Original Research

DAX 40 Agent Success 2026: the numbers

Thirty-six DAX 40 brands scored on two axes in Wave Q2 2026 Pilot. Every number in one table.

Hyperize 7 min read

The DAX 40 Agent Success Index, Wave Q2 2026 Pilot, scores 36 of 40 brands on both axes: Deutsche Telekom first at 7.0, Brenntag last of 36 at 2.5. [S1] A buyer opens ChatGPT and asks for a product. The agent has to find the brand, then finish the job on the brand's own site. Both steps are scored, with every task frozen before the wave ran. [S3] The full table is below, the data is CC BY 4.0.

01

Two axes, one score, tasks frozen before the wave.

The score is one number from two measured axes and one confidence layer [S2]. AI Visibility asks whether an AI agent finds and recommends the brand when a buyer asks without naming it. AI Usability asks whether an agent that has arrived can complete a real task on the brand's own surface.

The weighting is public and fixed: AI Visibility × 0.20, AI Usability × 0.70, Evidence × 0.10, shown on a 0 to 10 scale [S2]. Evidence is a confidence note on the measurement, not a third axis to sell against.

Every task is frozen and published before the wave runs, under the Task Selection Doctrine [S3]. One task per brand in this wave, run across three AI providers. The measurement language is German, except for Qiagen, measured in English as the buyer language for expert diagnostics [S5].

Confidence is C for most brands: one measured task each, one wave. Qiagen carries B on full datapoint coverage [S5]. adidas and Zalando carry D, because every measured agent class was blocked before a working surface could be scored [S1].

Two axes, one number, and every task frozen before anyone ran it.

02

Thirty-six brands ranked, from 7.0 down to 2.5.

Six brands score 5.0 or higher, thirteen stay below 4.0, and the median sits at 4.1 [S1]. At the top, Deutsche Telekom at 7.0: every measured agent class reaches the cart [S4]. At the bottom, Brenntag at 2.5: the buy-critical specification is not on the page, so no agent could confirm the order [S8].

Rank, brand, ticker, sector, the two axes and the score. Each brand name links to the reference page that carries the task, the access profile and the sources.

Wave Q2 2026 Pilot · 36 brands · Agent Success Score, 0 to 10
Rank Brand Ticker Sector AI Visibility AI Usability Score
1 Deutsche Telekom DTE Tech & Telecom 33.8 80 7.0
2 Qiagen QIA Pharma & Healthcare 11.5 80 6.8
3 DHL Group DHL Logistics 46.0 68 6.5
4 Infineon IFX Industrials 39.1 68 6.2
5 Siemens SIE Industrials 44.0 57 5.6
6 Merck MRK Pharma & Healthcare 38.3 54 5.2
7 Deutsche Bank DBK Insurance & Finance 40.5 48 4.9
8 MTU Aero Engines MTX Aerospace & Defense 52.5 45 4.8
9 Commerzbank CBK Insurance & Finance 38.0 48 4.8
10 Fresenius Medical Care FME Pharma & Healthcare 45.0 44 4.7
11 Siemens Energy ENR Industrials 41.9 45 4.6
12 Fresenius FRE Pharma & Healthcare 40.6 44 4.6
13 Symrise SY1 Chemicals & Materials 36.3 44 4.5
14 Vonovia VNA Real Estate 42.8 41 4.4
15 Hannover Rück HNR1 Insurance & Finance 28.2 45 4.4
16 Munich Re MUV2 Insurance & Finance 44.2 38 4.2
17 GEA Group G1A Industrials 40.3 38 4.1
18 Henkel HEN3 Consumer & Retail 35.3 39 4.1
19 BASF BAS Chemicals & Materials 37.7 39 4.1
20 Volkswagen VOW3 Automotive 49.7 34 4.1
21 Beiersdorf BEI Consumer & Retail 34.4 39 4.1
22 Siemens Healthineers SHL Industrials 33.3 38 4.0
23 Mercedes-Benz MBG Automotive 43.6 34 4.0
24 Rheinmetall RHM Aerospace & Defense 49.4 30 3.7
25 Daimler Truck DTG Automotive 21.4 38 3.7
26 E.ON EOAN Utilities 26.3 35 3.7
27 Airbus AIR Aerospace & Defense 43.9 30 3.6
28 Continental CON Automotive 26.4 34 3.6
29 Heidelberg Materials HEI Chemicals & Materials 40.2 30 3.6
30 Zalando ZAL Consumer & Retail 43.3 28 3.5
31 adidas ADS Consumer & Retail 35.0 28 3.4
32 Allianz ALV Insurance & Finance 19.8 30 3.2
33 Bayer BAYN Pharma & Healthcare 22.1 30 3.1
34 SAP SAP Tech & Telecom 29.4 26 3.1
35 BMW BMW Automotive 37.6 23 3.0
36 Brenntag BNR Chemicals & Materials 24.2 19 2.5
Source: Hyperize DAX 40 Agent Success Index, Wave Q2 2026 Pilot, as published on 2026-09-14. Score = (AI Visibility × 0.20) + (AI Usability × 0.70) + (Evidence × 0.10), shown 0 to 10. AI Visibility and AI Usability as published per brand. CC BY 4.0.
Coverage

