Read Time: 10 min
By David Carneal – Digital Efficiency Consulting Group
If you want a warm, inspirational story about “continuous improvement,” IBM in the early 1990s is not the fairy tale you’re looking for.
This is the story where the building is on fire, the sprinklers are on fire, and someone is still asking, “Should we loop in Legal before we put out the fire?”
IBM didn’t “optimize.” IBM survived. And it wasn’t because they rebranded, “leaned into synergy,” or discovered a new font.
They survived because they got brutally honest about reality, cut internal drag, and started fixing how work actually moved through the company.
If you’re thinking, “We’re not IBM,” you’re right. You probably don’t have IBM’s size, IBM’s brand, or IBM’s ability to lose billions and still get invited to important meetings.
But you do have one thing in common with IBM back then: your company runs on processes. Those processes either help you grow… or quietly pickpocket your margin every day while everyone says, “We’re just really busy.”
So let’s talk about what IBM was facing, how ugly it got, what they did, and what your company can steal without needing a multi-billion-dollar crisis to motivate you.
When the math stopped being cute
In January 1993, IBM announced it lost $4.96 billion in 1992, described as the biggest annual U.S. corporate loss at the time. [1] That’s not “a tough year.” That’s a crater with a souvenir shop.
And then the fourth quarter of 1992 decided to outdo the whole year. IBM lost about $5.46 billion in Q4 (Los Angeles Times). [1] The Washington Post reported it as $5.5 billion and quoted an analyst saying, “It’s very ugly.” [2]
Also from the Washington Post: IBM reportedly lost about $45 million on day-to-day operations in that quarter. [2] Translation: they weren’t losing money because of one weird expense. They were losing money because the core math was broken.
Hard question, and I’m not asking to be dramatic:
If your company started losing money on day-to-day operations, how quickly would you know?
Because “eventually” is where profit goes to quietly expire while everyone is busy updating a report nobody reads.
Why it happened: the world changed, and IBM didn’t move fast enough
IBM wasn’t losing because customers suddenly hated technology. IBM lost because the market shifted hard, moving away from expensive proprietary systems and toward cheaper PCs and networks. [5]
Here’s the part that matters for your business: when a company gets big and complex, internal friction starts eating the company from the inside.
Friction is the invisible monster. It doesn’t scream. It just quietly slows decisions, adds handoffs, multiplies approvals, creates rework, and turns “simple” into “why is this taking three weeks?”
Hard question:
How long have you been in “crisis mode,” but calling it “busy”?
If your org runs on heroics, you don’t have a process. You have a very expensive hobby.
The stakes: bitter medicine, big numbers, bigger consequences
By mid-1993, IBM wasn’t dealing with “headwinds.” It was dealing with survival.
In July 1993, IBM announced an $8 billion quarterly loss and said it would eliminate another 35,000 jobs by the end of 1994. [3] That loss was driven largely by about $8.9 billion in extraordinary charges tied to restructuring, layoffs, and plant closings. [3] [6]
And Lou Gerstner, newly installed as CEO, described the experience with a line that should be printed on every executive dashboard that’s currently lying to people:
“We have to get behind us, the Chinese water torture, we’ve been going through, quarter after quarter and year after year.” [3]
Hard question:
What is your company’s “water torture”?
The recurring issue that drips on everyone’s forehead every day that the organization has decided is “just how it is.”
The “vision” trap and the rare move IBM made
When companies get squeezed, they usually do one of two things:
- Cut costs and hope.
- Announce a grand vision and hope harder.
IBM had to cut costs, but Gerstner refused to hide behind the comforting fog of “vision.”
In July 1993 coverage, he said: “The last thing IBM needs right now is a vision.” [3]
Because when execution is bleeding, a “vision” can become a fancy way of avoiding the uncomfortable truth:
Your workflows are broken.
You can’t PowerPoint your way out of a bad process. You can’t town-hall your way out of approval paralysis. And you definitely can’t “culture” your way out of a workflow that needs six handoffs and two spreadsheets just to ship something.
Hard question:
Where are you using “strategy” as a substitute for fixing the machine?
The pivot: re-engineer the business, don’t just trim it
By January 1994, IBM reported its first quarterly profit in more than a year: about $382 million in Q4 1993. [5] That’s real progress.
But the year overall was still brutal: roughly an $8 billion loss for 1993, with full-year revenue around $62.7 billion. [5] IBM’s SEC filing shows 1993 total revenue of $62.716B, net loss of $8.101B, and restructuring charges of $8.945B. [6]
So no, IBM wasn’t instantly “fixed.” They were stabilizing while rebuilding.
Here’s the key line worth stealing: the Los Angeles Times described Gerstner rejecting the grand-vision routine and focusing on “re-engineering” IBM to make its processes more efficient. [5]
Because here’s the mistake most companies make under pressure: they cut people, but they don’t redesign the work.
So the waste stays… and now it’s louder because fewer humans are left to absorb the chaos. That’s not efficiency. That’s just suffering with a smaller payroll.
What “re-engineering” looks like without the buzzwords
“Re-engineering” can sound like consultant-speak for “we made a new org chart and everyone is now confused in fresh new ways.”
