Cogent Communications Holdings, Inc. (CCOI) Earnings Call Transcript & Summary

Good morning, and welcome to the Cogent Communications Holdings First Quarter 2025 Earnings Conference Call. As a reminder, this conference call is being recorded, and it will be available for replay at www.cogentco.com. A transcript of this conference call will be posted on Cogent's website when it becomes available. Cogent's summary of financial and operational results attached to its press release can be downloaded from the Cogent's website. I would now like to turn the call over to Mr. Dave Schaeffer, Chairman and Chief Executive Officer of Cogent Communications Holdings. Please go ahead.

Thank you, and good morning, everyone. Welcome to our first quarter 2025 earnings conference call. I'm Dave Schaeffer, Cogent's Chief Executive Officer and on this call this morning with me is Tad Weed, our Chief Financial Officer. We have received numerous comments from investors related to the structure of our earnings call. We greatly appreciate their observations and constructive comments, and we've implemented a number of those suggestions in the script that we are using for this call. Please continue to provide additional suggestions to help us refine our reporting. We are well aware that Cogent has undergone significant changes over the past two years, and we want to fully address the impact of those changes on our strategy and our prepared remarks and strive to focus on our growth plans going forward. For the quarter, I'd like to touch on some significant milestones that we achieved. I want to recognize that these achievements are what some of the milestones that we've achieved. We are now offering wavelength services in 883 data centers with 10-gig, 100-gig and 400-gig capabilities. We have materially been able to reduce our provisioning times to today approximately 30 days. Our wavelength revenues for the quarter were $7.1 million, an increase of 114% over the same period in 2023 -- in '24. Sequentially, our wavelength connections increased by 18% sequentially and our wavelength revenue increased by 2.2%. The vast majority of our connections were provisioned near the very end of the quarter. We have sold wavelength services now in 329 locations. We have provisioned and cleaned up our former backlog of wavelength orders. We currently have a backlog and funnel of 3,433 wavelength opportunities with more wave provisioning experience and the actual ability to deliver services, we now anticipate that between 4% and 5% of this funnel will be installed each month going forward. We also expect, based on the growth in the sales activity that by year-end, there will be 10,000 unique wave opportunities in our funnel. We currently have provisioning capacity to install 500 waves per month. We intend to capture 25% of this highly concentrated North American market within three years. Our IPv4 leasing revenue for the quarter increased sequentially by 14.8% to $14.4 million and increased 42% year-over-year. Due to the scarcity of this valuable asset and the terms of our customer contracts, we have been able to increase our IPv4 leasing pricing. We maintain a consistent acceptable use policy and did retrieve a significant number of addresses in the first quarter from a customer who violated these policies. Our average revenue per IPv4 address sold was $0.49 for the quarter, a 63% increase from the $0.30 installed base number at the beginning of the year. We have titled to nearly $38 million before addresses, which is more than any other service provider. We have realized the remainder of our targeted $220 million in cost savings that we outlined at the acquisition of Sprint. We expect to achieve at minimum another $20 million of cost savings through the second quarter of 2026. Demonstrating the impact of these savings on our cost of goods sold, they declined from $31.6 million in the first quarter of last year and our gross margin increased by 790 basis points from the first quarter of 2024 to 44.6%. Additionally, our SG&A declined by $3.8 million from the first quarter of last year. $10.6 million of the sequential increase in SG&A expenses was due to traditional typical seasonal factors, including annual CPI increases, the timing of cases taken and the accruals associated with them and the reset of payroll taxes. We are now connected to 3,500 on-net buildings. We have reconfigured several Sprint acquired facilities. These facilities have been added to our 1,668 carrier-neutral and 101 Cogent data center footprint. Our Cogent data centers have 183 megawatts of installed and available power. We have converted additionally, 79 smaller Sprint facilities into edge data centers. These edge data centers, each have approximately 40-rack capability and in total have about 28 megawatts of additional installed power. So on a combined basis, Cogent has 180 data centers, edge and core, with 211 megawatts of installed power available for customers. After the quarter ended, we repurchased approximately 100,000 shares of our common stock for approximately $5 million at an average price of $53.07 under our stock buyback program. A total of $17.4 million remains available under that program through year-end. A comment on tariffs. We do not anticipate any material impact of tariffs on our business or our CapEx projections. Much of our data center and network conversion equipment has been ordered pre tariff and a majority has been received. A portion of our network equipment purchases do have tariff input costs, but these are minimal. We recognize that we have increased our leverage due to these activities and our Board of Directors has elected to slow the rate of dividend growth, but continuing that dividend growth rate at $0.005 per share per quarter. Our dividends for the quarter rose from $1.005 to $1.01. This represents the 51st consecutive sequential increase in our regular quarterly dividend and an annual dividend growth rate of 3.6%. Now that the Sprint business is combined with our legacy business, and we have fully analyzed the revenue burn off of undesirable revenues, we are adjusting our long-term annual revenue growth rate to 6% to 8%, and we are increasing the rate at which we anticipate our EBITDA as adjusted margin to expand annually to 150 basis points. Our updated revenue and EBITDA targets are meant to be multiyear goals and not designed to be specific quarterly or annual guidance. We are nearing the ending of grooming of undesirable revenues from Sprint contracts that are set to expire. We expect to return to total top line revenue growth by mid-Q3 2025. Finally, I would like to take a moment to recognize one of our long-serving Board members, Blake Bath, for his outstanding counsel and service to Cogent. Blake has served on our Board since November of 2006 and elected to retire, and we wish him well in that retirement. Now I'd like to turn things back over to Tad to read safe harbor language and give some additional color on our operating performance.

