From 300 to 3,300 SKUs: When Clean-in-Place Becomes the Bottleneck

When Product Variety Outpaces Process Design

Just like manufacturing knowledge loss in the face of the Great Retirement, SKU proliferation is quietly reshaping food and beverage manufacturing. Washington Post recently noted the consumer shift in demand has resulted in an era of "peak sauce" or the "Golden Age of Condiments."

This shift has a direct impact on manufacturing operations. Plants designed to run a handful of hero products are now managing hundreds or thousands of variations—and the cleaning and changeover processes weren't built for it. The result: clean-in-place cycles get longer, line availability shrinks, and every new SKU adds operational drag.

One sauce manufacturer recently shared a story that stuck with me: When he joined his company 12 years ago, they had around 300 SKUs. Today they have 3,300.

And those are active SKUs. That means if a co-manufacturer gets an order for any one of those SKUs, they're ready to run it. They're prepared. They have the recipe.

That's the level of agility and readiness our customers need now. The question is whether the cleaning and changeover process can keep up.

Key Takeaways

• SKU proliferation changed the game: Pre-COVID, plants optimized for hero products and long runs, not an ever-evolving or growing product mix. Post-COVID, consumer demand for variety exploded with larger and larger product catalogs, but the factories stayed the same.

• CIP recipes scale poorly: A cleaning program built for 10 products can't handle 100 without adding buffer time (and that buffer compounds across every shift).

• OEE misses the real problem: The line looks productive during cleaning, but line availability for value-added production is what actually pays the bills.

• Real-time visibility changes the math. When you can see inside the pipes, you can clean to conditions instead of worst-case presets... and stop relying on chemical suppliers to tell you what's optimal.

What is SKU proliferation?

SKU proliferation is the rapid increase in the number of SKUs or product variations a manufacturer must produce on the same equipment. A facility that once ran 10 products might now manage 100 or 1,000, each with different ingredients, quality standards, and cleaning requirements.

‍

How Consumer Demand Shifted After COVID Resulting in Higher SKU Count

You can divide the sauce, food, beverage world into pre-COVID and post-COVID. Our relationship with food changed, concinciding with a generational change as well.

Pre-COVID, a lot of lines were configured for hero products. The whole focus was capacity, productivity, just keep it running, running, running all the time. OEE was the metric, and all OEE really measured was utilization—as long as the line's running and every piece is running, the plant is producing. Everything was static. Make the same thing over and over again.

Post-COVID, trends changed. People started ordering out, trying new flavors, testing new things. Preferences changed. The younger generation especially came out of it thinking "what's the flavor of the day?" Food became about new experiences instead of "I have an experience and I love that experience." And it's caught on—it's not just the younger generation anymore. Everybody is addicted to more variety.

The factories, though? They remained the same.

Checkout our recent blog post for more insights on how manufacturing is shifting in the post-pandemic world: When Manufacturing Expertise Walks Out the Door.

‍

Why Process Manufacturing Can't Easily Adapt

In discrete manufacturing — electronics, automotive, industrial machinery — plants can reconfigure fairly quickly because the work is discrete. Move a robot out. Redo the setup of the plant to adjust for a new operational methodology.

In process engineering, that's not the case. Once the pipes are laid down, the setup is very static. Once a line is configured for a particular product, everything in that line is catered to what the output is. Large tanks, large pumps — these are millions of dollars of asset investments. Lines and recipes are very pre-configured.

So now think about what happens when a sauce plant goes from running three SKUs in four batches — a very predictable production schedule for the week — to running constantly changing products with different inputs, different quality standards, different packaging and labeling requirements.

‍

How Co-Manufacturers Manage SKU Complexity

Co-manufacturers are a great example of what's happening across food and beverage. Our favorite fast-food chain's signature sauce isn't made by them. Three vendors might be producing it, and the same facility making that sauce is also making dips for a competing burger chain and caramel for a national coffee brand.

A food or sauce plant today is catering to a whole bunch of OEMs. So the question becomes: how does the plant keep everything running, but also optimize for profitability? What batches need to run for Company A? What batches for Company B? The plant wants to run all of them, but the products are different, the inputs are different, the quality standards are different, the packaging is different.

How does a manufacturer take the existing capitalization and change it from running long batches with infrequent changes to continuous changes, switching constantly?

Why CIP Recipes Don't Scale with SKU Proliferation

Here's where it gets practical.