36 of 40 member brands are scored on both axes. Scout24 is measured as a marketplace subject in its own lane and is not blended into this ranking. Hochtief is scheduled for Wave Q3 2026.

RWE and Deutsche Börse sit outside the commerce lane: no retail transaction takes place on their brand surface, so there is nothing to measure there [S1]. Universe as of 2026-07-01, reviewed after each Deutsche Börse index review.

Thirty-six scores, one rule. The median DAX brand sits at 4.1.

03

Three patterns recur across the 36.

Three patterns explain more of the table than the sector column does. Each comes from the published reference page of the brand that shows it most clearly.

Qiagen: a near-flawless surface that mass-market AI rarely finds.

Qiagen scores 6.8, rank 2 in the set, with AI Usability 80 and AI Visibility 11.5 [S5]. Ask about a specific diagnostic kit and Qiagen surfaces reliably. Ask a broad category question and it comes up about one time in ten.

Once an agent arrives, everything works: server-rendered pages, a working product finder, an add-to-cart that needs no login [S5]. The build is ahead of being found. Pattern one: a surface can be ready long before discovery is.

Allianz: comparison portals capture the category before the brand page.

Allianz scores 3.2 with AI Visibility 19.8 [S6]. Ask an agent for car insurance and it reaches allianz.de about one time in five. Comparison portals such as Check24 and Verivox capture the category first.

Once reached, the surface works: a full browser agent runs the calculator to a quote, while a plain fetch gets a 403 [S6]. The reference page classifies the interception as avoidable, because agents route directly when the prompt names the brand or when the brand's own answers rank above the portals. Pattern two: the channel decides before the surface does.

adidas: the door closes before product and cart.

adidas scores 3.4 with every measured agent class blocked [S7]. Ask an agent to add an Ultraboost 5 to a cart on adidas.de and the answer is an Akamai 403 page at homepage entry, before search or product detail.

The reference page records this as a documented hard-block finding, not a measurement of a working surface, at Confidence D [S7]. Pattern three: a bot wall that turns away crawlers also turns away the agent that came to buy.

Ready but not found. Found but intercepted. Not even through the door.

04

Talent is measured separately and never blended.

The Talent lane asks a different question: can a candidate's agent find one live software or IT posting on the brand's own career site and reach the application form [S9]? The task was frozen on 2 July 2026, one posting per brand, and no application was submitted anywhere.

AI Usability is measured for 36 of 40 member brands [S9]:

  • 26 brands: the agent reached the application form.
  • 9 brands: the posting was found and the specification matched, but the form was not reached.
  • 1 brand: blocked before the posting could be read.
  • 4 brands: no score this wave, because no single posting could be locked for a complete measurement.

AI Visibility for this lane is still pending, so there is no Talent composite. Talent rows never blend into the Commerce ranking above or into any index average. Two lanes, two tables, no mixing.

05

How to cite these numbers.

The index dataset is published under CC BY 4.0: quote the numbers, name the source, link the index [S1]. The suggested citation and the machine-readable records are below.

Suggested citation

Hyperize (2026). DAX 40 Agent Success Index, Wave Q2 2026 Pilot. https://www.hyperize.ai/en/dax40-index

Numbers as published on 2026-09-14. The index is a living dataset and re-scores each wave; this page is the dated record of the Q2 2026 wave.

The numbers move each wave. A brand that wants a different number next wave needs two things: a surface where an agent can finish the task, and an answer agents cite before the portal. The weighting rewards the first at 0.70 and the second at 0.20 [S2].

Questions about the data: hello@hyperize.ai.

