PCBA quoting in Excel is slow & expensive

80% of customer BOMs are defective. The data shows just how much this is costing you, and why quoting spreadsheets can’t fix it. 

EMS companies consistently report 80% of customer BOMs have errors. A foundational iNEMI study (Arnold, Motorola) documented the impact of this problem:

• 80% of received BOMs are defective

• BOMs average 100 line items

• Up to 40% of line items have errors

• 10 minutes of staff time are consumed fixing each error

That’s 5.3 hours of staff time per quote. For every 8 quotes, that’s one full-time employee. It’s not obvious because this time is spread across different job functions, including program management, procurement, and engineering. 

Recent studies have documented that engineers waste, on average, 68% of their time searching, configuring, and recreating parts. 98% of OEM suppliers think improvements in communications are necessary. 

In this article, we’ll show why your Excel-based quoting workflows need to change, and what manufacturing quoting software does that spreadsheets can’t.

Quoting is a chain of dependent steps  

A PCBA quote moves through seven dependent steps:

1. BOM intake

2. BOM scrub

3. Supplier RFQs and follow-ups

4. Labor estimation

5. Cost rollups

6. Internal validation and approvals

7. Post-award scrub

If any one of them breaks, the whole chain slows down. Here’s what each looks like in practice:

1. BOM intake

Customer data comes in every format you can imagine and a range of file types. Excel files, CSV exports, PDFs, sometimes multiple files stitched together. One BOM might list a single manufacturer part number per line; another might stack alternates across multiple rows, all tied to the same internal part number.

Even within a single file, structure varies. Different column names, missing fields, inconsistent units, duplicated entries, or customer-specific conventions that only make sense if you’ve seen them before.

Before anyone can price anything, all of this has to be translated into a consistent structure that the quoting process can actually work with. Until that happens, nothing downstream can run cleanly.

2. BOM scrub

This is the time sink everyone knows but rarely measures. Part numbers get corrected, manufacturers verified, duplicates removed, and formats aligned. Missing data gets filled in, obsolete parts get flagged or replaced, the list goes on. And it’s all done over email.

This is where most of the hidden engineering and procurement time disappears, one line item at a time.

(...) we calculated material prices in one tool and then manually transferred them to Excel. The data from work preparation was passed on to purchasing via email; it was all very complex.

— Mario Rehbehn, Head of Sales Team, Prototype Service, BMK

Mario came to call this “email ping-pong” — the constant back-and-forth between work preparation, purchasing, and customers before they automated the workflow.

3. Supplier RFQs & follow-ups

This is where complexity explodes.

One internal part number can map to multiple manufacturer part numbers. Each of those manufacturer part numbers can have multiple authorized distributors. Suddenly, a single BOM line item turns into several supplier RFQs.

Multiply that across an entire BOM, and you’re dealing with hundreds of component quotes.

Buyers send out requests, wait for responses, chase down missing quotes, and reconcile conflicting information across all these options. Then it all has to be aggregated back into a single, comparable view.

This portion of the process is almost always manual. Supplier quotes come back in emails, spreadsheets, PDFs. Someone has to extract the relevant data, map it back to the correct line items, and re-enter it into the quoting spreadsheet.

That re-entry step is where errors creep in, versions drift, and time disappears. Email becomes the system. Response times drive cycle time more than anything else.

Before, it was writing emails and calling people. That showed us we needed software support to make our processes better and, at the same time, raise the level of process reliability and transparency.

— Christian Gerland, Head of Procurement, TQ Group

4. Labor estimation (SMT, THT, test)

There are a few ways this gets handled, but a common approach is to estimate labor based on component packaging. Each package type, 0402, QFN, BGA, and connector, is mapped to an expected placement or processing time. Add up the counts, make some assumptions, and you get a rough estimate of the cycle time.

But this only works if the packaging data is available and accurate. If it’s missing, someone has to find it. Datasheets get opened. Parts get classified manually. In some cases, engineers are making judgment calls just to fill in the gaps so a labor estimate can move forward.

Even when the data is there, these models are simplified. They rarely account for board complexity, panelization, feeder setup, or line balancing. So the estimate can look precise in a spreadsheet while still being directionally off. And once that labor estimate is set, it gets rolled up directly into the total cost.

5. Cost rollups

This is where the math gets messy.

For each line item, you’re comparing multiple supplier quotes across multiple manufacturer part options, each with different price breaks at different quantities.

If a customer asks for pricing at 1, 10, and 100 units, the optimal supplier can change at each tier. So now you’re solving a matrix of decisions across the entire BOM, not a single selection.

On top of that, you have to account for SMT attrition (extra components needed to cover feeder setup, handling losses, and placement scrap), scrap rates, and other process assumptions that impact true material consumption.

All of this logic ends up buried in spreadsheets. Multiple tabs, nested formulas, copied workbooks. Small changes ripple through the model, and it becomes harder to trust the output over time.

🚀 Accelerate your quoting process

Eliminate manual steps and deliver high-quality quotes at scale.

👉 Get started

6. Internal validation & approvals

Engineering, procurement, and operations all need to agree. Questions come back. Assumptions get challenged. The quote loops internally before it ever reaches the customer.

But validation itself is harder than it should be.

The data is spread across multiple tabs, formulas, and versions of the spreadsheet. Material costs in one place, labor assumptions in another, supplier selections buried in email threads or separate files. To understand the quote, you have to reconstruct how it was built.

For an approver, this becomes a problem. You’re expected to sign off on margin, risk, and assumptions, but you don’t have a clean, consolidated view of what’s driving the number. So validation turns into spot-checking, asking questions, and relying on the person who built the quote to explain it.

