GridForge - WFW VC Teardown #04

B2B SaaS · Seed · Climate Infrastructure · Autonomous Energy Orchestration · Evaluated by James Palmer, Blackbird Ventures

Verdict

What GridForge Does

GridForge helps electricity network operators manage growing grid pressure – from EV charging spikes, batteries, renewables, and data centres – by coordinating distributed energy assets in real time using software, instead of waiting years for new physical infrastructure to be approved and built.

Think: air traffic control for electrons - a system that sees where congestion is building across the grid and automatically shifts load before anything breaks, without a human needing to make each decision.

Raising: $6M USD seed · Founder: Samir Patel

Interested in learning more about GridForge?

GridForge Intro Email

3 Things That Could Kill This

1. The first customer choice doesn't make sense

GridForge's proposed wedge is mid-sized utilities managing EV adoption spikes – specifically those wanting to defer substation upgrades. The problem: electricity network operators are typically regulated monopolies. They are hard, slow customers with long procurement cycles. More importantly, the more capital infrastructure they have in the ground, the more money they make under a regulated return model. Deferring a $10M substation upgrade is not compelling to a customer who profits from building it.

Why this matters: The quality of a founder's first customer choice is a direct signal of their commercial intelligence. Choosing a customer who is structurally disincentivised to buy the product is not a mistake that gets fixed later, it reflects a fundamental misread of the market.

2. The "why now" doesn't pass the sniff test

The technology GridForge describes, reinforcement learning, real-time dispatch, distributed asset coordination, is not new. Grid orchestration has been attempted for years. The pitch presents a list of constraints the grid faces, but doesn't explain what has specifically changed in the last two to three years that makes this company possible today. Is it a regulatory shift? A technology unlock? A behavioural tipping point among utilities? Without a clear "why now," the natural investor question is: has this been tried and failed, or is it simply not a real opportunity?

Why this matters: A "why now" is the force that clears a path for a new company to exist. Without one, incumbents and the absence of prior winners both become evidence against the opportunity, not just context around it.

3. This is a network effects business with a nasty cold start problem – and the pitch ignores both

GridForge's moat compounds with every connected asset, every congestion event, and every market integrated. That is a long-term advantage. It is also, by definition, a two-sided network with a painful cold start. The pitch doesn't acknowledge this tension or explain how the first customers get recruited before the network has value. Skipping this question doesn't make it disappear, it makes investors wonder whether the founders have thought through the hardest part of building this business.

Why this matters: The cold start problem in network-effects businesses is where most of them die. The best founders have a specific, credible answer to how they get the first nodes on the network. Silence on this point is not reassuring.

Why This Might Actually Work

1. The team is the strongest argument for a meeting

Samir Patel spent years modelling national transmission upgrades as a grid systems engineer. Mei Chen completed a PhD in distributed systems optimisation and worked on DeepMind's energy optimisation team. Their advisory board includes a former national grid operator CTO. This is a team with genuine, hard-to-replicate unfair advantage in exactly the technical and operational domain this problem requires.

2. The traction is real for a business this hard to start

Three paid pilots, 42MW under orchestration, $420k ARR with early contracts converting to multi-year, and a demonstrated 9% peak load reduction is meaningful progress for a company operating in one of the hardest sales environments in technology. Utility procurement is slow, cautious, and bureaucratic. The fact that GridForge has paying customers at all is signal worth taking seriously.

3. The underlying market problem is genuine

Electrification is accelerating. Physical grid upgrades take 5–10 years and cost billions. The gap between the pace of electrification and the pace of infrastructure delivery is real and widening. Whether GridForge is the company that solves it is an open question – but the problem itself is not in doubt, and that matters as a starting point.

4. Incumbents are structurally slow

Siemens, GE Grid, and Schneider Electric sell hardware-heavy enterprise contracts with innovation cycles measured in years. A software-native, learning system that improves with every connected asset is genuinely difficult for them to replicate without cannibalising their existing business model. That is a real, if impermanent, structural advantage.

