Permitting and grid-capacity OS for European AI campuses and factories racing to secure bankable megawatts.

1. Europe-wide electrification targets are raising the volume of new power-load projects that need to be permitted and connected.
2. Repeated energy price shocks make faster grid access economically urgent rather than merely operationally convenient.
3. Permitting delays are already publicly identified as a competitiveness bottleneck, creating urgency for workflow tools instead of more advisory labor.
4. AI data-center demand is increasing competition for the same scarce grid-ready sites, which raises willingness to pay for faster site qualification.

Catalyst. Norrsken's 50% electrification push explicitly names permitting delays, high power prices, and rising AI data-center demand, turning grid-readiness from an engineering detail into an immediate growth blocker.

The product is a workflow system for project developers to go from target geography to submission-ready power plan. It ingests site attributes, tariff assumptions, local planning constraints, and project load shape, then recommends a bankable configuration: utility interconnection path, storage or flexibility add-ons, and a sequenced permit checklist. It auto-builds the evidence pack for utilities, municipalities, and internal investment committees, so teams stop paying consultants to recreate the same package from scratch on every site. Over time, the company builds a proprietary dataset of which site profiles, application structures, and flexibility designs actually get approved, creating a feedback moat around permit success by megawatt and geography.

What's different. Most incumbents attack this problem as bespoke advisory work after a site is already chosen. This company wedges in earlier with software that helps developers decide where to pursue capacity at all, then turns the messy interconnection-plus-permitting process into a reusable system of record. The moat comes from approval outcome data, jurisdiction-specific templates, and design recommendations that connect grid access with storage and flexibility choices rather than treating them as separate workstreams.

Jobs to be done

flowchart LR
  Buyer[Developer or energy lead] --> Pain[Uncertain grid access and permit delays]
  Pain --> Product[Grid-readiness workbench]
  Product --> Outcome[Faster approved megawatts and lower project risk]

Executive takeaways

• The scarce asset is not generic clean power; it is bankable, permitable megawatts at specific sites. JLL shows FLAP-D vacancy at 6.3% and connection lead times up to 10 years, while Ramboll says interconnection queues beyond five years are pushing land banking and pre-powered land. [3][29]
• Policy is moving toward the thesis, not against it: the EU wants to at least triple data-centre capacity in 5-7 years while simultaneously tightening data-centre reporting and speeding grids, storage, PPAs and permitting. [18][11][13][16][17]
• The workflow is still fragmented and services-heavy. Paces is the closest software analogue, but it is US-centric and broader than Europe-specific load-side permitting; Ramboll shows how much value still sits inside bespoke due diligence and permit strategy. [19][20][21][28][29]
• Nordics and the Netherlands are credible beachheads. Fingrid says Finland already has 100 pending data-centre projects, while Dutch industry reports frame the market as having moved from boom to bottlenecks. [10][4][5]
• Willingness to pay should exist because the avoided cost is large: powered-land premiums, dead-end site-control work and consultant cycles dwarf a six-figure workflow subscription if the product kills bad sites early or shortens submission cycles. [3][21][29]
• The main venture risk is not demand creation; it is execution. The startup must win earlier in the workflow than consultants and then compound local approval data fast enough to become a system of record rather than a nicer study generator. [15][19][28][29]

Market definition

This market is narrowly defined as workflow software for large-load project developers that need to qualify a site, model path-to-power options, coordinate interconnection inputs and generate submission-ready permitting packages for 10-100 MW projects in Europe. It includes AI campuses, colocation developments and adjacent power-intensive sites where grid access and local approvals are gating; it excludes generic ESG reporting, utility asset management, and purely upstream renewable project origination tools. [3][18][28][29]

Customer and buyer

Initial users are development managers, energy-infrastructure leads and origination teams at European colocation and AI-campus developers. The economic buyer is typically the VP/Head of Development or country GM who owns site control, utility engagement and internal investment approvals. The painful jobs are screening sites before competitors lock them up, proving a credible path to power, and assembling a board- and regulator-ready package without waiting on slow consultant cycles. [3][19][21][28][29]

Buying triggers

• Anchor-tenant LOI, expansion mandate or land-option deadline forces a developer to prove path-to-power quickly. [3][19][29]
• Grid scarcity makes pre-commitment the default in major markets, pulling diligence earlier in the funnel. [3]
• New or faster permitting windows around grids, PPAs and storage make earlier submissions economically valuable. [13][16][17]

