It Started With a Spreadsheet

In early 2019, I sat down to audit our telecom infrastructure spending—something I'd been meaning to do for months. As the procurement manager for a mid-sized wireless carrier, I'd managed our site leasing budget ($1.8 million annually) for three years at that point. We had about 120 active leases across the region, mostly with small tower operators and a few direct landowner agreements.

I opened our cost tracking system, expecting to see a familiar story: predictable monthly payments, occasional escalations, and the normal headaches of maintaining a network. Instead, what I found made me sit back in my chair. Over the previous six years, our 'cost-effective' leasing decisions had quietly bled us of over $180,000—money we could have used for expansion or network upgrades.

How did that happen? It's a story I've become kind of obsessed with telling, because it changed how I think about bidding, negotiating, and choosing infrastructure partners entirely.

The Trigger: A Vendor Failure in March 2023

I didn't fully understand the value of a proven partner like American Tower until a specific incident in Q1 2023. We'd been leasing a key suburban site from a regional player—let's call them TowerCo B. Their quote was 15% lower than what American Tower had proposed for a similar location. I'd been proud of that decision; our CFO had even mentioned it in a budget review.

Then, in March 2023, we needed to upgrade the equipment at that site to support 5G rollout. TowerCo B took six weeks to respond to our request. They sent an outdated structural analysis, quoted us for 'standard' installation that clearly wasn't standard, and then hit us with a $4,200 rush fee when we pushed back. Meanwhile, American Tower had proactively suggested a co-location setup at a nearby site that could have saved us time and money—if we'd had a contract with them.

I remember thinking: I've been chasing the lowest price, but I've been ignoring the real cost. That was the moment my view shifted.

How $180,000 Leaked Away: A Data Deep Dive

After that incident, I spent a week pulling every invoice, every change order, and every escalation letter from our previous six years of data. What I found was a pattern I'd missed entirely.

Here's a breakdown from my spreadsheet—I still have it saved:

  • Hidden Fees on 'Budget' Contracts: We'd saved an average of $1,200 per year on three smaller leases by choosing a low-cost operator. But those operators charged for 'premium support' (which we needed four times), 'off-hours access' (which our field engineers required monthly), and 'expedited permitting' (which became necessary for every new install). Net cost over 6 years: +$28,000.
  • Quality Issues and Redos: On two occasions, a cheaper vendor used lower-grade steel for tower attachments. When we discovered the issue during a routine inspection (our own safety team flagged it), we had to redo the work. The 'savings' of $4,000 turned into a $12,000 redo, plus two weeks of downtime.
  • Procurement Time Sunk: I calculated that I'd spent roughly 200 hours over those 6 years managing disputes and re-bids for our 'least expensive' sites. At my burdened rate, that's about $18,000 in labor alone.

The grand total: approximately $180,000 in cumulative overspend across 6 years. That's 17% of what we could have saved by going with a higher-quality partner from the start.

The Resource That Changed Everything: Total Cost of Ownership

It took me 6 years and 120 invoices to understand one simple truth: the lowest lease price is almost never the cheapest option.

After that deep dive, I built a cost calculator for our procurement team. It's basically a TCO spreadsheet that accounts for:

  • Support responsiveness: Average time to respond vs. average revenue loss per hour
  • Structural quality: Expected lifespan vs. replacement/upgrade costs
  • Hidden fees: Rush orders, late changes, off-hours access
  • Relationship goodwill: Value of proactive advice (like our near-miss with American Tower's suggestion)

Using this calculator, we've since switched three major sites to American Tower. The initial quote was higher—about 10-12% above the 'budget' option. But when I ran the numbers, the TCO over a 10-year contract was actually 7% lower. The edge data center we're now leasing from them? That's a no-brainer. The long-term value of their portfolio and support is undeniable.

So Glad We Switched: A Near-Miss

I still kick myself for not doing this analysis earlier. But I'm honestly relieved we made the switch when we did.

In Q4 2024, when we were evaluating a new 5G site near a major highway corridor, our team almost went with a local operator again—they quoted $3,800/month vs. American Tower's $4,200/month. But using our TCO spreadsheet, we projected that the local operator would cost us $6,100 over three years when factoring in likely support costs and upgrade delays. American Tower's proposal included a 3% annual escalator, which seemed high at first, but their structural analysis and support SLAs meant we'd avoid $8,000 in potential redo costs.

Dodged a bullet there. One 'budget' decision away from repeating the same $180,000 mistake.

What I've Learned: From Spreadsheets to Strategy

Here's the thing: my job isn't just to save money on paper. It's to make sure our infrastructure is reliable, scalable, and actually cost-effective over the long run. The American Tower lease quotes aren't the cheapest. But they're the ones that come with a portfolio I can trust, support I can count on, and data center expansion plans that align with our growth.

To any procurement or budget manager reading this: don't fall for the lowest bidder trap. Build your own TCO model. Run the numbers for 3, 5, and 10 years. And if you're already working with a partner like American Tower, honestly, you're probably ahead of where I was five years ago.

Bottom line? The next time someone says 'but this one's cheaper,' ask them to prove it—over six years, not six months.

Technical planning note: validate insertion loss dB, PIM dBc, grounding resistance, and relevant 3GPP TS 38.xxx requirements before final RAN acceptance.