Thank you, Dave, and good morning, everyone. This earnings conference call includes forward-looking statements. These forward-looking statements are based upon our current intent, belief and expectations. These forward-looking statements and all other statements that may be made on this call that are not historical facts are subject to a number of risks and uncertainties, and actual results may differ materially. Please refer to our SEC filings for more information on the factors that could cause actual results to differ. Cogent undertakes no obligation to update or revise our forward-looking statements we use non-GAAP financial measures during this call, you will find these reconciled to the corresponding GAAP measurement in our earnings releases that are posted on our website at cogentco.com. Now some comments on results. Our revenue per quarter was $247 million. Our rep productivity increased by 9% to 3.8 units per full-time equivalent reps this quarter which was an increase from 3.5 units per full-time equivalent last quarter. Our EBITDA, as adjusted, was $68.8 million, which was a $1.9 million increase and our EBITDA as adjusted margin increased sequentially by 130 basis points to 27.8%. Our EBITDA as adjusted is adjusted for Sprint acquisition costs, if any, during the period and payments under the IP Transit agreement with T-Mobile. In accordance with our IP Transit services agreement, we received three monthly payments totaling $25 million this quarter the same as last quarter, $25 million last quarter. A year ago, we received $87.5 million in the first quarter of 2024 as those payments stepped down in that quarter. We will continue to receive an additional 32 monthly payments of $8.3 million each until November of 2027. There are further payments related to lease obligations we assumed at closing that total at least $20 million. This amount is to be paid to us in four equal payments from November '27 to February '28. We analyze our revenues based upon network connection type, which is on-net, off-net, wavelength and noncore. And we analyze our revenues based upon customer type and we classify our customers into three types, NetCentric, corporate and enterprise. Our corporate business represented 44.9% of our revenues this quarter. It decreased 11.4% year-over-year and 2.1% sequentially. These decreases in our corporate revenue were primarily due to the continued grooming of low-margin off-net connections and the elimination of noncore products. Our NetCentric business continues to benefit from the growth in video traffic activity related to artificial intelligence, streaming and wavelength sales. Our NetCentric business represented 37.5% of our revenues for the quarter, increased 0.7% year-over-year and declined sequentially by 1.1%. Our quarterly NetCentric revenue under our commercial services agreement with T-Mobile, declined sequentially by $0.8 million and was $0.7 million for the quarter and a declined $2.5 million year-over-year. The decline in revenue from the commercial service agreement from T-Mobile and the negative impact of FX, which was $0.5 million sequentially and $1.3 million year-over-year had a negative impact on our NetCentric revenue results. Our enterprise business represented 17.7% of our revenues for the quarter. Net revenue decreased by 11.3% year-over-year and sequentially by 4.1%, primarily due to a reduction in noncore and low-margin enterprise revenues. On-net revenue, we serve our on-net customers and our 3,500 total on-net buildings. We continue to succeed in selling larger 100-gigabit connections and 400-gigabit connections in carrier-neutral data centers and selling 10-gigabit connections in selected multi-tenant office buildings. Our on-net revenue was $129.6 million for the quarter, a year-over-year decrease of 6.5% and a sequential increase of $0.9 million or 0.7%. Our sequential on-net revenue results were negatively impacted by the same contract with T-Mobile, the commercial services agreement, the $0.8 million sequential decline in on-net revenue and also negatively impacted by $0.9 million of negative FX. Our off-net revenue was $107.3 million for the quarter, a year-over-year decrease of 9.2% and a sequential decrease of 5.2%. Our off-net revenue results are impacted by our migration of certain off-net customers to on-net and the grooming -- and continued grooming and termination of low-margin off-net contracts. Comments on pricing. Our average price per megabit for our installed base decreased sequentially by 6% to $0.20 and decreased by 25% year-over-year. This is consistent with historical trends. Our average price per megabit for our new customer contracts for the quarter was $0.10, a sequential price per megabit decrease of 10% and 5% year-over-year. Some ARPU, churn statistics. Our ARPUs for the quarter were, our on-net ARPU was $496. Our off-net ARPU was $1,266, our wavelength ARPU was $1,945, and our IPv4 revenue per address for the quarter was $0.49. On churn, our on-net monthly churn rate was 1.4% and off-net monthly churn rate was 2.2%. Our network traffic was flat sequentially for the quarter but increased 8% year-over-year. Foreign currency comments. Our revenue earned outside the United States is reported in U.S. dollars and was about 18% of our revenues this quarter. The average euro to USD rate so far this quarter is $1.12 and the average Canadian dollar rate is $0.72. So these averages remain at the current levels for the remainder of this quarter. The FX conversion impact on sequential revenues would be $2 million and the positive impact year-over-year will be $1.2 million. We believe that our revenues and customer base is not very highly concentrated. Our top 25 customers were 80% of our revenues for the quarter. CapEx. Our total CapEx for the quarter was $58.1 million. Our principal payments on capital leases declined to $8 million for the quarter. We are continuing our network integration of the former Sprint network and legacy Cogent network into one unified network and converting former Sprint switch sites into Cogent data centers. We have accelerated and expanded our data center conversion program due to the high level of demand for our power availability. This program will require capital spending for the first half of '25, similar to the last half of '24 and then decline in the second half of '25. Our total gross debt at par, including our finance lease obligations was $2 billion at quarter end, and our net debt was $1.8 billion. Our total gross debt for the last 12 months EBITDA as adjusted ratio was 6.69 at quarter end and net debt was 6.08, have calculated under our note indentures. Our leverage ratio was 5.86. Secured leverage ratio was 3.44 and fix coverage was 2.8. Finally, our day sales outstanding was 29 days at quarter end, the same as the end of the year, and our bad debt expense was $2.1 million, which was less than 1% of our revenues this quarter. I'm turning the call back over to Dave.

