The distinction between a milestone being achieved and the underlying economic right actually being secured is particularly important. Investors spend a lot of time diligencing whether a company can execute its roadmap, and perhaps not enough asking whether the external system will still assign the same value to each milestone when it gets there.
The two halves arrive on completely different schedules. The company tells you how execution is going every month, because someone is paid to tell you. Nobody tells you when the world outside stopped valuing that milestone the same way. There's no report and no meeting for it. What I've settled on is writing down, at entry, the two or three outside conditions the case depends on, and what would have to show up in a filing to prove them wrong. Without that, there's no moment when anyone looks. You find out when it's already in the price.
“A change in meaning is easier to miss because the vocabulary remains intact” is probably the key insight here. This extends well beyond data centres: permits, certifications, customer commitments, manufacturing capacity and even “commercial validation” can all retain the same label while becoming materially weaker evidence.
Certification might be the sharpest version. The certificate stays valid. What changes is how narrowly it gets read. Nothing expires, nothing gets reissued, so there's no event to spot. You usually find out when a customer's own approval process turns it down, which is years after the milestone went into the model as done.
This feels especially relevant to deep tech because so much enterprise value sits several milestones into the future. In software, an external constraint can often be worked around relatively quickly. In industrial, energy or biotech companies, a change in regulation, infrastructure availability or qualification standards can rewrite several years of the roadmap at once!
And it compounds. In software the fix for a constraint is usually separate from what caused it, so you route around it in a quarter. In industry the fix needs the same permit, the same queue, or the same supplier that caused the problem. The fix inherits the delay. That's why one regulatory change can move several years of roadmap at once instead of one milestone.
The distinction between a milestone being achieved and the underlying economic right actually being secured is particularly important. Investors spend a lot of time diligencing whether a company can execute its roadmap, and perhaps not enough asking whether the external system will still assign the same value to each milestone when it gets there.
The two halves arrive on completely different schedules. The company tells you how execution is going every month, because someone is paid to tell you. Nobody tells you when the world outside stopped valuing that milestone the same way. There's no report and no meeting for it. What I've settled on is writing down, at entry, the two or three outside conditions the case depends on, and what would have to show up in a filing to prove them wrong. Without that, there's no moment when anyone looks. You find out when it's already in the price.
Thanks Eden for sharing your insights and ongoing work!
“A change in meaning is easier to miss because the vocabulary remains intact” is probably the key insight here. This extends well beyond data centres: permits, certifications, customer commitments, manufacturing capacity and even “commercial validation” can all retain the same label while becoming materially weaker evidence.
Certification might be the sharpest version. The certificate stays valid. What changes is how narrowly it gets read. Nothing expires, nothing gets reissued, so there's no event to spot. You usually find out when a customer's own approval process turns it down, which is years after the milestone went into the model as done.
This feels especially relevant to deep tech because so much enterprise value sits several milestones into the future. In software, an external constraint can often be worked around relatively quickly. In industrial, energy or biotech companies, a change in regulation, infrastructure availability or qualification standards can rewrite several years of the roadmap at once!
And it compounds. In software the fix for a constraint is usually separate from what caused it, so you route around it in a quarter. In industry the fix needs the same permit, the same queue, or the same supplier that caused the problem. The fix inherits the delay. That's why one regulatory change can move several years of roadmap at once instead of one milestone.
Great article! Thx for sharing your insights
Thank you Jenny, glad it was useful.