With 10 products, a plant probably has maybe four recipes for cleaning. A three-step clean-in-place process for some products, a five-step CIP process for others. For six of those SKUs, the three-step CIP works because the plant isn't dealing with strawberries.

The other three have strawberries, so they need an extra step to ensure no seeds get stuck in line. Or maybe they have more oils and fats, so they need something additional to avoid microbial issues or residue.

That's it. Done. Just run it.

But going from 10 recipes to 100 recipes raises the question: are the same three-step and five-step going to work for 100 products with that much variation?  

It is trial and error leading to standardizing the worst-case CIP cycle.

The plant runs it, and then quickly finds out that the old recipe used to work, but when a new concentrate gets added—caramel concentrate or something in beverage — it sticks more, leaves behind more residue. And quality says no, that cleaning is not going to work anymore.

What's the first thing that happens? Buffer it up.

How Buffer Time Reduces Line Availability

When cleaning times increase, line availability goes down.  

The sanitation team says "the quality guy told me it's failing, so we're adding extra 10 minutes to every cleaning." Now the plant is running multiple batches, and this has to become a planning consideration.  

Can the cleaning happen at the end of the shift when everybody's done? Sure. But then quality says no, we need an extra cleaning step during the shift. Planning has to accommodate. The plant manager has to get involved. Everyone is asking: what does this mean for the number of cases we produce?

For capacity-constrained co-manufacturers, every minute counts. That's the implication of SKU proliferation. Business can be great, but unless the plant can keep it productive, it doesn't matter.

‍

Why OEE Doesn't Capture Line Availability Problems

Here's the thing about SKU proliferation: OEE alone as a metric stops being useful.

‍

The line's running all the time. It looks productive. The line's idle only because cleaning is happening, and cleaning takes 30 minutes. From an equipment efficiency standpoint, everything is fine. The cleaning cycle is running perfectly, then the plant waits and does the next thing.

There's no downtime recorded. That line's not down. It's just waiting for cleaning to get done. So even though everything looks right and the efficiency metric looks good, what's actually shrinking is line availability.

The old way of measuring equipment efficiency isn't going to help. Manufacturers have to look at the process itself and ask: can the process be faster? Where does the next time bonus come from so there's more agility? The plant can stay more available, but still has to meet changeover and cleaning requirements.

This is the balance people are trying to strike.

Not sure how much time you're losing to buffer? Use our free CIP Analysis Tool to benchmark your current cycles.

The Hidden Cost of Worst-Case CIP Presets

Every time a quality swab fails, the CIP matrix gets bigger. And if all the quality tests pass, there's still a doubt — there must always be a buffer in there. Because how can everything pass? But then everything must pass.

The problem is nobody knows how many pennies are left in the cushion. How much can actually come out?

Are Chemical Suppliers Optimizing for Your Time or Their Revenue?

This is where the chemical supplier comes in. What's the number one biggest expense? Not water, not energy. In order of magnitude, time is the number one most expensive. Then chemicals, then water, then energy.

Because everyone wants the most time, the question becomes: how can that chemical step get shorter? And the chemical supplier is on both ends of the spectrum — selling the chemicals and advising on how to optimize chemical use.

So the supplier says: time savings come from increasing dosage. Put more chemicals in for a shorter time. The supplier gets the revenue and the plant gets time back.  

There is a better way to do it. that reduces chemical usage and still gives you line availability. You should look at the science behind cleaning instead.

How Self-Optimizing CIP Unlocks True Agility and Supports SKU Growth

This is what one of our sauce manufacturing customers is doing, and it's very interesting.  

They're running an design experiments using Laminar inline sensors. They're going to run some sub-cycles with 0.2% concentration, some with 0.5%, and one with 1%. And they're going to look at the cycles — Laminar ML models dynamically adjust these CIP cycles.

The experiment seeks to answer these questions:

• What is the true minimum chemical concentration needed to be effective?

• Is more chemicals effective?

• Where is the inline drop off point for effectiveness for the factory in total and for the requirements for specific product runs?

‍

With Laminar’s patented spectral sensors monitoring real-time conditions inside the pipes and sub-second, ML control, manufacturers are able to optimize every cycle.

When adjustments and tweaks happen in real time, they have confidence because the product is never at risk. CIP underwash protection is built in. If something goes wrong, it's not going to make a bad quality product.  

