TobiasConrad

Don’t bend the tool to the mess. Fix the mess and adopt the standard.

Companies keep buying CRM and AI to fix problems that are not technical. Others take a proven, global system and customize it until it mirrors every quirk of a broken organization. Both fail the same way.

A standard system already carries the experience of thousands of companies and years of refinement. Bend it to preserve unclear decision rights, broken hand-offs and legacy complexity, and you do not get insight. You get an expensive monolith that no one can maintain. I have seen exactly this happen, and seen what it costs to undo.

So the work is never the tool. It is fixing the organization, the process and the data so a standard system can do what it was built to do. Get that right, and the technology finally earns its price.

Read and rebuild, in one pass

There are two ways to fix a broken commercial organization, and each is half an answer.

I do both in one pass.

I read how the organization actually works, who really decides, where the hand-offs break, which people carry the business and which block it. At the same time I read the commercial engine, where the pipeline leaks, whether there is a single source of truth, and the one metric that moves the model. Then I rebuild both together, because they are one system. Decision rights and people on one side, pipeline, data and forecast on the other. Wire them separately and one half always undoes the other.

That is the difference. Not a faster diagnosis and not a harder cut, but an operating model that holds because the organization and the revenue engine were rebuilt to fit each other. The result is a platform that runs, scales and can be handed over without me.

Four movements, fast and in sequence

Twenty years where these problems show up first

I have built and fixed commercial organizations across digital platforms, AdTech and media, the field where these problems appear first and hardest, at transaction speeds and product complexity most industries never see. Whoever can industrialize that can do it almost anywhere.

The range runs from leading five people to leading more than three hundred, across PE-backed, family-equity and listed structures, with P&L responsibility up to nine figures. Inside a Waterland-backed platform I consolidated a fragmented commercial stack onto one system and one source of truth. As founding CEO of a media joint venture I led a turnaround through to merger clearance, and saw firsthand how a global standard system, customized to preserve every legacy quirk, had become an expensive monolith no one could maintain. Fixing the organization first, and returning the tool to standard, was the way out.

I have worked with Salesforce since 2005, as an operator who carried the number, not as a consultant. That is rare, and it is the difference between knowing the tool and knowing what an organization has to look like for the tool to work.

Selected transformations

Range

Leadership Principles, Harvard Business School · Agile Management, University of Maryland · MBA, University of Hamburg

The part a US playbook cannot copy

In Switzerland, Germany and Austria, putting AI on customer data and conversations is not a feature. It is a legal act.

Recording a call without the express consent of everyone on it is a criminal matter, analyzing it goes further still, and under the revised Swiss data protection law the personal liability can sit with the executives, not the company. Most playbooks are imported from markets where none of this applies.

I build the commercial model and the consent and governance layer together, so the organization gets the upside of AI without the exposure. It is the part a US playbook cannot copy, and the part a Swiss board cares about most.