The McCaw Cellular Story
WMMM #106- This week, I share an example of a failed negotiation.
Jeff Keplar Newsletter January 10, 2026 9 min read
Empathetic Negotiation
Continuing with the theme from last week, I asked Google NotebookLM:
"What is empathetic negotiation, and how do I reveal this skill in my content?"
The response was: "empathetic negotiation is the practice of prioritizing the long-term relationship and the counterparty's emotional and business reality over short-term extraction."
"It involves moving beyond 'positions' (what they say they want) to understand their 'interests' (what they actually need or fear) and solving for their internal pressures.
Sounds about right.
"Your content reveals this skill through specific stories."
The stories cited by NotebookLM included: T-Mobile Sidekick, VarTec, Cingular Wireless, and Verizon DBaaS.
It called out four different demonstrations of empathetic negotiation:
1- Distinguishing Interests from Positions
2-Protecting the Relationship over the Transaction
3-Solving Personal Pain to Earn Trust
4-Preparation as a Form of Respect.
Remembering each of those stories reminded me of another one.
Today, I share a story I have not written about before.
Taking Us Back to that Time
The year was 2002.
It was not a good time to be dedicated to the telecom vertical if you were a technology sales professional.
But that is precisely where I found myself.
After hitting a peak in 1999-2000, the telecom bubble burst alongside (and partly because of) the dot-com crash.
That was in 2001.
By 2002, the damage was unmistakable: massive writedowns, carrier bankruptcies, and a near-freeze in capital spending.
Mimicking this timing, the Oracle sales organization responsible for the telecom vertical experienced a brutal contraction.
In a 6-month period that began in late 2001, a management team of 17 shrank to 1 (me).
It was May, Oracle's fiscal year-end, and I found myself temporarily reporting to the Vice President of US Sales, George.
The Account
I chose to call the account McCaw Cellular for this story.
That's because "AT&T Wireless," the accurate name, might confuse many of us.
Today's AT&T Wireless is formerly Cingular Wireless.
Cingular was the outcome of a merger between BellSouth Mobility and Southwestern Bell Mobile Systems, companies formed within BellSouth and SBC, respectively.
Most of these accounts fell under my responsibility for much of my career with Oracle.
McCaw Cellular pioneered national-scale cellular service in the United States with the launch of a roaming-capable network they branded "Cellular One."
McCaw was bought by "the old AT&T" in the early 1990s and rebranded as AT&T Wireless Services, becoming one of the largest cellular carriers in the country.
It was purchased by Cingular in 2004.
My story is about AT&T Wireless Services, the account that was the former McCaw Cellular, based in Seattle.
The Transaction
In that era, many large Oracle accounts had "Enterprise Agreements (EAs)."
The number of users accessing Oracle software had become increasingly more complex to administer and costly to license.
This was especially true if companies purchased licenses on an as-needed basis.
The problem was exacerbated with fast-growing companies.
The Telecom industry had been booming for nearly a decade, and McCaw was one of the fastest-growing companies in the sector.
Oracle's sales team went to companies like McCaw and sold them licenses for their "enterprise."
A single license - no need to count the number of users accessing the software.
But there was a provision in these EAs called an annual "true-up."
Oracle required compensation for the increased use of its software.
They tied that to the growth of the enterprise - "shared reward."
"We only get compensated when you do well."
Many of these EAs used the number of employees as the metric for the true-up.
With McCaw Cellular, we had the tied annual true-up to the number of wireless subscribers.
The number of subscribers was a more accurate barometer of McCaw's growth and success.
The true-up date was in May, which conveniently coincided with Oracle's fiscal year-end.
Every May, McCaw's Procurement team would prepare for its true-up with Oracle by obtaining a certification document from an authorized McCaw executive stating the number of subscribers as of that month.
They would compare that number with the number of subscribers McCaw was licensed for in the Oracle EA and calculate a true-up number.
In the EA, we placed a table showing pricing for different subscriber levels.
