting responses instead of revenue. A survey with a decent response rate can still miss the ten accounts that carry the book, which is the argument for account-based CX. The third is measuring and then going quiet. A detractor who hears nothing for a month has already started a conversation with a competitor, and the detractor recovery playbook exists for exactly that gap.

A hypothetical: Nordika A/S and its lender

Imagine Nordika A/S, a mid-sized manufacturer with a credit facility, a payments arrangement and an insurance programme with the same financial group. Nordika's finance director scores the relationship 9 on the lender's survey. Nordika's operations manager, who deals with the payments desk daily, would score it 4 if asked. The institution sees a promoter and files the account as safe.

Nobody asked the operations manager, so the average is a story about one person. This is the typical B2B financial services failure: the relationship score reflects the stakeholder who is easiest to reach. The fix is not a bigger survey. It is a survey design that asks at least two roles per key account and treats disagreement between them as the most valuable signal in the data. With small account numbers, read the results with care: how many responses you actually need is a question of method, not of ambition.

How should a B2B financial institution benchmark itself?

Use the sector figures as context and your own accounts as the benchmark. In practice:

First, weight by revenue or exposure, not by head count of respondents. Second, track the same accounts over time, so a change means something. Third, compare roles inside each account before you compare yourself with anyone outside. Fourth, close every detractor response with a named owner and a date, and measure how long the loop takes. Fifth, translate the result into money, as we describe in Customer Satisfaction and Revenue. A relationship score that cannot be linked to retention, cross-sell or margin will lose every budget discussion, and it deserves to.

The contrarian conclusion: a high sector ranking is not good news for you. It raises the expectations of your clients and your board without telling you anything about the accounts that matter. A lower, honest number on your own revenue-weighted base is worth more than a flattering rank from someone else's population.

How does SurveyGauge help financial institutions?

SurveyGauge combines a feedback platform and strategic CX advisory in one subscription. For financial institutions that means relationship surveys designed per account and role, results weighted by revenue, detractor alerts routed to the person who owns the account, and an advisor who helps you turn the pattern into decisions. We are a partner in the work, not a tool you have to figure out alone.