What comes next
Wave Q3 2026 adds brands and the next sector slate on the same two axes, with Hochtief scheduled. This page stays the dated record of the Q2 2026 wave. The live index carries the current numbers, and the freshness strip below carries the date this page was last checked against it.
Sources

Evidence and provenance.

[S1]

DAX 40 Agent Success Index, Wave Q2 2026 Pilot (HTML page and JSON sibling)

Hyperize · 2026-09-14 · External

All 36 scores on both axes, AI Visibility and AI Usability per brand, the coverage counts (36 of 40 scored, one marketplace lane, one brand scheduled, two commerce-lane exclusions), the universe date, per-brand confidence and the CC BY 4.0 licence. Fetched 2026-09-14 from the JSON sibling, index last modified 2026-07-21.

https://www.hyperize.ai/en/dax40-index

[S2]

Agent Success Score: methodology

Hyperize · 2026-09-14 · External

The two measured axes, Evidence as the confidence layer, the public weighting 0.20 / 0.70 / 0.10 and the 0 to 10 scale.

https://www.hyperize.ai/en/methodology/agent-success-score

[S3]

Task Selection Doctrine

Hyperize · 2026-09-14 · External

Tasks frozen and published before the wave, one task per brand, cross-brand fairness rules.

https://www.hyperize.ai/en/methodology/task-selection

[S4]

Deutsche Telekom, reference page

Hyperize · 2026-09-14 · External

Score 7.0. Every measured agent class reads the tariff price and the browser agent reaches the cart in the live shop; the gap is the unbranded query, where discounters answer first.

https://www.hyperize.ai/en/dax40-index/brands/deutsche-telekom

[S5]

Qiagen, reference page

Hyperize · 2026-09-14 · External

Score 6.8, Confidence B, AI Visibility 11.5, AI Usability 80. Server-rendered pages, a working product finder, add-to-cart without login; broad category questions surface the brand about one time in ten; measured in English.

https://www.hyperize.ai/en/dax40-index/brands/qiagen

[S6]

Allianz, reference page

Hyperize · 2026-09-14 · External

Score 3.2, AI Visibility 19.8. Agents reach allianz.de about one time in five on the car-insurance task; comparison portals (Check24, Verivox) capture the category first; a full browser agent runs the calculator to a quote while a plain fetch gets a 403; interception classified as avoidable.

https://www.hyperize.ai/en/dax40-index/brands/allianz

[S7]

adidas, reference page

Hyperize · 2026-09-14 · External

Score 3.4, Confidence D. Akamai 403 at homepage entry for every measured agent class on the add-to-cart task; recorded as a documented hard-block finding, not a measurement of a working surface.

https://www.hyperize.ai/en/dax40-index/brands/adidas

[S8]

Brenntag, reference page

Hyperize · 2026-09-14 · External

Score 2.5. The locked buy-critical specification (technical-grade isopropanol, 200-litre drum, delivery in NRW) is not confirmable on the public surface; the inquiry form stalls at a validation wall.

https://www.hyperize.ai/en/dax40-index/brands/brenntag

[S9]

DAX 40 Agent Success Index, Talent lane (Talent view of the hub and the talentLane block of the JSON sibling)

Hyperize · 2026-09-14 · External

Task frozen 2026-07-02; AI Usability measured for 36 of 40 member brands: 26 application-ready, 9 at spec-match, 1 blocked, 4 without a score; AI Visibility pending, no Talent composite; talent rows never blend into the Commerce ranking.

https://www.hyperize.ai/en/dax40-index.json

Related

Three ways in. Into the method. Into the index.

Methodology · Agent Success Score

Two axes, one number.

How AI Visibility and AI Usability are measured, how the 0.20 / 0.70 / 0.10 weighting works, and why Evidence is a confidence layer rather than a third axis.

Read the methodology

Index · DAX 40 Agent Success

The living dataset behind this table.

36 brands with a reference page each: the task the brand was tested on, both axes, the access profile, the sources, and a JSON sibling for agents.

Open the index

Snapshot · your brand

Same method, off-cycle.

A free Agent Success snapshot for one brand on the same two axes, results in about 48 hours. The form sits on the homepage.

Get the free snapshot
Last updated · 2026-09-14 Next review · 2026-11-14 Tier · B (audit-derived) Confidence · C (one task per brand; Qiagen B, adidas and Zalando D) Wave · Q2 2026 Pilot · data as of 2026-09-14 Machine-readable record

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. Responses received are published in full alongside the findings. Full methodology and editorial-coverage notice: coverage statement.