The cost per subscriber would decrease as the subscriber count increased.
The decrease in cost per subscriber was a concession we built into the EA in return for more business - larger deals.
For each level, there was a dollar amount.
McCaw could find its level with the recent subscriber certification.
It was a simple calculation to determine the amount of the true-up (compare the amounts for the current level for which they were licensed with the amount for the level they currently belong to based on the recent certification).
Subtract the "current" from the "belonged" to obtain a true-up amount.
We made it simple.
The Negotiation
One of the benefits of an EA is that it reduces the number of license purchase transactions a customer would have to make with Oracle to one per year.
Our customers were pretty vocal about their dislike of Oracle's business practices that, as it turns out, were not materially better or worse, fair or unfair, complex or simple, clear or vague than the other large tech suppliers of the day: IBM, Sun, HP, EMC, Microsoft, or SAP.
Regardless, reducing their exposure to once per year improved productivity and customer satisfaction.
As with many true-up transactions, the procurement department took the lead for McCaw.
Unless McCaw wanted to add an Oracle product to their existing license rights, there really wasn't an opportunity to "negotiate."
McCaw had no additional products to add to their EA in May of 2002.
Yet, Procurement needed concessions - they wanted to improve the terms of the EA.
They did not agree to the terms in the EA and told us they would not approve the true-up in May.
Grow at All Cost
Oracle customers had been conditioned to receive extraordinary incentives in May.
May was Oracle's fiscal year-end.
Oracle had been a growth company for the past ten years.
It was still a growth company in 2002, growing much faster than the rest of the software industry.
Oracle built great software, and its secret weapon might have been its highly incentivized sales force.
But it wasn't.
It was its CEO, Larry Ellison.
Mr. Ellison is often credited with building a distribution channel featuring a meat-eating culture of aggressive sales tactics.
He threw fuel on this fire with a compensation plan that made it possible for salespeople to make life-changing money when they exceeded their quotas.
Mr. Ellison learned that the "run-of-the-mill" salesperson, with modest sales skills, would use whatever concessions they could to close enough deals to exceed quota.
Price discounts were the easiest.
"The price goes up on Monday Close," as demeaning and elementary as that sounds, became the mantra.
"Monday" meant June 1st.
It was the quickest and easiest path to the gold.
Mr. Ellison calculated that a "grow at all cost" strategy would accelerate Oracle's ascent to a dominant market position.
Enabling even an average salesperson to be successful would allow Oracle to accelerate at scale.
Once there, he knew he could enforce a change in the sales organization's behavior toward something a little more professional.
So he embraced the behavior of the everyday salesperson.
Because he knew something else.
For every dollar of Oracle license sold (discounted or not), Oracle received $2.20 over the next ten years.
So what?
That $2.20 was virtually guaranteed to Oracle.
It was an Oracle Support fee.
With Support, a customer received software upgrades, new releases, 24/7 customer service, and bug fixes.
It was optional, but no customer would want an Oracle license without it.
What is the most important thing to know about that $2.20?
The pre-tax margin exceeded 90%.
Surely Not Everyone Was Kung Fu Fighting
No, not everyone.
But enough sales teams took advantage of the situation.
Some might say they exercised the freedom given to them.
Others might say they exploited whatever means necessary to enrich themselves.
Either way, there were enough May deals done with aggressive terms that our customers expected it.
McCaw's procurement group did.
When it came to EAs, they were already aggressively priced with special concessions attached.
Oracle did not approve or encourage this behavior for larger accounts with EAs.
There was nowhere for me to go.
It highlights the problem with this genius, but goofy approach.
When Oracle's behavior tells its customers that the terms of the contract mean little in May, where do the concessions end?
When is enough enough?
When does no mean no?
Tag, You're In
McCaw's procurement believed that the only thing that saved Oracle in situations like this was the clock.
Oracle would be forced to continue conceding until there was no time left in the fiscal year.
