Fermentation Manufacturing Cost: Why Startups Can Improve the Process and Still Get Surprised by the Economics
- Gustavo Valente
- Apr 6
- 8 min read
The Economics of Biotech Scale-Up:
What actually drives fermentation manufacturing cost
In most biotech and foodtech startups, the technical team is doing exactly what it should be doing.
They are improving yield, shortening fermentation time, increasing robustness, improving recovery, reducing variability, and making the process more scalable.
That work is essential.
Without process optimization, there is no path to a credible commercial process.
But here is the uncomfortable reality:
A startup can make real technical progress and still move only marginally closer to a commercially viable manufacturing process.
That is where many teams get caught off guard.
Not because the science is weak. Not because the technical team is focusing on the wrong things.
But because technical progress alone does not always reveal whether the wider manufacturing system is becoming commercially stronger too.
This is where techno-economic analysis becomes valuable.
Not because it replaces science. And not because it tells the R&D team to stop optimizing.
But because it adds a wider lens and asks a bigger question:
Is this process becoming more manufacturable, more financeable, and more commercially robust at scale, or just technically better in the lab and pilot plant?
That distinction matters more than many startups realize.
Short answer: what actually drives a fermentation process manufacturing cost?
Fermentation manufacturing cost is rarely driven by one variable in isolation. It is usually shaped by the interaction between:
scale
utilization
downstream burden
recovery losses
throughput
utilities
labor intensity
facility-related overhead
capital intensity
and the commercial assumptions behind how the plant will actually operate
That is why a process can show genuine technical improvement while cost of production remains stubbornly high, or improves far less than expected.
Fermentation cost is not a lab metric. It is a manufacturing system outcome.
The real risk is not lack of optimization. It is lack of visibility.
This is an important nuance.
The problem is usually not that startups are optimizing too much. The problem is that not every improvement changes the economics equally.
A better yield is valuable.
A shorter fermentation cycle is valuable.
A cleaner downstream step is valuable.
A lower-cost feedstock is valuable.
But from a commercial manufacturing perspective, the real question is:
How much does each improvement actually improve the economics once the process is translated into a real plant?
That is where surprises begin.
Because some changes that look highly significant in development only move the economics modestly, while other less visible factors end up dominating cost at industrial scale.
A process can improve scientifically while remaining commercially fragile.
Fermentation cost is a system outcome, not a single-variable problem
This is one of the most important insights early TEA can provide.
Founders often hear simplified statements like:
yield is everything
feedstock is the main driver
downstream is 60% of production cost
we can cut cost dramatically by optimizing one section of the process
Sometimes these statements are directionally useful.
But they are often repeated too broadly and without enough context.
In reality, fermentation manufacturing cost emerges from how the whole system behaves once you account for:
material flows
bottlenecks
recovery across the full process
installed plant requirements
fixed-cost absorption
utilities
labor
commercial operating pattern
and realistic plant utilization
That is why early TEA is not just a financial exercise.
It is a way of connecting scientific progress to manufacturing reality.
The wrong commercial bottleneck often stays hidden too long
There are good reasons why startups focus first on variables like:
yield
fermentation time
feedstock cost
downstream efficiency
These are visible, measurable, and closely tied to the technical work happening every day.
But when a process is translated into a commercial manufacturing scenario, the economics are often driven by a wider combination of factors.
1. Scale and utilization
A process may look attractive at one commercial scale and much weaker at another.
Plant size affects:
equipment count
CAPEX
labor structure
utility demand
fixed-cost dilution
and whether the process ever reaches a competitive cost position at all
Utilization often matters just as much.
A plant that looks strong at 90% utilization can become economically fragile if the company only fills 30 to 50% of capacity in the early years.
That changes:
depreciation per kilogram
fixed-cost absorption
overhead burden
and the credibility of the COGS story presented to investors
A process may be technically sound and still commercially weak if the scale and utilization assumptions are unrealistic.
2. Recovery and downstream burden
A process may look strong up to the fermenter and still struggle commercially if too much value is lost downstream.
Recovery losses affect:
effective output
raw material use per unit sold
upstream burden carried forward
and the scale needed to hit target production volumes
Downstream complexity matters too, sometimes enormously.
A process with multiple purification, concentration, polishing, drying, or waste-handling steps can accumulate cost through:
yield loss
equipment count
cleaning burden
utilities
labor
cycle time
and capital intensity
In some cases, downstream burden is not only driven by process complexity, but by unchallenged commercial assumptions.
I have seen startups trying to replicate the exact form or purity of an incumbent product, for example forcing a bio-based product into a powder because the conventional version is sold that way, or targeting a very high purity that may not actually be necessary for the end use.
But part of the TEA conversation should sometimes be:
why does this product need to look exactly like the incumbent?
Could a liquid form work?
Could a lower purity still be commercially acceptable because the impurity profile is different?
Questions like these can radically change the downstream burden and, with it, the economics of the whole manufacturing system.
The point is not simply whether DSP is “important.”
It is whether downstream complexity is making the whole manufacturing system economically heavy.
3. Utilities, labor, and facility-related overhead
Utilities are easy to underestimate early because they sit in the background of technical development.
But at scale, steam, cooling, chilled water, compressed air, electricity, process water, and wastewater treatment can become major contributors.
