Most cities that talk about building their own internet never build it. Provo built it, wired every house in the city, and then spent nine years trying to get out from under it.
The story is usually told in one line — Provo sold its fiber network to Google for a dollar — and that line is true and almost entirely misleading. It leaves out that the city had already sold the network once before, that it got it back, that residents were paying for it on their power bills for years before Google showed up, and that the dollar did not include the debt.
Here is the whole sequence, with the numbers as they were actually published.
Why a city built a fiber network at all
In the late 1990s Provo convened a task force to look at whether the city should be in the telecommunications business. The consultants studied it, the committee recommended going ahead, and a phased plan emerged: first fiber rings for the city's own use, then a demonstration project in one neighborhood, then the whole city.
The logic was not exotic at the time. Provo owns its own electric utility — one of a small number of Utah municipalities that does — so the city already had crews, easements, poles and a billing relationship with every household in town. If you already run wires to every house, running one more kind of wire looks like a smaller step than it does from outside.
By 2004 the city was building in earnest, and had borrowed to do it.
The number nobody agrees on
The size of the borrowing is printed at least four ways in the public record:
- $36.5 million — cited in taxpayer-advocacy write-ups, alongside a separate $5.4 million the telecommunications fund owed the city's Energy Department
- $39 million — the Salt Lake Tribune's figure, and the one used in most 2013 coverage
- $39.5 million — the Reason Foundation's analysis
- $39.6 million — the Daily Herald's figure at the time of the Broadweave sale
These are probably not four claims about one number. They are different snapshots — principal versus principal-plus-fund-transfers, original issue versus balance at a later date — reported by people who each had a reason to round differently. The structure is more informative than the headline anyway: the city committed to $278,000 a month for twenty years, which is about $3.3 million a year.
Remember that monthly figure. It is the thread running through everything that follows.
The wholesale model, and why it did not hold
Provo did not sell internet service. The city owned the fiber and leased capacity to private internet providers, who did the retailing, the marketing and the customer support. Several signed up — Mstar, Nuvont, Veracity, and others.
On paper this is the elegant version of municipal broadband: the public builds the road, private companies run the trucks, nobody has to trust a city IT department with billing disputes.
In practice the arrangement had one structural weakness. Provo's bond payment was fixed and the providers' revenue was not. The $278,000 came due every month whether or not anyone subscribed. When providers went out of business or fell behind on their payments to the city, the gap did not vanish; it moved onto the city's books. Meanwhile Provo was competing against an incumbent telephone company with far deeper pockets, in a market where the pitch — fiber to the home, in 2005 — was ahead of what most households thought they needed.
By April 2008 the Reason Foundation estimated the network's cumulative losses would top $10 million by the end of that fiscal year, separate from the borrowed principal still outstanding.
Sale number one: Broadweave, 2008
On May 6, 2008, Mayor Lewis Billings announced that Provo would sell iProvo to Broadweave Networks, a Lehi company, for $40.6 million — enough, in principle, to retire the bonds.
Read the terms and the picture changes:
- The purchase was through a city-financed loan spread over 19 years. Provo was, in effect, selling the network to a buyer using money the city itself was lending.
- No money was required up front. The loan was backed by a $6 million letter of credit from Sorenson Capital.
- Ownership would revert to Provo if Broadweave could not make its payments.
- Broadweave would become both the wholesale operator and the sole retail provider, buying out the existing service providers.
The city also paid about $2.6 million in costs associated with the sale.
So the risk had not left Provo. It had been moved one company away, with a return path written into the contract.
The return path got used. In April 2009, Broadweave scaled back marketing, saying the cost of setting up new customers was prohibitive. In August 2009, Broadweave and Veracity merged and asked the city to take over part of the bond payment for 18 months, offering to repay it later with interest; Provo drew on its Energy Department reserves — about $1.4 million — to cover the shortfall. Local politics that year turned on the question of whether the city could afford to let the buyer fail. By July 2011 the city was preparing to take the network back.
And in September 2011, Provo added a surcharge to every utility ratepayer's bill to service the bond, and began shopping the network. By 2013 that surcharge was widely reported at $5.35 a month for a residential account — a figure worth knowing, because it is small enough that most households never questioned it and large enough that, across a city, it covers a $3.3-million-a-year obligation.
That is the moment the cost became universal. Before it, iProvo was a line item in a city fund. After it, every household in Provo was paying for the network every month, including households that had never subscribed to it and never would.