Willingness to pay

There is real budget pressure behind this problem because developers already pay indirectly via powered-land premiums, consultant-led feasibility studies, optioned-but-unbuildable sites and delay against scarce grid capacity. JLL and Ramboll both show that speed-to-power now changes land economics, while Paces markets faster modular diligence as a direct alternative to slow expert studies. [3][21][28][29] [3][21][28][29]

Category dynamics

Growth signal ~20% implied annual EU data-centre capacity growth if the EU triples capacity over 6 years

Validation signals

• The EU now explicitly targets a tripling of data-centre capacity within 5-7 years.
• FLAP-D vacancy has fallen to 6.3% and 83% of pipeline is pre-let, indicating genuine scarcity.
• Finland reports 100 pending data-centre projects and plans more than 4,000 MW of added southbound transmission over eight years.
• Dutch market reports have shifted from growth framing to bottleneck framing, matching the startup thesis.
• Piclo’s DSO/TSO profiles show flexibility markets expanding into outage management and local constraints, increasing the value of integrated load-flex planning.
• Paces is shipping data-centre-specific and fresher-grid workflow features, validating that buyers want software, not just studies.

Regulatory & technical constraints

• Grid-connection rules, thresholds and feasibility depend on local TSO/DSO conditions, so no single Europe-wide data model is enough.
• Data-centre reporting and rating rules raise the evidence burden on energy, water and renewable sourcing.
• Useful recommendations increasingly depend on linking interconnection, PPAs, storage and flexibility rather than treating them as separate workstreams.
• Outputs must be explainable to internal committees, consultants and authorities; black-box scoring will not be enough for live permitting.
• Local community acceptance, water access, noise and land-use constraints remain outside the direct control of a software vendor.

The direct-software field is still relatively thin. Paces is the closest product analogue on siting, due diligence and path-to-power workflows; Neara is a utility-side digital twin that can answer network-design questions but is not built as a developer system-of-record; Piclo owns flexibility-market procurement and DSO/TSO distribution, not site selection or municipal permitting; Ramboll represents the default incumbent: high-end services that span site selection, due diligence, design and permitting. In practice, the dominant substitute remains a patchwork of consultants, utility interactions and internal GIS models. [19][20][21][22][23][24][27][28][29][30]

Why incumbents do not win by default

• Engineering consultancies. Consultancies do not win by default because they monetise bespoke effort after a site is already in motion; a developer-side workbench can move earlier, kill bad sites faster and retain learnings across the portfolio instead of restarting each study. [21][28][29]
• Utility digital-twin platforms. Utility-side tools such as Neara help network owners model resilience and design, but they are not designed as buyer-side permitting CRMs for colocation developers managing site options, municipal workflows and internal investment committees. [27][30]
• Flexibility and market-access platforms. Piclo can help unlock flexibility contracts and visibility into constrained zones, but it solves monetisation and dispatch of flexible assets rather than the full site-ranking and permit-packaging problem. [22][23][24]
• In-house spreadsheets and GIS stacks. Developers can assemble internal workflows, but Paces’ own category framing shows that disconnected tools, stale models and sequential diligence are the core failure mode; a product can win if it materially reduces time-to-no-go and time-to-submission. [19][20][21]

This company is a Europe-focused workflow platform for developers that need to turn a candidate site into a permitable, grid-connectable 10-100 MW load project before competitors consume scarce capacity. The initial customer is a colocation or AI-campus development team in the Nordics or Netherlands that has an anchor tenant, land option, or expansion mandate and must prove path to power on a board-level timeline. The product wedge is not generic grid intelligence; it is a system that ranks sites, models interconnection plus storage or flexibility options, and generates submission-ready utility and municipal packages. The strategic logic is to win earlier than consultants in the site-screening workflow, then retain the project as the system of record through submission and approval. Research supports the pain, timing, and policy tailwinds, but not yet the buying behavior; the most important unknown is whether developers will pay for software before they have already hired consultants. The beachhead market is credible but narrow, so venture scale depends on proving repeatability in 2-3 jurisdictions and then expanding the same workflow into factory electrification and adjacent flexible-load projects. The company should therefore raise a pre-seed round to validate the 48-hour no-go workflow, convert design partners into annual contracts, and build jurisdiction-specific approval data rather than expand horizontally too early.

