0 Briefly introduce yourself:
With BiV for 5 years
BMW’s venture capital arm that invests in everything along the automotive value chain; from:
Vehicle itself, to production tech, to supply chain and enterprise software
We currently have more than $1Bn in AuM
300M Fund 3 Today actually
$1-10M checks
Backed >90 startups
1.) The Pivot/ Turnaround story: Could you each share one specific Valley of Death moment you've personally witnessed or navigated and the one surgical move that kept the lights on?
We see ups and downs working with teams
outside: looks easy:
See press release of excitign funding round
+Hopefully news of an amazing exit
nobody talks about between
Close to portco selling into automotive:
relationships take years to develop
pass extensive testing, safety, security
Company: passed all of it: BUT: long road with limited PoC revenue
Timelines kept pushin back
Need to scale up production capacity ahead of volume
Growht investors: want to see the revenue come in - we had the moment with last paycheck paid and 2 weeks before shut down:
secured a vehicle program in the nine-figure contract
multi-year, real volume, long program timeline
When contract comes: it comes big
=> Produt was working. Quality was there. In deeptech: Valley of Death is often one contract away. One contract later than your runway suggest
2) Navigating Investors and Corporates:
2 A) Does the Physical AI label actually help close rounds in heavy industry, or does it just import software-style expectations that hardware companies can't realistically meet? How do you tell a great capital partner from a tourist?
We’ve been investing in the space for long; long before the physical AI label was around
(+) Current developments interesting: Label has done:
sofware investors who never looked at robotics or industrial companies are now writing checks
BUT: also come with software-trained expectations on gross margins, time-to-revenue, deployment speed
What I see as the biggest miss: IMPLEMENTATION
Physical AI comes with real-world implementation
No click and distribute motion
Deploy on factory grounds, logistic hubs, in vehicles
Where is agreement: these industries represent a massive market opportunity
Disruption is real
greatest companies in Physical AI are yet to be made
My question mark: is the capital going to be there to get the amazing startups beyond valley of death?
2 B) What do you screen out in a first meeting or diligence calls, and how does the relationship change before versus after the check?
Understands unit economics and how those change at scale
Needs to know right off what the cost saving and ROI is for the end customer
Who are departments, customers, niches that will adopt fast because of a big pain point?
2 C) Are there things you should be looking for in strategics versus institutional investors?
Startegic: come in different shapes
some havea financial mandate (set up like a regular VC)
others more strategic; invest off balance sheet
Strategic investors can open doors other’s cant:
We position at business unit directly
as well as among decisoin makers and senior mgmt
But also: invest independent of corporate buy in and with financial returns
2 D) Are there any term-sheet trap doors that founders consistently miss?
Term sheets and deals can get complicated fast
Trap 1: Terms stick
What you accept in seed round carries through to next rounds
Senior liq pref in seed: A, B, C will demand the same
Trap 2: TS is a foundation for a long collaobration; you don’t end there and win or lose
Work just starts once the investment is wired
Trap 3: Look at combination of all terms
what happens to effectie price per share? Effective founder dilution?
Valuation is one thing;
what happens to ESOP increase, conversion of notes etc.?
3. Escaping Pilot Purgatory (Juan & Tom)
3 B) From the investor seat, which pilot metrics actually matter and which are vanity?
Early stage: great validation
Problem: many founder’s don’t have visibility on the actual process it takes on turning it to a real contract
People differe: production vs innovation teams
+ Founders burry non-converted pilot and never talk about those
4. The RaaS Debate (Jared & Thomas)
4 A) RaaS was supposed to be the holy grail. Is it still, or has it become a liability that drags valuations down and complicates exits? And where do venture debt or alternative structures actually fit in crossing the Valley?
RaaS = great if deployed and used correctly
But also: can be difficutl, costly, and neck breaking at scale
especially in terms of capital requirement to deliver now while revenue and cash collection in future
you are leveraging your own balance sheet to grow
when to deliver a large amount quickly: need much more capital to finance your deployment than you have
Need to know your paymant times and bound capital exactly!
Large enterprise: lower cost of capital - prefer capex over RaaS
Where RaaS (+): finance it with SPVs or debt
NOT EQUITY
Venture debt: lease-back models work great!
4 B) How does RaaS show up in valuation conversations and on the cap table at exit?
RaaS = traction easier to understand; BUT: not getting the same multiples on it at exit!
Buyers: often run sum-of-the-parts valuation
beginning: services and implementation heavy
then: hardware pay back
over time: software / Licensing revenue to stack up
Gross margin curve over time: negative to positive
4 C) Looking back, would you choose RaaS, CapEx sale, or hybrid again — and why?
I have bias for enterprise manufacutring:
Capex initially, with subscription on top; Not RaaS
4 D) Revenue quality versus revenue volume? What should founders optimize for in the first five million of ARR?
Most will say quality
My view is probably what you won’t execpt: show the $5M run rate as quickly as possible!
Find your wedge:
where to find customer groups / use cases that will convert fast
Fast bcs their pain is so big
Shiny logos: often take long
BUT: quality of course also matters: if all PoC= no go!
Honest take: teams underestimate how important it is to show commercial proof and trajectory
6. The AI Inflection Point (All Panelists)
6 A) Do you expect the rise of Foundation Models will actually shorten the Valley of Death by making robots more adaptable?
Foundation models will likely make the valley longer and harder to overcome:
but if you overcome it: ever more rewarding
With the developments: customers will increase the bar
VoD: in my view often a commercial problem
Pilot cycles will stay
Competition for deployments, integration and trust increases!
You won’t impress a client by just pilot + vision anymore
it’s now about deployments that work
and work at scale!
6 B) Are you interacting with foundation model companies today?
We are limited in our fund size and can write max $10M checks:
Reality is the investment sizes have moved well beyond where we play
6 C) In this new era, where does the durable moat actually live? Is it the model, the data, the embodiment, or the deployment layer?
bottleneck: proprietary data from real-world use caes
But: if you solve a real pain point:
customers will be willing to open up their data
Also think: is the data you collect part of your customers moat
or will it just make them more efficeint and reduce cost
Outro: What’s the one thing you wish every founder in this room knew before they spent the first dollar on a prototype?
If you are about to raise venture money: the clock is ticking
its not about building the best product; but
crossing the right milestone before you run out of cash
You have to reach technical proof points and commercial proof points year after year
Every dollar you spend on a prototype, needs to translate into a milestone that unlocks your next dollar to grow
Plan your milestones backward
financing risk kills most companies!
Think about what your dollar spent on a prototype will actually unlock in terms of insights, tech de-risking, cost of manufacturing etc.
Zuletzt geändertvor einem Monat