Sale number two: Google, 2013
John Curtis took office as mayor in January 2010 and has been blunt about what he inherited: the network was the biggest headache waiting for him, subscribers were not coming, the equipment was already aging, and the press had described it as a millstone around the previous administration's neck.
In April 2013, Google announced it would take over the network, and the Provo Municipal Council approved the deal unanimously. The price was one dollar.
What Google actually agreed to was the expensive part: finishing and upgrading a network built to mid-2000s standards, and operating it as Google Fiber. Provo became the third Google Fiber city, after Kansas City and Austin.
It was also the odd one out, in a way that mattered enormously to residents. In the other launch cities Google was digging streets. In Provo the trunk fiber was already under them.
The distinction that matters, and that most retellings flatten: passing a house is not connecting it. Provo's network ran past essentially all of the city's roughly 35,000 homes, but at the time of the sale only about 9,000 were actually hooked up — roughly a quarter. Google's commitment included finishing that job, connecting the two-thirds-plus of households that had fiber in the street and nothing coming into the building. Under iProvo, making that street-to-home connection had cost about $700.
So Provo did not get a finished network for a dollar. It got the expensive, slow, disruptive part — the trenching — already paid for, in 2004, by itself. The payoff arrived nine years later in a form nobody had planned.
What the dollar did not buy
The sale transferred the asset. It did not transfer the debt.
At the time of the deal, coverage put the continuing obligation at roughly $3.3 million a year for about another twelve years — the same $278,000 a month, still landing on utility bills. The city also faced about $722,000 in equipment costs of its own so that municipal operations could keep using the gigabit service.
This is the single most misreported thing about iProvo. "Provo sold its network to Google for a dollar" implies a city that walked away clean. What actually happened is that Provo gave away the asset and kept the liability, on the reasonable judgment that keeping both would cost more.
What Google actually promised
The dollar makes a good headline and tells you nothing. The obligations Google took on are where the deal lives:
- Finish the build-out. Connect the homes that had fiber in the street and no service in the building — the large majority of the city.
- Upgrade the equipment. As Provo's deputy mayor put it at the time, the fiber itself was still good; it was everything around the fiber that had aged. Estimates of what that would cost Google ranged from about $18 million (the Utah Valley Chamber of Commerce's figure) to $20–30 million in independent analyses. For scale, Google's Kansas City build was reported at around $94 million — which is the clearest possible statement of what Provo's 2004 bond had already bought.
- Free basic service for seven years. Any home on the network could take a 5 Mbps down / 1 Mbps up connection with no monthly charge for seven years, for a $30 activation fee. The comparison that makes that number land: activation in Google's other cities was around $300, and connecting a Provo house under iProvo had run about $700.
- Free gigabit for city facilities for the same seven-year period.
- Free gigabit for 25 public institutions and nonprofits under Google's Community Connections program, awarded by application.
- An escape clause. If Google changed its mind, it could sell the network back to Provo — again for $1.
Paid service was priced in line with Kansas City: about $70 a month for gigabit, or roughly $120 bundled with television.
Two honest footnotes. First, a 5 Mbps free tier was already modest in 2013 and is thin now; for some student households it was slower than what they were already buying. Second, the free-service commitment had an end date built into it. By early 2021, with the seven years nearly up, Provo City and Google Fiber were negotiating a new agreement; council members noted at the time that Google had signaled no immediate plan to switch the free tier off, and separately that Google was winding down its television product. Anything about what is available at your address today belongs in a current internet guide, not in a history.
The road not taken: UTOPIA
Provo's story is usually told alone. It should be told next to the other one.
While Provo was building its own network by itself, a coalition of other Utah cities built UTOPIA together — a shared, publicly owned fiber network now owned and run by more than a dozen city governments. The two projects are cited constantly in the same national arguments about municipal broadband, usually as a matched pair of cautionary tales.
But they made opposite decisions on the question that turned out to matter most: open access.
UTOPIA was built as an open network — the public entity owns the fiber, and any qualifying retail provider can sell service over it. Provo started there too. Then, in the 2008 Broadweave sale, the city handed the network to a single company acting as both wholesaler and sole retailer, ending the open-access arrangement and leaving several of the earlier providers badly burned.
That decision is the reason the Google deal was possible. Google Fiber is a closed network: Google owns the pipe and sells the service. A network legally and structurally committed to open access could not simply be handed to it. Provo's could, because Provo had already given that model up five years earlier for unrelated reasons.