Within that safety net, the plant doesn't need to trial and error to find the right chemical concentration and allow technology to optimize the cycle independently. Now the chemical supplier can focus on chemical supply, and Laminar runs your CIP cycle to it's fullest optimization.

‍

What Manufacturers Need to Support 3,300+ SKUs

Imagine a hundred recipes or a thousand. If somebody comes up with the 3,301st variation, how can a manufacturer make sure there's good cleaning, fast changeover, and good line startup with a self-correcting, self-healing process so that doesn't become the bottleneck?

There's already a whole bunch of considerations when making a new SKU. What's the recipe? How does it mix? What's the viscosity? How does it fill? What's the label on the box? Cleaning and changeovers can't become another issue. If a plant spends a lot of time there and has downtime, it screws up all the value-added activity.

Changeovers and CIPs are necessary evils of production. They're not considered value-added activities. Value-added activity is when the line's just filling product. So how do manufacturers make the necessary evils very optimal and really fast?

Why Self-Driving Clean-in-Place Systems Are Now Table Stakes

Given that manufacturers have no control over the proliferation (e.g. where the high demand comes from, how agile the sku portfolio needs to be, what's coming next week) the plant is always going to be responding. Maybe it's a new OEM, or maybe it's the company's own marketing department. Someone at Nestlé or Unilever sits with the food scientists and says "for this summer, we're putting a new flavor of drink out there." And everybody has to dance to that tune.

All of these opportunities, all of this uncertainty about what's coming next — it's an exciting world for sure. But manufacturers need more self-correcting, self-adjusting, closed-loop systems so they can focus on the value-added activities and not have to worry about what should be table stakes.

Producing a good quality product and executing a perfect changeover and line startup should be table stakes. Let's not screw up on the table stakes. Manufacturing leaders need a strategies that puts technology in their facilities that keeps the plant clean, reports on sustainability, shows how things are improving — does all of that digitally, without human intervention, so humans can innovate and focus on getting the productivity out.

That's really the story.

Frequently Asked Questions

What causes SKU proliferation in food and beverage manufacturing?

Consumer preferences shifted dramatically after COVID. People want new experiences, new flavors, variety. The younger generation especially came out of the pandemic with a "flavor of the day" mindset—food as novelty rather than routine. That attitude has spread. Product variety is now the expectation, not the exception. This high demand for variety, combined with co-manufacturers serving multiple brands on the same lines, has driven SKU counts from hundreds to thousands at many facilities.

How does SKU proliferation affect CIP cycles?

When a plant goes from 10 products to 100, the same two or three cleaning recipes can't handle all the variation of that many skus. New ingredients behave differently—more oils, plant-based colorants, proteins — and leave different residues. Quality teams respond by adding buffer time "just in case." That extra 10 minutes per cycle compounds across every changeover, every shift, reducing the time available for actual production.

Why doesn't OEE capture line availability problems?

OEE measures whether equipment is running, but a CIP cycle counts as "running." The line isn't down — it's cleaning. So OEE can look healthy while actual availability for value-added production shrinks. With SKU proliferation, manufacturers have to look beyond utilization metrics and ask whether the process itself can be faster.

How can manufacturers reduce changeover time with multiple SKUs?

The key is moving from time-based cleaning to condition-based cleaning. Instead of running a 60-minute cycle because that's what the recipe says, the plant detects when each phase is actually complete and ends it there. Real-time visibility into what's happening inside the pipes lets manufacturers clean to conditions, not worst-case presets. Plants using this approach typically reduce CIP cycles by 20-30 minutes.

What is condition-based CIP optimization?

Condition-based CIP optimization uses real-time sensors to monitor what's actually happening inside the pipes during cleaning. Instead of running fixed-time cycles validated by manual sampling at the end, plants can see when each phase reaches completion and adjust dynamically. This approach reduces over-washing without risking under-washing, because there's continuous visibility and CIP underwash protection built in.

How do chemical suppliers influence CIP cycle times?

Chemical suppliers are often in a conflicted position: they sell the chemicals and advise on how to optimize chemical use. A common recommendation is to increase dosage to reduce time—which generates more chemical revenue. Without visibility into what's actually happening during the cycle, manufacturers can't verify whether higher concentration actually delivers proportional time savings. Real-time monitoring lets plants run experiments and see the actual relationship between concentration and cycle time.

Ready to see how much time you're leaving in your CIP cycles? Try our free CIP Analysis Tool or contact us to discuss how real-time visibility can support your SKU complexity.