---
title: "Deutsche Bank — DAX 40 Agent Success Score | Hyperize"
description: "Deutsche Bank: AI Visibility 40.5/100, primary gap Usability. Hyperize DAX 40 Agent Success Index, measured under a public Task Selection Doctrine."
canonical: https://www.hyperize.ai/en/dax40-index/brands/deutsche-bank
lang: en
last-updated: 2026-07-02
---

# Deutsche Bank.

[Home](/en/) / [DAX 40 index](/en/dax40-index) / Deutsche Bank

Original research · Insurance & Finance



Last measured · 02 Jul 2026 Wave · Q2-2026-W2 Tier · proprietary Confidence · C

Brand ![Deutsche Bank](/logos/dax40/deutsche-bank.svg)

Deutsche Bank AG

Agent success

## Found and form-ready across every agent class.

Bottleneck Usability

Found, but the surface stalls the agent.

4.9 /10

Agent Success Score

AI Visibility 40 / 100

Found & recommended by AI agents

AI Usability 48 / 100

Can an agent use the surface to get the job done

Coverage · 2 of 6 lanes measured Commerce lane · Wave Q2-2026-W2

Commerce 4.9

Talent usability measured

After-sales —

Procurement —

Investor —

Press —

This page measures two lanes. Commerce: can an agent find Deutsche Bank, and once it arrives, transact. Talent: can a candidate's agent find an open engineering role on Deutsche Bank's own career surface and prepare the application. After-sales, procurement, investor and press lanes run on different surfaces and are not yet measured.

Your brand isn't measured yet

## What does an AI agent do with your website?

The same measurement as Deutsche Bank, free for your domain. Five agent classes, one real task, your score in 48 hours.

The test

## Found, parsed, and form-ready across every class. KYC is the last mile.

We asked five kinds of AI agent to reach an account-opening state for the Deutsche Bank AktivKonto, for an employee on 3,500 euro net per month. A plain fetch and a script both held the locked GT (the AktivKonto and its monthly fee), and the browser agent reached the opening form in all three runs, no login required to start. The application itself completes only partly under an agent: identity and KYC verification are the regulated last step.

One label, many breeds. From a plain reader to an autonomous operator, the kinds behind ChatGPT, Perplexity, and Claude Code:

Plain reader reads your raw page text, no browser Succeeded

Search assistant finds you through search Not yet run

Coding agent a script hitting your site Succeeded

Computer-use agent clicks and types like a person Succeeded

Autonomous operator runs the whole task unattended Partial

Scope. This is one product for one persona. Deutsche Bank offers current accounts, cards, loans, mortgages, investments, and private banking, each with its own flow.

Commerce lane

## Found, and able to transact?

Two questions, measured separately. A brand can be recommended and still un-buyable, or perfectly buyable and never found.

AI Visibility

When someone asks an agent to open a current account, does it route to Deutsche Bank?

40 / 100

Deutsche Bank comes up about four times in ten. The rest of the time, an agent recommends an alternative first.

Discoverability · 18-datapoint audit

AI Usability

Once an agent is on Deutsche Bank's site, can it open an account?

48 / 100

Plain reader and Coding agent and Computer-use agent reach the booking form; Autonomous operator get there only in some runs.

Browser-class agents reach the booking form; full\_automation-class agents reach it only partially, submission was not completed · breakdown above

Evidence · 70 / 100 A measure of how provable and consistent the result is, grounded in cross-method ground-truth agreement (the methods that ran returned the same price), not a separate measured run. A confidence layer on the two scores above, not a third sales axis.

Talent lane

## The talent lane.

A candidate's agent, sent to find an open software-engineering role at Deutsche Bank and prepare the application. Same brand, second surface: the career portal.

AI Usability · Talent lane

### Application-ready for agents.

Both non-browser agent classes read the full job posting in raw HTML with no JavaScript execution required; the application entry is identified on the same page, but submitting requires a Workday candidate account (login) and an English-language CV.

Furthest close-state reached: application ready · frozen task locked 2026-07-02

68 AI Usability / 100 · derived, not hand-rated

Plain reader fetches the posting page raw, no browser pass Plain curl (default UA, no JS) on https://db.wd3.myworkdayjobs.com/en-US/DBWebsite/job/Frankfurt-Theodor-H-A-IBC/Senior-Oracle-Database-Engineer--f-m-x-\_R0425683 returned HTTP/2 200, ~22.4 KB HTML, no bot-challenge, no W

Coding agent parses the posting into structured fields pass Machine-extracted from an embedded JobPosting JSON-LD-style config block in the raw HTML (not a \`<script type="application/ld+json">\` tag, but an equivalent structured object with @type:JobPosting, identifier, jobLocatio

Computer-use agent clicks through the career portal like a person untested

The score above is provisional where a class shows untested: untested classes are excluded from the calculation, so the pending check can only confirm or raise the value, never silently lower it.