Thanks, Tad. Now for a couple of comments on our NetCentric business. At quarter end, we directly connected to 8,240 other networks, of which 22 of these are peers and 8,218 are Cogent transfer customers. We remain focused on our sales force productivity and continue to manage out underperformers. Our sales force turnover rate was 7.1% a month. This is down from the peak of 8.7% per month, but was slightly above our historical average of 5.7% per month. At quarter's end, we had 296 professionals focused solely on selling NetCentric, 319 professionals focused on the corporate market and 14 professionals focused on the enterprise market. We remain excited about our ability to deliver profitable on-net and off-net IP services to enterprise and corporate customers. We are enthusiastic about our wavelength opportunity. The portfolio of buildings that we now connect to and the backlog on that funnel of nearly 3,433 wavelength opportunities. We have completely refreshed that funnel and cleaned out older orders that had been accumulated over a year period while we were doing network reconfiguration. We have diligently worked on accelerating the cost savings of the Sprint network integration. We have exceeded our initial targets and raised those targets. We are able to continue to monetize IPv4 addresses, fiber assets and excess data center spaces either through sale or long-term leases. We are in active discussions beyond the LOI stage with multiple counterparties. And since our inception, we've offered superior service, expedited provisioning and disruptive pricing. That is why Cogent remains an industry leader in the services itself. With that, I'd like to open the floor for questions.