It is an uncomfortable piece of causation. The compromise that looks worst in the 2008 story is the one that made the 2013 exit available. Whether that counts as luck or as vindication is a question people in this valley still argue about, and it is a better argument than the one about whether cities should own fiber at all.
Is the bill finally over?
The original structure — $278,000 a month for twenty years from the mid-2000s — retires somewhere around the middle of this decade. The 2013 reporting, counting twelve more years from that point, lands in the same window.
We are deliberately not printing a specific end date. The city restructured and refinanced pieces of this along the way, which is exactly the kind of detail that turns a clean-looking number into a wrong one. If you want the real answer for your own household, two places have it and this page does not: your current Provo utility statement, and the city's published budget documents, which itemize the telecommunications debt service.
If the surcharge has come off your bill, that is the end of a twenty-year story you may not have known you were part of.
What it actually cost, and what it actually bought
Two conclusions get written about iProvo, and each one is usually presented as though the other did not exist.
The financial verdict is bad, and it is not close. The network never covered its own debt service. It was sold to a buyer that could not pay, using money the city lent, and it came back. It was then given away. Residents paid a surcharge for years on infrastructure the city no longer owned. Judged by the standard Provo set when it borrowed — that the network would pay for itself — iProvo failed, twice, expensively.
The infrastructure verdict is genuinely good, and it is also not close. The fiber exists. It reaches essentially the whole city. It carries gigabit service today. Provo has had a national-brand fiber network since 2013, which is earlier than almost any comparable American city and considerably earlier than it would have arrived if the city had waited for a private company to decide Provo was worth the trenching. Ask anyone who has moved here from a similar-sized city about their internet options before and after.
Both are true. The interesting question is not which one to believe but what the trade actually was: Provo's residents bought their city's fiber infrastructure at a price they did not choose, on a schedule they did not set, and received it years earlier than the market would have delivered it. Whether that is a scandal or a subsidy depends mostly on what you think infrastructure is for, and people in this valley genuinely disagree.
What is not in dispute is that the fiber went in the ground once, and it is still there.
The timeline, compressed
- Late 1990s — city task force recommends building a municipal telecom network; a phased plan is adopted
- 2004 — iProvo construction underway; city commits to about $278,000 a month for 20 years
- 2004–2008 — wholesale model: city owns fiber, private ISPs retail service; several fail or fall behind
- April 2008 — Reason Foundation estimates cumulative losses will exceed $10 million
- May 6, 2008 — Mayor Lewis Billings announces sale to Broadweave Networks for $40.6 million, financed by the city over 19 years, with reversion on default
- April 2009 — Broadweave cuts back marketing on customer-acquisition costs
- August 2009 — Broadweave and Veracity merge; ask the city to absorb part of the payment for 18 months; Provo draws about $1.4 million from Energy Department reserves
- January 2010 — John Curtis takes office
- July 2011 — city prepares to take the network back
- September 2011 — utility-bill surcharge added for every ratepayer; city begins shopping the network
- April 2013 — Google Fiber deal announced; municipal council approves unanimously; sale price $1; bond stays with Provo
- Mid-2020s — the original 20-year payment schedule runs out
Why this still matters here
Three reasons, and none of them are nostalgia.
It is the reason your internet is good. Provo's unusual fiber position is not the result of being a tech hub. The causation runs the other way around, and it runs through a 2004 municipal bond. For what is actually available at your address now, see our internet service guide.
It is the honest counterweight to the Silicon Slopes story. The valley's tech narrative is a series of wins — WordPerfect and Novell, Omniture, the companies founded here. iProvo is the same region making a large technology bet with public money and losing it in financial terms while winning in physical ones. A tech-history conversation that only includes the exits is not a history.
It is a live template. Cities across the country are still deciding whether to build municipal broadband, and iProvo is one of the most-cited cautionary examples in that argument — usually by people who have never been here and who leave out the part where the fiber worked. If you live in Provo, you are living inside the case study.
For the valley's other large public bet on infrastructure, and what it left behind, see Geneva Steel.
Frequently Asked Questions
What was iProvo?
Why did Provo sell iProvo to Google for one dollar?
Did the Google deal wipe out Provo's debt?
Am I still paying for iProvo on my Provo utility bill?
Was Provo the first Google Fiber city?
Who was mayor during the iProvo saga?
Was iProvo a failure?
What internet options does Provo have now?