What the Founder Should Do Next

• Identify a better first customer → Regulated utilities that profit from infrastructure spend are the wrong wedge. Find a customer whose P&L is directly damaged by grid constraints and who has no regulatory reason to prefer physical infrastructure – large industrial energy users, data centre operators, or EV fleet managers facing charging costs are all more promising starting points. The first customer choice should demonstrate commercial intelligence, not just technical accessibility.
• Answer the "why now" in one sentence → What has changed in the last two years that makes GridForge possible today? If the answer is the DeepMind-trained reinforcement learning model that Mei Chen brings, say that explicitly. If it's a specific regulatory change in the NZ or Australian energy market, name it. This single sentence needs to be in the first paragraph of every investor conversation.
• Translate MW into dollars → Every metric in the pitch is expressed in megawatts. Investors think in dollars. What does 42MW under orchestration generate in ARR? What does a utility save per MW of peak demand reduction? What is the payback period for a customer? Build the unit economics model and lead with it.
• Confront the cold start problem directly → Explain specifically how GridForge recruits the first ten customers before the network has enough data to be self-evidently valuable. This is the hardest commercial question in the business – answering it proactively signals exactly the kind of commercial thinking that turns a technical team into a fundable company.
• Recalibrate the Series A milestone → $2M ARR is no longer sufficient to unlock a strong Series A round unless the growth slope is exceptionally steep. Reframe the milestone around the network metric that actually demonstrates compounding value – MW under management, number of integrated asset types, or demonstrated congestion reduction at scale across multiple markets.

What This Teardown Teaches

Pattern 1 – First customer choice is a commercial IQ test

Investors evaluate early-stage companies on the quality of the founders' thinking as much as the quality of the product. Choosing a first customer that is structurally disincentivised to buy – or one that is simply the most obvious rather than the most winnable – signals a gap in commercial intelligence that compounds as the company scales. The best founders choose their first customer with the same rigour they apply to their technology.

Pattern 2 – "Why now" is not optional

Every mature industry has smart people who saw the opportunity years ago. If those people didn't build this company, investors want to know why you can build it today. The answer is the "why now" – a specific technology unlock, regulatory shift, or behavioural change that has cleared a path that didn't previously exist. Without it, the absence of prior winners reads as evidence against the opportunity.

Pattern 3 – Market sizing must connect to your actual product and customer

"We can capture a fraction of a $20 trillion infrastructure market" is not market sizing – it is arithmetic on a number that has no relationship to what GridForge sells or who buys it. Real market sizing starts with a customer, a price point, and a count of reachable buyers – then builds up. For a network-effects business in particular, the relevant question is not what fraction of the market you can capture. It is who wins the market, and why that is you.

30-Minute Meeting?

Yes – on the strength of the team

The pitch has real commercial problems that need to be fixed before the next investor conversation. But the team – a national grid engineer paired with a DeepMind-trained distributed systems PhD and an advisory board with operational grid experience – is singularly credible for this problem. In early-stage investing, people outweigh ideas. The meeting is worth taking to assess whether the commercial thinking can catch up to the technical capability.

Full VC Notes

James Palmer

Connect with James on LinkedIn

Principal at Blackbird Ventures, a leading Australasian VC firm with over $7B under management.

He invests in ambitious Aussie and Kiwi founders from the earliest stages, often before the idea has a name. James has led investments into Cotiss, Ivo, Kwetta, Nextwork, Vertus Energy and Watchful and supported Blackbird’s bets on OpenStar and Ternary Kinetics.

Too much jargon. Clarity of the value proposition is lost.

Talks about value in MW terms rather than dollars – if possible convert to a metric where the financial value is obvious. It’s not clear that managing 42MW of capacity is a profitable exercise?

Traction is good enough to warrant

This doesn’t articulate a problem instead it outlines constraints. Who is the customer that feels the urgency of these constraints. Problems at a market level are too fuzzy because they’re not linked to a P&L or an employee trying to be promoted. Good problems are painful and urgent.

