STEN.AIDeutschContact

AI transformation · Sten Franke & ETHORITY

Five eras. One discipline. Now: AI that works.

A web directory in 1997. Search marketing, internet monitoring with machine learning, social media intelligence, and today AI agents. In every era the same three questions decided success: what changes in the work, who decides, and how do we prove the result? ETHORITY turns that experience into a system for your AI transformation.

Five rings for five technology eras from 1997 to today, with client logos on the rings, inside an outer ring of 961 points for the AI offerings in the Good AI Prism.19972001200920182026AI that works1997 → 2026
Select an era

2026 · AI agents and the AI Business OS

Evidence, decisions and operation in one system.

961 AI offerings in the Good AI Prism today

What the market needs

High use. Thin proof.

The Market Study 2026/27 reads more than 30 economies. Its ten findings describe a market that uses AI widely, struggles to prove the value and can rarely compare offerings on facts.

The value gap

Nearly 9 in 10of survey respondents report regular AI use in at least one business function
37 %of respondents report a positive contribution to their company’s EBIT

Source: Good AI Prism Market Study 2026/27, Global edition v2.6, pp. 5–7, citing Gartner, McKinsey, Eurostat, Bitkom, national statistics and Regulation (EU) 2026/1744

  • US$2.67tnAI spending forecast for 2026, mostly into infrastructure. AI has become a running cost.Gartner, Sep 2026
  • 1 in 5EU firms with ten or more staff use AI. Large firms adopt 1.6 to 4.3 times as often as small ones.Eurostat, Dec 2025
  • 85 %of German firms with 20+ staff that do not use AI say they lack technical know-how.Bitkom, Sep 2026
  • 2 Dec 2027EU AI Act Chapter III, Sections 1–3 (except Article 6(5)) start to apply to systems classified as high-risk under Article 6(2) and Annex III.Regulation (EU) 2026/1744

Five needs follow from it.

  1. 01Comparable facts before buyingBuyers can rarely compare offerings on facts. They need sources, dates and limits for every option.Good AI Prism →
  2. 02A business case before scalingOnly 37 % see an EBIT effect. Investments need a baseline and a value that can be measured.AI Opportunity Case →
  3. 03Proof in real operationAgents are the growth story. They have to work with real cases, real volumes and clear acceptance criteria.AI Trust Pilot →
  4. 04Accountability that holdsDuties are dated. Every system needs an owner, evidence and a way to stop it.AI Trust Record →
  5. 05Know-how in your own teamKnow-how is the barrier named most often, and the size gap widens with agents. Capability has to stay in the company.Transformation programme →

The system we are building

From the first question to AI in operation.

Each step answers one of the five needs and ends with a decision. You can start at any point. Sten leads the first conversation personally; ETHORITY provides research, advisory and implementation.

Personal project leadership by Sten Franke

  1. Discover

    Good AI Prism

    Free

    961 AI offerings by task, industry and evidence, each with source and date.

    Ends withA shortlist you can defend

    View offer →
  2. Decide

    AI Decision Brief

    €195 net · one-time

    The right AI offering for one question, compared on facts.

    Ends withA recommendation with source and check date

    View offer →
  3. Plan

    AI Business Blueprint

    €495 net · one-time

    One objective, translated into a 30/60/90-day roadmap.

    Ends withA plan your team can start on Monday

    View offer →
  4. Invest

    AI Opportunity Case

    Priced after scoping

    Economics, risks and the evidence you need before the budget is released.

    Ends withPilot, redesign, expand or stop

    View offer →
  5. Prove

    AI Trust Pilot

    €25,000 net · one-time

    Six weeks, one use case, real cases and explicit acceptance criteria.

    Ends withA measured result in your operation

    View offer →
  6. Govern

    AI Trust Record

    €4,500 net · one-time

    One AI system, ten working days: evidence, open questions and the accountable decision.

    Ends withA record that holds up to review

    View offer →
  7. Run & grow

    Transformation programme

    Priced after scoping

    After the pilot ETHORITY builds and runs the system with your team, as an AI1st organisation with its own AI Business Operating System.

    Ends withAI in daily operation, with capability in your team

    View offer →

Proof

Real projects, from now back to 1997.

Every entry names the client or initiative and what was done. The last line says what it means for AI today. Drag or scroll sideways.

  1. The AI systems behind ETHORITY
    2026

    The AI systems behind ETHORITY

    ETHORITY

    An AI Registry, Navigator and Explorer that track more than 100,000 observed models, tools, datasets, research objects and providers; a Research Factory that turns sources into verified claims; decision and trust records with owners, evidence and stop conditions.

    For AI todayEvidence has to be built into the system, not added at the end.

  2. Good AI Prism and Market Study 2026/27
    2026

    Good AI Prism and Market Study 2026/27

    ETHORITY

    On 23 October the Good AI Prism opened with 961 AI offerings in 24 topic areas, each with its source and date. On 27 October the Market Study followed, covering more than 30 economies.

    For AI todayBuyers need comparable facts before they need more tools.

  3. AI3M: six stages of AI maturity
    2024

    AI3M: six stages of AI maturity

    Workshops with companies

    Six stages from first experiments to responsible AI leadership. Companies use the model in workshops to see where they stand; alongside it came readiness assessments, risk models and governance roadmaps.

    For AI todayMaturity is a leadership question before it is a technology question.

  4. European Central Bank
    2021–2022

    Digital assets and blockchain

    European Central Bank · Solaris · a family office

    A market study and risk assessment on crypto assets for the European Central Bank, a digital-asset strategy for Solaris and an investment and marketing strategy for a family office.