Labor intensity can also become heavier than expected when a process involves:
manual intervention
complex batch handling
cleaning steps
monitoring points
multi-step transfers
or frequent operational adjustments
Then there is the broader facility burden.
Not just the core equipment, but the wider plant environment:
buildings
support areas
utility infrastructure
HVAC
special warehousing
maintenance support
quality systems
operational overhead linked to running the site
A process can carry a heavy facility burden simply because it requires too much manufacturing infrastructure relative to the output it generates.
4. Capital intensity and commercial ramp
Even when a process performs well technically, a capital-heavy configuration can still distort the economics.
The lab can tell you whether the process is getting better. TEA helps tell you whether the business case is getting stronger too.
Higher CAPEX usually brings:
more depreciation
more financing pressure
higher fixed-cost absorption needs
and a more demanding commercial ramp to justify the investment
This is one reason some processes look viable in principle but become fragile in practice.
A stronger process is not always a stronger business case.
What matters is not just whether the process can work.
It is whether it can work under realistic commercial conditions.
Why repeated industry phrases can mislead founders
One thing I see often in biotech and fermentation circles is the repetition of simplified statements such as:
downstream is 60% of cost
optimizing DSP will reduce cost of production by 30%
feedstock is always the main driver
yield is the main economic lever
These phrases spread because they sound plausible, and in some cases they reflect real situations.
But they can become misleading when repeated without context.
For example, in some processes downstream equipment may indeed represent a large share of the direct process equipment CAPEX.
But that does not automatically mean downstream dominates:
total installed CAPEX
total project CAPEX
or full manufacturing cost
Once utilities, indirect project costs, buildings, waste treatment, labor, and full operating assumptions are included, the picture can change significantly.
The same applies on the OPEX side.
In some cases, DSP is indeed a major economic driver.
In others, the larger issue may be:
underutilized assets
low throughput
poor recovery
a capital-heavy facility
energy demand
or a cost structure that never gets diluted properly at realistic commercial volumes
What dominates direct process equipment CAPEX does not necessarily dominate total project cost or COGS.
That is exactly why broad rules-of-thumb can distort priorities if they are used in place of a real TEA.
This is how startups get surprised by the economics
Most startups do not get surprised because they ignored technical development.
They get surprised because technical progress created confidence while the wider manufacturing and commercial logic remained underexplored.
That can look like:
a process that performs better in the lab but still needs too much capital at scale
a COGS model that looks attractive only at unrealistic plant utilization
a business case that depends on a recovery level not yet demonstrated across the full process
a plant concept that works only if downstream complexity stays artificially simplified
a commercialization plan that assumes cost competitiveness before the full burden of manufacturing is understood
In practice, this is also where early TEAs can change decisions that would never emerge from lab work alone.
I have seen companies:
revisit feedstock sourcing strategies
realize they need feedstock or production partnerships
identify more viable production locations
or change the location they originally had in mind for the process to work commercially
Those are not scientific failures.
And they are not questions the lab can answer on its own.
They are part of the wider manufacturing and commercial logic that only becomes visible when the process is pressure-tested as a real production system.
Some of the most important commercial decisions in scale-up are not about improving the biology, but about creating the conditions under which the biology can be manufactured competitively.
These are not failures of science.
They are failures of visibility.
And that is exactly the gap TEA is meant to close.
What a better early TEA conversation looks like
A better conversation does not ask only:
how do we improve yield?
how do we shorten the fermentation?
how do we make DSP more efficient?
It also asks:
which variables actually move total manufacturing cost the most?
what happens once realistic plant utilization is included?
are we talking about direct equipment cost, total project CAPEX, or full cost of production?
does the process still look commercially credible once utilities, labor, waste treatment, and facility burden are included?
are we improving a technically important variable, or one that truly changes the economics?
what would actually need to be true for this process to work at commercial scale?
That is the value of TEA at this stage.
It helps startups see the difference between:
technical improvement
and commercial strengthening
The best teams do both.
Final thought
A startup can be doing the right technical work and still get surprised by the economics if the wider manufacturing and commercial logic is not pressure-tested early.
That is not a criticism of science.
It is a reminder that science alone does not automatically reveal the full commercial picture.
Fermentation manufacturing cost is rarely explained well by one variable or one slogan.
It is shaped by how the whole manufacturing system behaves once scale, utilization, recovery, utilities, labor, facility burden, and capital structure are taken into account.
That is why early techno-economic analysis matters.
Not because it replaces process development, but because it helps ensure that technical progress is translating into a commercially stronger path to scale.
If your team is making real progress in the lab but still has limited visibility into what is actually driving the commercial cost structure, this is exactly the stage where an early TEA can help. It can reveal whether the wider manufacturing system is improving too, or whether important risks are still hiding outside the immediate technical focus.
If that is a question you are actively working through, feel free to reach out. I’m always happy to discuss how to pressure-test manufacturing cost logic before those surprises show up later in fundraising, pilot design, or commercial planning.
Gustavo Valente
Director, Sustech Innovation
WhatsApp: +52 55 3405 0552
Series note
This article is part of The Economics of Biotech Scale-Up — a series exploring the real manufacturing decisions founders and investors face when moving from lab to commercial reality.
ScaleUpReady™ note
These are exactly the kinds of manufacturing decisions explored through the ScaleUpReady™ approach: using techno-economic analysis not as a static model, but as a decision framework that evolves with the process.