[Operator Instructions] And your first question comes from the line of Jim Schneider from Goldman Sachs.

It's Josh in for Jim. I guess, within the waves business, are you seeing any change in competition, be it pricing or terms or otherwise as sequential revenue growth was not where most had expected? And do you think the Crown Castle Zayo deal will change the landscape at all? And then separately, you've referenced in the past a few more quarters of potential revenue headwinds in corporate. Can you give us an update on these trends and the mix pricing and customer growth, especially as we think about your new revenue growth targets?

Yes, sure. Thanks a lot Josh. So first of all, our two primary competitors in the wavelength market have struggled to provision and do not have the ubiquity of coverage that we have. I think that is why many customers were willing to sign agreements with Cogent when Cogent couldn't even give them a firm delivery date or a commitment to exact endpoints. As we've purged that older funnel, the funnel that we have built now is a much more accurate representation of orders that will continue to install. We've developed enough of a cadence to know that about 5% of that funnel will convert each month. In the quarter, we installed virtually all of the incremental units, the 18% sequential growth at the very end of the quarter. The reason for that is many of the customers were not ready when we were ready because they had weighted. Many of those customers did take service at the end of the quarter. We expect that the competition will continue to improve on their ability to offer ubiquity and offer faster installs, but today, we think we have a significant advantage. With regard to the Crown Castle Zayo combination, that is probably a year away. I know that each of those companies represents a number of previous acquisitions that are still being integrated, and I would anticipate even post closing, it will take several years based on the pacing that each of those companies have achieved in integration for the entire company to be functioning as a unified organization. We should always be paranoid about competitors. But at the end of the day, we think this is a fairly distant threat, not something immediate. Now pivoting to your question on corporate revenue. We have gone through the undesirable revenue in the Sprint base. We have churned the vast majority of that revenue. We know that, that has inflected our total top line growth rate negative. We believe that we will be through that process by the mid-part of Q3 based on our need to honor certain contract commitments. From that point forward, we anticipate Cogent's total revenue growth to be positive and continue after the roughly 18 years of positive growth that we demonstrated pre-acquisition of Sprint, our growth did turn negative when we acquired a business that was declining at 7.4% a year and represented 40% of the revenue of the combined company. We further accelerated that decline by electing to terminate these unprofitable noncore services. And I think the fact that we've been able to increase our EBITDA in absolute terms even with revenue decline is a very clear indication of the unprofitability of this business. But I think the corporate segment, as all of our segments, will be growing by the end or the middle of Q3 2025.

Your next question comes from the line of Greg Williams at TD Cowen.

Dave, on the $0.005 per share dividend growth per quarter, is it like a temporary move? And would we think about returning to the $0.01 growth quarter-over-quarter? And if so, what would the milestones need to be? I imagine leverage targets would be one of them to return to that growth if you choose to do so? And then just back on the waves cadence. You mentioned 4% to 5% of your bookings will be in 12 months. That's about 150 circuits a month. In the past, you said you'd get to 500 circuits a month. What needs to happen to get to that target? And when could that be reached at this point?