    For AI todayA new technology needs a risk view before it needs a roadmap.

  5. since 2019

    AI in cancer research

    EpiTrace.AI · Dr. Rurainski Foundation

    EpiTrace.AI combines AI with DNA-methylation analysis for research into early cancer detection. Its secure research platform has been running since 2025.

    For AI todayWhere the stakes are high, data governance comes first.

  6. Customer loyalty across all channels
    2018–2021

    Customer loyalty across all channels

    Deutsche Telekom · T-Mobile

    T-Mobile

    What keeps customers, at which touchpoints: a loyalty strategy across all channels, including a conversational AI concept for inbound sales.

    For AI todayAI belongs at the touchpoints that keep customers.

  7. BMWDisney
    2015–2017

    Customer experience in Los Angeles

    BMW Financial Services · Disney

    Customer-experience and social media strategy for BMW Financial Services, and digital marketing strategy for Disney.

    For AI todayCustomer experience is where technology earns its keep.

  8. Social media intelligence for brands
    2012–2014

    Social media intelligence for brands

    Telefónica O2 · Bayer · AIDA Cruises

    Telefónica O2Bayer

    From the Astraturm above Hamburg harbour the team worked for Telefónica O2, Bayer and AIDA Cruises. The Social Media Prisma grew to version 6.0 with 261 platforms and tools.

    For AI todayMarkets move faster than org charts. A good map helps teams choose.

  9. Industry leadership
    2007–2010

    Industry leadership

    BITKOM · BVDW

    Chair of the BITKOM working group on customer satisfaction, a talk at the BITKOM CE Forum in the Sony Center Berlin, and chair of the BVDW Social Media Council from 2009 to 2010.

    For AI todayNew channels need shared rules before they need scale.

  10. Internet monitoring with machine learning
    2001

    Internet monitoring with machine learning

    Gridpatrol · Aldi · AOL · ECB · Daimler

    Aldi SüdEuropean Central Bank

    An early internet monitoring service built on machine learning, with clients including Aldi, AOL, the ECB and Daimler. In 2003 heise reported on its ranking of brand images on the web.

    For AI todayMachine learning is not new. Making it useful for a brand is the work.

  11. Search marketing for the young web
    1997

    Search marketing for the young web

    mediatime · T-Mobile · Kraft Foods · Lufthansa

    T-Mobile

    media-time.de, a directory for the young German web, then international search marketing for T-Mobile, Kraft Foods and Lufthansa, and 1a-Suchmaschinen.de, a paid SEO software service.

    For AI todayBeing found is the first condition of being chosen, now also in AI answers.

Your industry

Where your transformation starts.

Choose an industry. You see the clients Sten has worked with there, a typical first decision and the path we recommend.

Telecommunications

Clients
T-MobileTelefónica O2
What Sten did

Search marketing for T-Mobile (1997); social media intelligence for Telefónica O2 (2012–2014); a loyalty strategy with a conversational AI concept for Deutsche Telekom and T-Mobile (2018–2021).

Scenario · a typical first decision

“Which customer requests should an AI assistant resolve on its own, and what counts as resolved?”

Financial services

Clients
European Central BankBMW
What Sten did

Internet monitoring for the ECB (Gridpatrol, from 2001); a crypto-asset market study and risk assessment for the ECB and a digital-asset strategy for Solaris (2021–2022); customer experience for BMW Financial Services (2015–2017).

Scenario · a typical first decision

“Where may AI prepare decisions on documents, and who signs them off?”

Pharma and life sciences

Clients
BayerRocheBoehringer IngelheimMerck
What Sten did

Stakeholder analysis and social media intelligence for Bayer; EpiTrace.AI, AI with DNA-methylation analysis for research into early cancer detection (since 2019).

Scenario · a typical first decision

“Which research or medical-information workflow can AI speed up without weakening the evidence?”

Automotive and mobility

Clients
BMWMercedes-BenzAudiContinental
What Sten did

Customer-experience and social media strategy for BMW Financial Services in Los Angeles (2015–2017); internet monitoring for Daimler (Gridpatrol, from 2001).

Scenario · a typical first decision

“Which customer or dealer process benefits first, and how do we measure it?”

Consumer goods and retail

Clients
Aldi SüdCoca-ColaHenkel
What Sten did

Search marketing for Kraft Foods (mediatime, from 1997); internet monitoring for Aldi (Gridpatrol, from 2001); social media intelligence for consumer brands at ethority (2005–2021).

Scenario · a typical first decision

“Which service or content process can AI take on, at what cost per case?”

Travel, media and entertainment

Clients
Disney
What Sten did

Search marketing for Lufthansa (mediatime, from 1997); social media intelligence for AIDA Cruises (2012–2014); digital marketing strategy for Disney (2015–2017).

Scenario · a typical first decision

“Which guest or audience question should an AI answer, from which approved sources?”

Industry and energy

Clients
SiemensBoschZEISSHoneywellE.ONRWEDHL
What Sten did

These companies are among the clients of Sten Franke’s companies. For manufacturing and energy, ETHORITY publishes sector briefs in the Good AI Prism study series.

Scenario · a typical first decision

“Which service, incident or documentation process gives a measurable first result?”

Start

One task. One team. One decision.

Bring a real workflow, such as a service request or an investment decision. With the sponsor, the business team and IT, we clarify who decides, what costs time or money today and which test would justify an investment.

Sten Franke