AI Visibility · in measurement — the audit run for this lane has not landed yet. No Talent-lane composite is shown until both axes are measured.

What this does not yet cover

This measurement did not test the submission itself: no login was performed, no account was created, no form fields were filled, and no application was ever submitted to Deutsche Bank. Only the publicly reachable job-detail HTML and the stated apply-entry requirements (login, CV) were verified via GET/HEAD requests.

Fairness note

This measurement tested read access and application entry for one live, publicly posted engineering role on careers.db.com, verified open at capture on 2 July 2026. No application was submitted, no account was created, and no personal data was used. Deutsche Bank's requirement of a Workday candidate account and an English CV is recorded as a documented condition of applying, not as a score deduction, and the score shown covers usability only while visibility measurement is still in progress.

What's next

## What this means for Deutsche Bank.

Diagnosis

On product discovery, Deutsche Bank is not hard to reach or parse, every class held the AktivKonto and its fee and reached the opening form. It thins at the regulated last step: identity and KYC verification complete only partly under an agent. The handoff is the gap, with discovery mid (found about four times in ten).

What changes the outcome

Make the regulated handoff legible, not another product page. A structured opening path an agent can carry to the verification step, and clear evidence of what the KYC step requires, so the agent hands off cleanly instead of stalling.

What proof looks like

The proof isn't a simpler fee table. It's agents carrying the account opening to a clean KYC handoff, re-measured each wave.

Audit · €1,900

### Commission an audit.

Where the BrandScore opens the question, an audit closes it. An interpretive engagement on your full surface, scored under the same methodology.

[Get the audit](mailto:hello@hyperize.ai?subject=Audit%20%C2%B7%20Deutsche%20Bank)

Founding · €4,500

### Found with us.

Strategic partnership for brands building agent success as a long-term capability, not a one-off engagement.

[Apply](mailto:hello@hyperize.ai?subject=Founding%20Program%20%C2%B7%20Deutsche%20Bank)

Snapshot

### Audit an adjacent property.

The BrandScore covers the primary domain. Get the same methodology applied to an adjacent property: a country site, a sub-brand, a category beyond the DAX-40 slate.

[Get a Snapshot](mailto:hello@hyperize.ai?subject=Snapshot%20%C2%B7%20adjacent%20property%20for%20Deutsche%20Bank)

This is Deutsche Bank. What about your brand?

## You just saw how an AI agent treats a DAX 40 brand.

The same measurement runs free against your domain: five agent classes, one real buying task, your Agent Success Score in 48 hours, in the exact format of this page.

✓

**Measured, not guessed.** Real agents against your real site, no questionnaire.

✓

**Publicly comparable.** Your score in the same grid as the measured DAX 40 brands.

✓

**Next wave closing.** Start now to be in the next index round.

Channel position

## No intermediary stands between agents and Deutsche Bank. The gap is being found, not the channel.

Measured from the AI Visibility audit: no displacing comparison portal surfaced as the route (0 of 18 responses named one; generic compare-accounts phrasing appeared, but no portal capture). Banking comparison portals exist in the German market; they did not displace Deutsche Bank in these responses. The gap here is being found, not the channel.

100% direct

0% via intermediary

Frozen task slate

## Hyperize-selected tasks.

One task from the public sector grid. Task list is frozen before each wave runs.

Deutsche Bank AktivKonto (current account)

Close state

the booking form

Bottleneck

Found, parsed, and form-ready across all classes; the application completes only partly under an agent because identity/KYC verification is the regulated last step.

Fairness note

Wave 2 (a fleet wave Finance). Single measured task (AktivKonto current account for an employee on 3,500 EUR net; booking\_form\_ready close state). AI Visibility 40.49 (18/18 valid datapoints across 3 providers); AI Usability derived from the a fleet wave access profile. Confidence C, single task; Fairness Review pending the sector fairness grid.