Yes, sure. So let me take each of those questions. Thanks for them, Greg. First of all, the Board reflected on the increase in leverage and realize that the fundamentals of the business remain strong, but that our leverage is going to continue to increase as we've outlined on several of these calls that our aggregate leverage will peak in Q3 of this year due to the decline in transit payments from T-Mobile. We have been able to affect a material amount of cost savings, but those still did not result and enough to fully offset the decline in monthly payments from T-Mobile from $29 million down to $8 billion. As a result of that, throughout the year after those payments declined, our leverage is going up. As our leverage begins to decline in Q4 of this year and going forward, the Board will continue to evaluate the pace of that delevering and is absolutely committed to returning capital to shareholders. Finally, we, I think, have continued to demonstrate our willingness to opportunistically enter the market and supplement our dividend with buybacks. That policy will continue going forward, but we absolutely believe that our ability to return cash flow will increase starting in the fourth quarter, and some of that will be used to delever. Some will be used for increasing the dividend. And finally, some will be used for opportunistic buybacks. So I think the milestone will be the reduction in net leverage. I'm now going to pivot to your second question, which is wavelength installation. As we have been very clear in previous discussions with investors, we built a funnel of wavelength opportunities with no defined installation windows. As it became clear that we could begin to install in select locations, in Q3 of 2024, we began the process of cleansing that funnel. And as expected, the majority of that funnel fell out. Those customers went elsewhere because they could not wait for our deliveries. The funnel that we have now of 3,433 orders is a completely rebuilt funnel that was rebuilt from the end of Q3 '24 to the end of Q1 '25. We now have much more clarity around locations and about timing to be able to install. At 883 locations, we can now install in 30 days. We have sufficient field resources, pluggable optics and service delivery coordinators to be able to provision 500 orders a month. With a 3,400 order backlog and funnel that represents with a 5% conversion rate about 160. So we have more installation capacity than orders that are ready to install. As we build credibility with customers, we will both see an uptick in the number of opportunities going into that funnel. Based on the sales forecast that we have, we anticipate that funnel to reach 10,000 from 3,430 by year-end. So in the next seven months. And while we are hopeful that the conversion rate monthly will be greater than the 5% we outlined, we're basing that on our three-month experience of actually being able to provision orders. The final point I would make really is the sequential pacing of growth. With 18% sequential unit growth, we did very well. However, 2% revenue growth was as a result of those orders installing near the very end of the quarter. And the issue was come January, we were ready to start installing at that point, it was 802 sites, and we grew that to the 883 at quarter end, and that number has continued to grow. But many customers who still wanted the services needed time to have their equipment ready to accept those services. And that resulted in most of the in-sell activity being back-end loaded at the end of the quarter. As we go into Q2 and beyond, we think that the pacing of installs will be more evenly distributed throughout the quarter. Hopefully, that helped clarify the question. And the goal is to be very specific with 500 capable installs per month, we think we will be hitting that target probably near the end of the year when the funnel reaches the 10,000 and the conversion rate remains at about 5%.

So next question is from the line of Alex Waters of Bank of America.

Maybe just first, can you maybe talk about the wavelength ARPU and kind of where you see that trending throughout the year? And then secondly, just on the data center monetization, can you talk just timing, scale and tie a bit some of these potential deals?

Yes. Thanks, Alex. Congratulation on your new more senior role. Our wavelength ARPU was just under $2,000 in the quarter, about $1,930. I think our basis now large enough, meaning we have enough visibility into the mix of 100-gig and 400-gig waves as well as contract duration and route length that using an ARPU of about $1,900 is probably a reasonable way to model the business. As the base continues to grow, the installed base, I think that number of the entire base will converge to something around $1,900 to $2,000 per wavelength. We are seeing a much higher uptick rate and higher capacity waves, 82% of our sales have been of 100-gig waves. That compares to the installed base in the industry of 55% of the base been ended waves. So the waves that Cogent has been selling tend to be at 100-gig with about 8% of sales being 400-gig-as compared to the industry base of about 3% at 400-gig. So -- and then in terms of route length, we now have nationwide or actually continental ubiquity and pretty good visibility into the orders that we have been booking. And as we get stability around the book-to-bill cadence, I think that $2,000 ARPU is a good modeling number. Now with regard to your second question of data center sales and monetization. We are continuing our work to convert those facilities and have just over 100 megawatts of power in what is now 24 facilities, we had originally targeted 23 that we are earmarking for sale or long-term lease. We have taken four of our letters of intent and move forward towards initial contract negotiations, we are still engaging with parties who are conducting site conditions, studies and engineering due diligence. We don't have an exact time frame, but we are highly motivated to sell this surplus capacity as it is not baked into our financial projections but would be the easiest way for us to quickly delever. Hopefully, that was helpful.

Your next question is from the line of Walter Piecyk from LightShed.

Dave, I first wanted to go back to that a couple of questions ago, just to make sure I heard you right. So -- because now there was this thesis that people circling wavelengths that it was just limited by your ability to execute. But I think you said you had a capacity but you're just waiting the customers to fill that capacity. And then if you can just talk about if you would installed at the start of the quarter, rather than 2.2% sequential growth, which was obviously impacted by the back-end loading, what that growth might look like so we get a sense of kind of unit growth conversion to revenue growth?

