Building a €220M pipeline engine across twenty European markets

The situation

Watson AI was a global product with a global narrative. My mandate as EU Marketing Director was to turn that into pipeline across more than twenty European markets, running a €55M budget and a 160-person organisation with twelve direct reports.

The structural tension in any multi-market brief is the same. Push the global strategy down unchanged and it fails locally: the wrong buying culture, the wrong channels, the wrong references, the wrong language for how the problem is described. Let each market adapt freely and you lose the ability to compare anything — twenty local strategies, twenty definitions of a lead, and a group number assembled from parts that do not add up.

The diagnosis

Markets were not failing because their tactics were wrong. They were failing because each had quietly built its own vocabulary. A qualified lead in Germany was not a qualified lead in Spain. Pipeline stages carried different meanings. When roll-up reporting added those together, the group figure was arithmetically correct and commercially meaningless.

The second issue was allocation. Budget had accumulated historically — markets held what they had held the previous year, adjusted for politics. It bore no relationship to where opportunity actually sat.

What I did

1. Centralised the definitions, decentralised the execution

One definition of a qualified opportunity. One set of pipeline stages. One reporting standard. These applied in every market without exception and were not open to local interpretation.

Everything downstream stayed local: channel mix, creative, language, partner selection, event strategy, the specific proof points used with a given buyer. Local teams got genuine autonomy over how they generated pipeline and none at all over what counted as pipeline.

2. Reallocated budget against opportunity, not history

Budget moved to where pipeline could actually be built — addressable opportunity, sales capacity to absorb it, and demonstrated conversion. Some markets gained substantially. Some lost. Both decisions were defensible because both were made against the same evidence, and the evidence was now comparable across markets for the first time.

3. Built the SMB motion as a separate engine

Enterprise and SMB were being treated as one funnel with different deal sizes. They are not. Enterprise buying is a long, multi-stakeholder, relationship-led process suited to account-based marketing. SMB is a shorter, more self-directed decision suited to volume demand generation and product-led entry points.

I separated them entirely — different teams, different targets, different channel economics. The SMB platform reached €30M ARR by year three and grew the SMB customer base by 73%.

4. Made pipeline the only number that counted

Marketing at enterprise scale generates an enormous quantity of measurable activity, most of which explains nothing about revenue. I reduced the reporting to pipeline generated, pipeline converted, and cost per qualified opportunity. Every other metric became diagnostic — useful for working out why a number moved, never presented as the number itself.

The results

What transfers to smaller businesses

The scale is not the transferable part. The structure is.

• Standardise definitions before you standardise anything else. A company operating in three markets has the same problem as one operating in twenty, just with less noise hiding it.
• Segments with different buying behaviour need separate engines. Enterprise and SMB, or supply side and demand side, cannot share a funnel and be measured honestly.
• Historical budget allocation is not a strategy. If you cannot explain why a market or channel holds the budget it holds, in terms of opportunity rather than precedent, it is allocated wrong.
• Reduce the reported metrics until only revenue-linked ones remain. Everything else is diagnostic, and mixing the two is how boards lose confidence in marketing.