That slows things down, and it introduces risk. Approval is happening without full visibility, or it requires pulling more people back into the process just to build confidence in the result.

7. Post-award scrub

After the order is won, the work shifts.

The quoted BOM now needs to serve as a source document for execution. That means translating it into ERP for purchasing and planning, and into the systems used to program SMT and inspection machines.

This is where data quality really gets tested.

The same BOM used for quoting now needs to drive machine programs, which in turn are used for line simulations. Those simulations produce the actual cycle times that flow back into ERP and determine real labor costs.

So there’s a loop: quote → machine program → simulation → ERP.

If the data coming out of the spreadsheet isn’t structured and complete, that loop breaks. Fields are missing, mappings don’t align, and assumptions have to be revalidated.

And because the quote spreadsheet was never designed to feed these systems, much of this translation is manual. Data gets reworked, re-entered, and rechecked. Issues found here don’t disappear. They show up as delays, rework, or margin erosion after the order is already won.

If you’re using Excel as the glue holding this together, you’re not alone. But Excel was never built to:

• Standardize incoming data

• Enforce validation rules

• Coordinate workflows across teams

  
  

💡 Build smarter with better BOM insights

Optimize your electronics sourcing with powerful tools designed for modern hardware teams.

👉 Get started

Bottom line: Excel quoting runs on people

Strip the spreadsheets away, and what’s left is people doing the work by hand, often late at night.

We’ve always turned around quotes [on the] same day. Customers just didn’t realize my team was sometimes working until 10 o’clock at night to do it.

— Hassan Malak, VP Sales, Omega EMS

Why Excel quoting is so expensive

Labor is what makes Excel-based quoting expensive. Every BOM issue, every supplier exchange, every validation step pulls human time. Here’s where that cost concentrates.

People are where the cost comes from…

Every BOM issue triggers the same chain of work:

• Identify the issue

• Interpret intent

• Find the correct data

• Validate the choice

• Communicate with suppliers or the customer

In an Excel-driven workflow, all of this is manual. That means your most experienced people are doing repetitive interpretation work:

• Engineers analyzing part discrepancies

• Buyers chasing supplier confirmations

• Operations validating assumptions that should have been resolved upstream

You’re paying skilled labor to solve the same categories of problems over and over.

According to the Engineering Efficiency Report, 48% of engineers spend an hour or more per day searching for parts.

… but not all BOM problems need people

Most BOM issues fall into three repeatable categories documented across EMS teams:

• Completeness — missing required data like MPNs, quantities, supplier info, or attributes

• Consistency —  mismatches such as quantities vs. reference designators or inconsistent naming

• Correctness — wrong or obsolete part numbers, incorrect manufacturer assignments, bad specs, repeated MPNs, phantom line items, Do Not Place (DNP) items.

These aren’t rare edge cases. They show up on almost every RFQ.

And importantly, they are structured problems. That means they can be validated, flagged, and often resolved without human interpretation.

When these categories are handled manually in spreadsheets, every issue becomes a task for engineering or procurement. When they’re handled systematically, most of that work disappears.

About 80% of all positions are recognized automatically after upload. That frees us to focus on technical guidance and optimization.

— Martin Schiedel, Technical Specialist & Production Planner

Excel vs. automation: what manufacturing quoting software actually does

Excel is good at storing and calculating.

It is not good at enforcing consistency or applying rules at scale.

It won’t:

• Validate a BOM against current market data in real time

• Flag lifecycle risks consistently

• Normalize supplier naming conventions automatically

• Enforce required fields before work begins

So every exception becomes a human task. Here’s how that plays out across the most common BOM issues 👇

Why Excel quoting doesn’t scale

Cost is one problem. Volume is another. As RFQs increase, the same workflows that already feel heavy start to compound — and the spreadsheet model has no answer for it.

The structural problem

This creates a scaling issue.

As RFQ volume increases:

• More BOM defects show up

• More supplier interactions are required

• More internal coordination is needed

Which leads to:

• Longer turnaround times

• Higher labor cost per quote

• Growing dependence on a few experienced individuals who “know how to fix things”

Excel-based systems also block automation. As data isn’t standardized and rules aren’t enforced, you can’t reliably automate any part of the process. Every attempt to improve quoting runs into the same constraint: the inputs are inconsistent, and the logic lives in people’s heads or buried in spreadsheets.

So instead of reducing effort, a spreadsheet system locks you in and actually prevents you from automating.

The core insight

If your quoting process depends on human interpretation, it will be slow and expensive. Excel-based workflows depend on human interpretation by design.

That’s the constraint.

How to fix the quoting bottleneck

Messy BOMs are part of the business. The lever you do control is how your organization handles that variability:

• Standardize how data is accepted and validated

• Encode decision logic for repeatable issues

• Reduce the number of steps that require human judgment

In practice, this means moving away from spreadsheets as the operating layer and toward systems built for the job.

Manufacturing quoting software does exactly this: it validates BOMs on intake, enriches part data, applies sourcing and costing logic consistently, and keeps engineering, procurement, and sales working off the same dataset.

When that happens, turnaround time compresses, labor cost per quote drops, and reliance on tribal knowledge decreases. Quoting stops being a bottleneck and becomes something you can scale.

If you’re evaluating this shift, manufacturing quoting software built specifically for EMS — like Luminovo — turns BOM variability into a structured, automatable process.

🚀 Ready to transform your workflows?

See firsthand how Luminovo helps teams move faster and make better decisions.

👉 Request a demo