Problem

• Large-load developers in Europe lose months on candidate sites that never secure grid capacity or local permits.
• Today's workflow is fragmented across consultants, utilities, municipalities, and internal spreadsheets, so teams discover fatal issues too late and pay repeatedly for bespoke studies.

Solution

• A grid-readiness workbench that scores candidate sites for path-to-power viability across grid access, planning constraints, and optional storage or flexibility design.
• Submission workflow software that turns project inputs into reusable utility, municipal, and investment-committee evidence packs instead of bespoke consultant decks.

Why we win

• We enter earlier than advisory incumbents by helping developers decide where not to spend time before site control hardens.
• Product defensibility compounds from approval-outcome data, jurisdiction templates, and design recommendations that link interconnection with BESS and flexible-load choices.

Milestones

flowchart LR
  Wedge[Beachhead wedge] --> MVP[MVP]
  MVP --> Proof[Proof points]
  Proof --> Expansion[Expansion motion]
  Wedge --> Data[Jurisdiction templates and approval data]
  Data --> Proof
  Proof --> Moat[System of record and benchmark moat]
  Moat --> Expansion

Founding team

Experiment roadmap

Risk assessment

What must be true

• At least half of interviewed colocation developers must confirm they can authorize workflow spend before full consultant engagement on a live project.
• The product must eliminate or deprioritize bad sites within 48 hours with acceptable accuracy on at least 10 back-tested projects.
• At least 60% of paid pilots must convert into annual contracts because the software stays in use beyond initial screening.
• Two initial jurisdictions must show enough process repeatability for templates to be reused across at least 5 projects each.
• At least one adjacent market beyond colocation must show the same workflow can command similar ACV without a full product rewrite.

Open diligence questions

• Who currently owns the budget line before site control: development, energy infrastructure, or external advisers?
• What minimum dataset is needed for a credible no-go decision in Finland, the Netherlands, and Germany?
• How much of the workflow can remain software-led before customers demand bespoke engineering sign-off?
• Which part of the process actually drives willingness to pay: site ranking, submission packaging, or internal approval speed?
• Do adviser and EPC partners accelerate distribution, or do they pull the company into low-margin services delivery?

Model sanity

• Revenue engine. Base-case revenue is driven by reaching 11 paid regional accounts by Y3 at a blended $240K annual account value tied to live project workflows.
• Must go right. The company has to win budget before or alongside consultants so founder-led pilots convert into annual software contracts fast enough to justify the Y2 sales hires.
• Model breaks if. A one-quarter sales-cycle slip or services-heavy delivery keeps Y3 revenue under $1.8M and pushes the cash low point toward the downside case.
• Next-round proof. Converting at least 2 pilots into annual contracts across 2 jurisdictions is the proof point that supports a seed raise for Germany and adjacent industrial expansion.

Scenarios

Sensitivity

flowchart LR
  TargetAccounts --> QualifiedLiveProjects
  QualifiedLiveProjects --> PaidPilots
  PaidPilots --> AnnualAccounts
  AnnualAccounts --> Revenue
  Revenue --> GrossProfit
  GrossProfit --> Cash

Flags: Gross margin only reaches the BP target if partner-delivered edge cases and bespoke engineering stay outside the core team. · The model assumes buyers pay before full consultant engagement; if that gating assumption is wrong, both CAC and sales cycle worsen quickly. · Beachhead revenue concentration is high because the first 8-11 accounts are all tied to large-load development budgets in a few European jurisdictions. · Cash turns negative without a pre-seed close, so fundraising timing is itself an operating dependency rather than a background assumption.

• Data availability. Utility queue, hosting-capacity, and local permitting data may be fragmented or hard to standardize across jurisdictions. Mitigation: Start in a small number of geographies with repeatable workflows and combine public data with customer-supplied documents and partner integrations.
• Services creep. Early customers may demand bespoke modeling and permit support that turns the company into a consultancy. Mitigation: Productize every recurring deliverable into templates, scoped modules, and partner-delivered service layers outside the core software team.
• Long enterprise sales cycles. Data-center and industrial developers may buy only when a live project reaches a gating event. Mitigation: Sell into concrete project milestones, price by active project, and use advisory and EPC partners to enter already-funded developments.