NBA by Channel: What Works on Online Banking, App Push, Email and in the Branch

Part 2 of 3 in our series on next best action in banking. Part 1 covers the decision logic behind NBA (Part 1), Part 3 looks at why transaction data is the most valuable signal (Part 3).

One recommendation, five possible delivery points

A next-best-action recommendation only works once it reaches the customer where they actually are — and in a form that fits the channel. The same recommendation, "offer customer X a savings product," might appear as a subtle tile in online banking, a short push notification, a detailed email, or a talking point for an advisor in a branch conversation. Which channel fits depends on three things: how strong the channel is in principle, how fast a message reaches the customer there, and what regulatory requirements apply to form and documentation.

How far digital channels have already come in DACH

The starting point can be stated precisely: 89% of respondents with checking-account access used online banking in 2025, up 8 percentage points from 2023 — and the Bundesbank explicitly notes that this increase occurred "across all population groups," not just among younger users (Deutsche Bundesbank: Payment behaviour in Germany 2025). Bitkom finds a similarly high peak of 86% for Germany and shows digital channels displacing the branch: 44% of online-banking users no longer visit a branch at all, according to that survey (Bitkom). Internationally, Accenture confirms this picture with concrete contact-frequency figures: in a study of 49,300 customers across 39 countries, the app logs 152 contacts per year, while the branch is down to just 8 contacts per year (Accenture Global Banking Consumer Study 2025). For an NBA strategy, that means the branch remains relevant, but as a rare, high-value touchpoint — not as the primary delivery channel for ongoing recommendations.

Why context makes or breaks the push channel

No other channel shows as clearly how much relevance drives performance as app push. Generic push notifications in banking and finance average just 8.8% (Android) and 7.2% (iOS) open rates — contextualized messages tied to actual customer behavior instead reach 14.4% versus 4.19% for purely generic pushes, more than triple (Batch: Push Notifications Benchmark 2025). Tight segmentation amplifies this effect further: it can push click-through rates as high as 9.35% — more than 14 times the average — and first-name personalization alone doubles click-through rate (Pushwoosh). For an NBA engine, this is the key lesson: push is the channel with the widest gap between poor and good execution, and closing that gap requires precise, behavior-based targeting.

The NBA channel playbook

The table below maps the five most relevant NBA delivery channels by strength, latency, suitable format, and compliance considerations.

Table sources: Deutsche Bundesbank, Bitkom, Accenture, Batch, Mailchimp, Emarsys/SAP, BaFin.

Channel choice and trust in AI-driven recommendations

Channel choice is not just a question of technical reach — it is also a question of how much customers trust AI-generated recommendations in the first place. A recent Bitkom survey shows a split picture: 27% of respondents could imagine letting an AI handle most of their financial decisions, while 49% reject AI in financial matters outright; 56% see AI in finance as an opportunity and 40% as a risk — with a clear age gap: 68% of 16- to 29-year-olds see opportunity, versus only 35% of those over 65 (Bitkom: A quarter want AI to decide on their finances). For channel strategy, the implication is clear: the same NBA recommendation should be framed differently depending on the likely skepticism of a customer segment — more restrained and more explanatory in digital channels for customers with lower AI acceptance, more direct for digitally native segments. This kind of differentiation is only possible with the channel and audience understanding described in Part 1 of this series.

The branch remains a case of its own

Despite declining contact frequency, the branch is not a fading channel for NBA — it is a different one. According to Accenture, 65% of customers still see branches as a symbol of stability, even though branch numbers in Europe have fallen by 40% over the past decade (Accenture Global Banking Consumer Study 2025). For NBA logic, that means a portion of high-value, advice-intensive recommendations — complex financing or retirement products, for instance — should be routed deliberately into an advisor's CRM queue, together with a rationale the advisor can act on. This kind of explainability toward the advisor is, in practice, the same requirement that MaRisk AT 4.3.5, item 6 imposes toward supervisors (BaFin) — except here it has to be delivered live, in the customer conversation.

Email and SMS as complements, not replacements

Despite shrinking attention, email remains an important channel for more detailed NBA messages that need rationale and multiple next steps — the business and finance sector average sits at a 31.35% open rate and 2.78% CTR (Mailchimp Email Marketing Benchmarks), while current 2026 omnichannel benchmarks put B2C open rates at 40.0% (Emarsys/SAP: Omnichannel Engagement Benchmarks for 2026). SMS, per the same benchmarks, reaches open rates of 90–98% with a tight format — suited to time-critical, short NBA prompts like due-date reminders, but not to product recommendations that need explanation.

How Acceleraid delivers across channels

Acceleraid's CLM/CVM orchestration translates a single NBA recommendation into the right form for each channel and delivers it in real time across online banking, app, email, and branch CRM (Platform). Contact-frequency limits and channel preferences are built into that orchestration, so a customer is not approached with the same message across multiple channels at once (Banking). This capability is the practical answer to the channel heterogeneity described in this article: one and the same recommendation needs five different implementations.

Conclusion

Channel selection is not a downstream detail of NBA logic — it is part of the decision itself. Data from the Bundesbank, Bitkom, and Accenture point to a clear shift toward online banking and the app as primary touchpoints, while the branch becomes rarer but more symbolically significant. Banks that fail to reflect these differences in strength, latency, and compliance requirements leave much of the uplift that contextualization — as shown with push — can deliver on the table. Part 3 of this series turns to the data source that makes this contextualization possible in the first place: transaction data.

Illustration: AI-generated. AI-assisted content: We use AI technologies and automated agents in the creation of our articles, including from Microsoft, Google, OpenAI, Anthropic and other providers. Topics, editorial direction and final approval remain with our team.