August 2, 2025
Why classical SaaS metrics fail in DeepTech and the GPT Moment for Material Informatics 3.0
Last week, we saw exciting developments in the DeepTech space. The strong momentum in DeepTech led Bavaria to claim Germany’s position as the #1 destination for venture investing, overtaking Berlin. An initiative in the Bavarian parliament to allow foundations to invest in VC funds is expected to further accelerate this trend. | |||||||||
On our fund level, our portfolio company ExoMatter launched a new feature with groundbreaking potential. Susanne wrote an article putting common misconceptions about DeepTech and ExoMatter’s new feature launch into perspective: | |||||||||
| |||||||||
When evaluating DeepTech startups, applying classical SaaS lenses often leads to wrong conclusions – and missed opportunities. | |||||||||
Let’s look at the facts. | |||||||||
| |||||||||
In a traditional SaaS business, growth is usually benchmarked via MRR increase, customer acquisition cost to annual contract value ratios, churn, and net revenue retention. The holy grail: repeatable, scalable Go-to-market playbooks and self-serve customer journeys. | |||||||||
These metrics work well when products solve horizontal problems in a standardized way – think marketing automation, CRM, or collaboration tools. | |||||||||
But DeepTech is different. | |||||||||
These companies operate at the frontier of science and engineering. Their customers are often R&D teams, labs, or industrial innovation units. Their sales cycles are longer, integration efforts are higher, and the level of required domain understanding is very different. | |||||||||
The result? | |||||||||
Revenue ramps more slowly – especially in early phases, as validation, regulatory approvals, or manufacturing compatibility take time.Customer collaboration is more intensive and less scalable – as projects are often co-developed and highly specific.Products are much “stickier” – once adopted, they are deeply embedded in the customer’s core IP, workflows or materials stack.The moat is not just product, but knowledge + trust + process alignment – which takes time to build, but is harder for competitors to disrupt. | |||||||||
For classical SaaS investors, this can feel uncomfortable. Product-market fit in DeepTech is typically not expressed in short-term MRR graphs, plug-and-play churn dashboards or viral product-led growth. The familiar indicators of traction and repeatability don’t always show up early – even when a startup is solving a mission-critical problem. | |||||||||
Yet this discomfort says more about the metrics than the market. | |||||||||
As McKinsey and BCG both point out in recent reports, scaling DeepTech startups requires new mental models and tailored operating frameworks. | |||||||||
One of the biggest bottlenecks we see across DeepTech startups: | |||||||||
How to scale deep customer collaboration without scaling headcount accordingly? | |||||||||
That’s where it gets interesting. | |||||||||
Let’s take the example of Exomatter, one of our portfolio startups developing AI tools for the discovery of next-generation sustainable materials – a field traditionally dominated by manual simulation and trial-and-error in the lab. | |||||||||
The problem: every industrial use case – from packaging to batteries to semiconductors – requires specific material properties and context. Onboarding a new customer can be as complex as a full R&D project. | |||||||||
Their solution: Exira – an AI agent purpose-built to scale deep technical collaboration. It guides users through discovery processes, contextualizes Exomatter’s models, suggests new research paths, and adapts its guidance to specific customer targets and constraints. | |||||||||
Instead of hiring a large technical services team, Exomatter has launched a scalable, intelligent interface that understands both the customer’s goals and the physics behind the product. | |||||||||
A promising step toward making DeepTech more scalable – without compromising its depth. | |||||||||
Have a great week! | |||||||||
PS: Wait …. Let us not forget this important development above: DeepTech’s momentum led Bavaria to surpass Berlin as Founders Capital #1. Read more in the Spiegel Article below (in German). |
Last week, we saw exciting developments in the DeepTech space. The strong momentum in DeepTech led Bavaria to claim Germany’s position as the #1 destination for venture investing, overtaking Berlin. An initiative in the Bavarian parliament to allow foundations to invest in VC funds is expected to further accelerate this trend. | |||||||||
On our fund level, our portfolio company ExoMatter launched a new feature with groundbreaking potential. Susanne wrote an article putting common misconceptions about DeepTech and ExoMatter’s new feature launch into perspective: | |||||||||
| |||||||||
When evaluating DeepTech startups, applying classical SaaS lenses often leads to wrong conclusions – and missed opportunities. | |||||||||
Let’s look at the facts. | |||||||||
| |||||||||
In a traditional SaaS business, growth is usually benchmarked via MRR increase, customer acquisition cost to annual contract value ratios, churn, and net revenue retention. The holy grail: repeatable, scalable Go-to-market playbooks and self-serve customer journeys. | |||||||||
These metrics work well when products solve horizontal problems in a standardized way – think marketing automation, CRM, or collaboration tools. | |||||||||
But DeepTech is different. | |||||||||
These companies operate at the frontier of science and engineering. Their customers are often R&D teams, labs, or industrial innovation units. Their sales cycles are longer, integration efforts are higher, and the level of required domain understanding is very different. | |||||||||
The result? | |||||||||
Revenue ramps more slowly – especially in early phases, as validation, regulatory approvals, or manufacturing compatibility take time.Customer collaboration is more intensive and less scalable – as projects are often co-developed and highly specific.Products are much “stickier” – once adopted, they are deeply embedded in the customer’s core IP, workflows or materials stack.The moat is not just product, but knowledge + trust + process alignment – which takes time to build, but is harder for competitors to disrupt. | |||||||||
For classical SaaS investors, this can feel uncomfortable. Product-market fit in DeepTech is typically not expressed in short-term MRR graphs, plug-and-play churn dashboards or viral product-led growth. The familiar indicators of traction and repeatability don’t always show up early – even when a startup is solving a mission-critical problem. | |||||||||
Yet this discomfort says more about the metrics than the market. | |||||||||
As McKinsey and BCG both point out in recent reports, scaling DeepTech startups requires new mental models and tailored operating frameworks. | |||||||||
One of the biggest bottlenecks we see across DeepTech startups: | |||||||||
How to scale deep customer collaboration without scaling headcount accordingly? | |||||||||
That’s where it gets interesting. | |||||||||
Let’s take the example of Exomatter, one of our portfolio startups developing AI tools for the discovery of next-generation sustainable materials – a field traditionally dominated by manual simulation and trial-and-error in the lab. | |||||||||
The problem: every industrial use case – from packaging to batteries to semiconductors – requires specific material properties and context. Onboarding a new customer can be as complex as a full R&D project. | |||||||||
Their solution: Exira – an AI agent purpose-built to scale deep technical collaboration. It guides users through discovery processes, contextualizes Exomatter’s models, suggests new research paths, and adapts its guidance to specific customer targets and constraints. | |||||||||
Instead of hiring a large technical services team, Exomatter has launched a scalable, intelligent interface that understands both the customer’s goals and the physics behind the product. | |||||||||
A promising step toward making DeepTech more scalable – without compromising its depth. | |||||||||
Have a great week! | |||||||||
PS: Wait …. Let us not forget this important development above: DeepTech’s momentum led Bavaria to surpass Berlin as Founders Capital #1. Read more in the Spiegel Article below (in German). |