Methodology

## How the score was produced.

Discoverability is audit-pipeline-derived. 3-provider sample (openai, perplexity, anthropic), 3 query variants per task, 2 runs per variant · 18 valid datapoints scored against a five-state handoff cascade. [\[S1\]](#sources)

AI Usability is derived from the access-profile above (usability-derivation/v1.2): how far the best agent reached (close state) modulated by how many agent classes succeeded. The per-class profile is the truth; the score is a reproducible summary of it, not a separate rating. Fleet phases (HTTP / Coding / Browser / ACT) produce the profile. [\[S2\]](#sources)

Formula

Agent Success Score = (AI Visibility × 0.20) + (AI Usability × 0.70) + (Evidence × 0.10)

On a 0–100 scale, displayed 0–10. AI Usability bundles the agent's reach + completion; AI Visibility is audit-derived discoverability. Weighting is public; the per-prompt derivation is not.

Measurement scope

Confidence C · one measured task on a 3-provider track (openai, perplexity, anthropic). Confidence promotes to B with a second task plus a fourth provider on the next wave.

History

## Measurement timeline.

Each wave appends; nothing overwrites. Frozen Wave Rule.

1.  Entry · 01
    
    22 May 2026
    
    Wave · Protocol
    
    WAVE-Q2-2026-W2
    
    ars-methodology/v1.1
    
    First v3 measurement. AI Visibility 40.49 (18/18 valid datapoints, 3 providers, DE). AI Usability derived from a fleet wave: HTTP/coding/browser all success holding the locked GT, ACT reached the account-opening form with no login to start, completion partial at identity/KYC. Confidence C, single task.
    

Sources

## Evidence and provenance.

Public methodology references and internal evidence pointers behind every claim above.

1.  \[S1\]
    
    AI Visibility audit · Deutsche Bank Wave Q2 2026
    
    Accessed · 22 May 2026
    
    Internal · Hyperize evidence
    
    -   · AI Visibility score (18/18 valid datapoints across 3 providers)
    -   · the close state reached (booking\_form\_ready)
2.  \[S2\]
    
    Hyperize fleet · a fleet wave Finance (access profile)
    
    Accessed · 22 May 2026
    
    Internal · Hyperize evidence
    
    -   · how each kind of agent fared (HTTP/coding/browser success, ACT opening-form partial via KYC)
    -   · the close state reached (booking\_form\_ready)
    -   · the tested product (AktivKonto + monthly fee)
3.  \[S3\] [Task Selection Doctrine](https://www.hyperize.ai/en/methodology/task-selection)
    
    Accessed · 22 May 2026
    
    Public · hyperize.ai
    
    -   · fairness declaration
    -   · Third-Party Interception framing

Related

## Read the doctrine. Challenge the score. Extend the slate.

[Methodology

### Task Selection.

The fairness doctrine behind the slate above. Five failure modes, six criteria, public before each wave.

Read](/en/methodology/task-selection) [Challenge

### Disagree with this score.

Send evidence under public Fairness Review. Failed reviews are documented with the named failure mode.

Challenge](mailto:hello@hyperize.ai?subject=DAX%2040%20Pilot%20%E2%80%94%20Deutsche%20Bank%20Reference%20Page%20Challenge) [Extend

### Submit your own task.

Open Surface Run · additive measurement. The Hyperize-selected slate stays frozen; your task gets the same methodology.

Submit](mailto:hello@hyperize.ai?subject=Open%20Surface%20Run%20%E2%80%94%20Deutsche%20Bank)

Last updated · 02 Jul 2026 Next review · 30 Sept 2026 Wave · Q2-2026-W2 Tier · proprietary Confidence · C Index score · 4.9/10 [Machine-readable record](https://www.hyperize.ai/en/dax40-index/brands/deutsche-bank.json)

Universe: DAX 40 composition as of 2026-07-01, reviewed after each Deutsche Boerse index review.

Editorial coverage

The DAX 40 Agent Success Index is a point-in-time snapshot of the agent-success of public digital touchpoints. Results are not statements about product quality, company performance, service quality, or the legal obligations of the brands named. Brand names and logos remain the property of their respective owners and are used solely for identification and reporting purposes in the context of editorial coverage (§ 23 MarkenG, Art. 5 GG).

Brands wishing to respond, engage, or correct a factual error may contact [hello@hyperize.ai](mailto:hello@hyperize.ai). Responses received are published in full alongside the findings. Full methodology and editorial-coverage notice: [coverage statement](/en/imprint#dax40-disclaimer).

---

Canonical: https://www.hyperize.ai/en/dax40-index/brands/deutsche-bank
Machine-readable index: https://www.hyperize.ai/llms.txt