If you weren't receiving concessions in your negotiation, simply hold off on signing until the clock struck midnight Pacific Time (Oracle's HQ was in Redwood Shores, CA).
McCaw figured they would sweat us out until midnight.
Surely Oracle would not let its fiscal year end without resolving this true-up with McCaw.
At around 10:00 pm PT, Mike, McCaw's CIO, took the reins from Procurement.
We had a couple of conversations, but I suspected something wasn't right.
I did not think he was in a position to execute this true-up.
Could it be that he had to extract a concession from me or lose face with the procurement team?
I hoped not.
I had nothing to give him.
My account manager and I had made that clear for weeks.
Mike then went silent on me.
It's a common tactic.
It's still uncomfortable.
Ten minutes before 2:00 am (my time), my phone rang.
It was George.
He wanted to know what was going on with McCaw.
I told him.
He said that Safra (yes, that Safra) had received a call from Mike with McCaw.
How did he get her number?
I told George I did not know.
I did not give it to him.
George told me that he needed to call Safra back and would connect with me later.
Did We Close the McCaw Deal?
Yes, but not on May 31, 2002.
George called Safra and told her I was holding Oracle's position.
Safra called Mike back and told him that he had Oracle's Best and Final Offer.
His salesperson wasn't holding anything back.
Mike let the clock run out without signing the true-up.
What would George think?
Would Safra be annoyed that one of my accounts called her?
Will I have a job tomorrow?
I learned the next morning that Mike had told Safra something that made sense of everything we were experiencing.
He said that McCaw's procurement had enlisted the help of a Big Six firm.
(I cannot recall which one, but if I had to guess, they were using Deloitte or KPMG.)
Their advice was that Oracle would sweeten the terms of their true-up to get it closed in May.
Everybody was Kung Fu Fighting!
Mike should escalate to Safra, a top Oracle executive not responsible for sales, if he didn't get satisfaction with his Oracle sales executive.
So Mike bypassed George and me and went to Safra.
The consulting firm gave him Safra's direct line.
The Cooling Off Period
George did not like that one of our accounts had called Safra.
He thought I should have had better account control.
But he appreciated that I held our line in the negotiation.
He possibly felt that way because Safra appreciated my transparency with Mike.
It turns out that she had grown tired of the heavy discounting, last-minute concessions, and "Price Goes Up on Monday" tactics of many in our field sales organization.
I did not call Mike back immediately.
He was given bad data from a trusted source.
I did not want to be "right."
Being "right" in this case would not be "effective."
I wanted to be effective.
How would I allow him to save face?
I needed to get this transaction closed this quarter (Q1 of Oracle’s next fiscal year.)
But I waited.
I thought he might call me (I was hoping he would).
Reengaging and Reaching Agreement
He did.
Mike had found a piece of data that we had yet to discuss to break the ice.
"Do we still have Support?"
I said that they did as far as I knew.
"But our renewal was part of the true-up.
And we did not get it done - it expired."
I double-checked and got back to him.
Their account was flagged, but they still had Support.
"Did you know that someone from Oracle Support told someone on our procurement team that if we did not renew, Oracle would shut us off.
You never threatened us.
I told our procurement team that I wasn't worried.
I've been working with Oracle executives since May 31st.
Whoever made that threat isn't part of Oracle's executive team.
You have bad data."
"And that wasn't the only bad data we received."
I heard about that.
Mike thanked me for my professionalism during the prior couple of months.
He delivered their true-up order within a few days.
Years later, Mike was announced as the new CIO at another telecom account, and when the Oracle account team met with him, he asked about me.
Playing the long game is a winning strategy and a key to empathetic negotiation.
Lessons…
1) We must have trust - at the time of this deal, I had not earned enough trust with the CIO.
2) Identify all influencers - the consulting firm being used by Procurement was a blind spot.
3) Just because everyone is doing something doesn't mean you should.
Thank you for reading,
Jeff