In real life, re-engineering is painfully simple. It’s asking three questions most companies avoid because the answers are inconvenient:
- What is the customer actually paying for?
- Which steps directly create that value?
- Which steps exist because we don’t trust ourselves, our data, or our systems?
Those “trust” steps are where waste breeds. They show up as double-checking, re-entering data, exporting and emailing, reconciling numbers, “just to be safe” approvals, and meetings to confirm what the system should already know.
If you want to spot opportunities fast, look for these patterns:
Pattern A: The handoff parade
Work changes hands so many times that nobody owns the outcome. Everyone owns a step. Nobody owns the result.
Pattern B: The approval staircase
Five sign-offs for a decision that should take five minutes. Most approvals don’t prevent disasters. They prevent discomfort.
Pattern C: The “special case” factory
The process works for a small slice of reality, so most work becomes exceptions. Exceptions are where costs go to party.
Pattern D: The spreadsheet economy
Spreadsheets exist because systems don’t integrate or no one trusts reporting. The spreadsheet becomes “the process,” and now the business is fragile.
Pattern E: The meeting tax
Meetings grow when information flow is broken. The calendar becomes a coping mechanism.
Hard question:
Which pattern is your company’s favorite hobby? If the answer is “all of them,” don’t worry. That’s common. It’s also fixable.
Efficiency isn’t cardio, it’s removing the drag
Efficiency is not “work faster.” That’s just exhausting.
Real efficiency is removing waste that hides in plain sight:
- Duplicate data entry because systems don’t talk
- Rework because standards are unclear
- Approval chains that exist to reduce anxiety
- Handoffs where work sits
- Outdated steps built for a world that doesn’t exist anymore
Waste is sneaky. Waste usually looks like “normal.”
Hard question:
What is normal in your operation that shouldn’t be?
How DECG helps without breaking your business
A common fear is: “If we change processes, we’ll break something.” Fair.
Companies break things when they change too much at once, or when they automate a messy workflow and create faster chaos.
So DECG keeps it simple:
1) Simplify
Remove steps that add no customer value. Unnecessary approvals. Duplicate checks. Manual transfers. Comfort reports.
2) Standardize
Define how work should flow, with clear handoffs, clear definitions, and a simple exception path.
3) Automate
Only after the logic is clean. Automation should remove low-value work, not fossilize bad habits.
Hard question:
Where are you currently automating confusion? If you’ve ever said, “The system made me do it,” you know what I mean.
A practical “IBM test” you can run this week
Pick one core process (order-to-cash, procure-to-pay, pick-pack-ship, service delivery). Then:
- Map it on one page
- Mark every place work waits
- Mark every time data gets retyped
- Mark every approval
- Ask at each step: “If we removed this, what would break?”
If the answer is “nothing,” congratulations. You found waste.
If you want help doing this without turning it into a six-month internal debate club, that’s what DECG is built for: diagnose the waste, design the fix, and hand you a plan your team can implement.
Because the most expensive way to improve efficiency is to wait until you have no choice.
Conclusion
IBM’s 1990s lesson isn’t “cost cutting saves companies.” It’s this: when the machine is broken, you don’t win by talking about the machine. You win by fixing how work flows through it.
So here’s the last hard question:
If I walked into your business next week and asked, “Where is the waste hiding, and what’s your plan to remove it?” would you have an answer?
If yes, keep going. Measure it. Improve it. Protect it.
If no, good. Now you know where to start.
Small improvements create real change.
And sometimes those small improvements are the difference between “we’re fine” and “why is everything on fire?”
CTA: Want a quick sanity-check on where work is getting stuck (before it becomes expensive)? Book a short discovery call.
Footnotes
- Los Angeles Times (Jan 20, 1993) — “IBM Reports Largest Annual Corporate Loss” https://www.latimes.com/archives/la-xpm-1993-01-20-fi-1540-story.html
- The Washington Post (Jan 20, 1993) — “IBM REPORTS A RECORD $5 BILLION LOSS” https://www.washingtonpost.com/archive/politics/1993/01/20/ibm-reports-a-record-5-billion-loss/9f27cba9-65d7-4f7e-ac4c-869734d4054c/
- Los Angeles Times (Jul 28, 1993) — “$8-Billion Loss Posted by IBM; More Layoffs Set” https://www.latimes.com/archives/la-xpm-1993-07-28-mn-17823-story.html
- Los Angeles Times (Jan 26, 1994) — “IBM Chalks Up 4th-Quarter Profit, but ’93 Loss Deepens” https://www.latimes.com/archives/la-xpm-1994-01-26-fi-15546-story.html
- U.S. SEC EDGAR (IBM) — 1993 Annual Report content incorporated in Form 10-K text file https://www.sec.gov/Archives/edgar/data/51143/0000950112-94-000794.txt
- Los Angeles Times (Jan 25, 1995) — “IBM profits soar as turnaround continues” (profit comparison cited in earlier draft) https://www.latimes.com/archives/la-xpm-1995-01-25-fi-23388-story.html
- IBM Newsroom remembrance of Gerstner era (used in earlier draft; verify exact page if needed) https://newsroom.ibm.com/
Leave a Reply