“Air traffic control for electrons.” This is a good analogy.

This says what it does but it lacks an earned insight; where is the magic? What secret did you discover that lies at the heart of this product? Every great start up attacking a mature industry has specific counter positioning and that isn’t clear here.

I also read this and think… man this is complex. In this value chain you have a grid owner/ operator, B2B customers, Generation, Retailers, consumers etc. You have behaviour and operational changes implied by the dispatch / throttling of demand. You have regulated and commercial incentive models.

Do you need to win everyone over to win?

it’s not clear this passes the ‘why now’ sniff test.

This looks / feels like traditional ML. What is the catalyst that clears a path for building out the network?

This is a network effects business which is always two sided… (1) long term moats; traded off against (2) nasty cold start problem.

This is solid traction if they’re less than 2 years old – for a company that will be inherently hard to get going.

BUT… it’s all pilots and all ANZ… this might be a blocker for us since networked businesses are inherently locally constrained.

I don’t think the wedge makes sense – substations are typically regulated monopolies and they’re hard customers. Capex referrals are not compelling because the more capex they have in the ground the more money they make.

I really care about who is a great first customer and this isn’t clear.

If one of the drawbacks of this business is a “nasty cold start problem”, how is it realistic of an investor to place such emphasis on first customers so early in the journey? Won’t the founder ‘figure it out over time’?

It is important because it expresses the quality of the founders commercial thinking. The best founders are the best learners, if they choose a ‘bad’ first customer it would erode my confidence in their ability to conquer the cold start problem which is all about understanding the ‘hard’ part of the market to get participate.

The market isn’t infrastructure. It is some fraction of this. Market sizing need to be tethered to the value of your product and the customers you can sell it to. Rather than the overarching market for infrasturcture.

Another pet peeve… I can capture a small portion of a big market and I’ll be awesome… that’s explicitly not true for network effect business that invite competition where the market structure should be winner takes most (at least regionally)

Is the following simplification fair?

1. Your market size is inflated.
2. You didn’t show how the pricing connects to the market size.
3. “We only need 1% of a big market” is weak logic.
4. If this is a network-effect business, the real question is who wins – not who gets a small slice.

James- sort of..it might read:

1. The market is mischaracterised – you do not replace the infrastructure.
2. Market sizing should connect pricing (and therefore customer value) with the customers you sell to.
3. 1% of a big market never works and is weak logic.
4. Yeah nice.

This shows only old school hardware as competition. Either that is true and it’s probably a bad market no one else believes in or it’s a dishonest presentation.

Can you break that comment down in its two parts?

This highlight traditional power electronics customers are competition. This makes me think one of two things:

1. The market is not really there. There is no observable ‘why now’ which makes me think this has been tried and failed or it’s not really an opportunity.
2. The founders aren’t being honest about competition from other start ups. I actually like an honest assessment of competition and a low level description of what they are getting wrong and why – this instead tries to avoid it.

Team looks strong for what they’re doing at a technical level. Probably singularly enough to take a meeting. Though pitch seems commercially naive.

Milestones are miscalibrated. $2M of ARR no longer enough to unlock a great round unless the slope of growth is

Overall

Good team. Very fuzzy on the commercial proposition – so it is a bad write up.

I am generally interested in unlocking capacity on the grid and finding commercial models to unlock value. They seem very weak on the second half of this but it’s worth a meeting.

30-Minute Meeting?

Yes, based on team and my interest in the overarching market. People > Ideas at the start.

Quick Navigation:

Back to VC Teardowns | How NZ VCs Evaluate Startups

🚨 Disclaimers: All startup companies, business models, products and founders described in VC Teardowns are fictional and created solely for educational purposes. Any resemblance to actual companies, persons or events — past, present or future — is purely coincidental. The opinion of each participating VC reflects their individual perspective and does not represent their firm as a whole. A teardown should not be treated as a universal rulebook. Founders are encouraged to engage investors early and build relationships — early conversations are often exploratory, not evaluative.