Yes. So if we had assumed that the orders had installed mid-quarter as opposed to end of quarter, our revenue growth rate would have been in the roughly 13% range. If we were fortunate enough that they had all installed at the beginning of the quarter, it would have been nearly 20% because the 18.2% unit growth actually understated the revenue growth because the ARPUs were actually slightly higher than the installed base. With regard to two very different metrics that we have given. The first metric is our ability to install. Our ability was limited by the repurposing of the Sprint network and we were constrained until the beginning of this year, where we had to do each installation that was done, the roughly 1,000 waves that we had sold on a custom basis and a limited number of sites. After the first of the year, we have the ability at 802 sites, which has now grown to 883 sites to install those services in 30 days. So that's on the Cogent supply side. We can do it in those sites at any of three speeds and deliver within 30 days. We had a funnel of orders that had been booked with customers over 1.5 years period as we were repurposing this network with literally no clarity to the customer on what we can install. It was not surprising to us, and we commented this extensively on our last earnings call that we were going through a process to purge that funnel of orders that customers had gone somewhere else where, they couldn't wait for us. Now we have rebuilt that funnel. We also know that many of those orders, the customers are not ready when we are ready. We expect that...

Dave, I don't want to say that you have to repeat what you already said. I heard all that. I'm just -- because it was just told to me rather than this install of 6% a quarter or whatever it is, which is how customers act that there was this old thesis the people are pitching that like, oh, it's fully about the supply and that they would just fill the supply as soon as that was available. So I think I heard you correctly. I was just trying to clarify that, which is now like that's the case. It's just customers have to be ready with their equipment and it's going to be -- and that's going to impact how that grows. Can I just move on to IP, which is like last quarter, had I think during the call, I specifically asked about what you could do each quarter and you said it's going to bounce back to $500,000, and that was February. I know in your -- that was late February. I know in your prepared remarks, you said you had basically disconnect some way that was misusing it. So with the disconnection like 700,000 IP addresses and can you talk about like what does someone have to do with their IP addresses that would merit getting disconnected. And I guess, similarly, should we just assume a shift down back to 500 a quarter going forward?

Yes. Okay. So one correction, Walter, I want to say 5%, not 6% is the conversion of the funnel and the growth in the funnel. So let me start with what someone has to do to be disconnected. And by the way, our acceptable use policy is clearly stated on our website. Typically, this will be one of three violations, either there is a government order saying that from any of the 57 countries that we operate in the world, that a customer announcing those addresses is doing something that, that government views is illegal. If that's the case, we immediately take it down. And it's upon the customer to resolve that issue with that government. The second area of abuse is typically copyright violations. And that is someone transmitting copyrighted information without the correct authority to do so. That is the case here that resulted in this fairly material take down. This customer was violating the digital rights management requirements of the U.S. government. And then the third potential area of abuse is if someone is using the addresses for a disruptive activity such as web scraping or spamming. Those are the three main categories of AUP violations. There were actually more than one customer in the quarter who had a significant digital rights management issue that resulted in a material decline in units, but also we were still able to grow revenues due to price increases. We absolutely anticipate to clearly answer your question, returning to a gross at a north of 500,000 incremental addresses a quarter, and it is difficult for us to predict or have been episodic periods in the past where we've had to take down blocks. This was a particularly larger instance this quarter.

So there could be additional churn going forward from this type of stuff, it's just too hard to protect for that. And then are you still going to 350 for the year and it's obviously going to be a pretty big set of a second half ramp to get there?

So we know that we have a very steep hill decline on EBITDA because we had a $104 million reduction in transit payments from T-Mobile. We feel comfortable we are continuing to improve our EBITDA and grow that and should be able to achieve the goals that we've outlined.

Your next question comes from the line of Chris Schoell from UBS.

Just a follow-up on your new long-term growth targets. What gives you confidence today to raise the target? Does any help breaking down that 6% to 8% revenue growth expectation by customer segment? And I believe in recent quarters for the legacy Cogent business, you provided a core growth rate for both corporate and NetCentric. What did core growth look like this quarter? And how do you expect that to evolve from here?

Thanks for your questions, Chris. So our growth was greatly impacted by absorbing Sprint's negative growth trajectory at closing and then accelerated by our attempt to purge undesirable services and revenues. That is how we have been able to actually grow cash flow while declining top line. This is on a customer-by-customer basis. We now have clear line of sight to the remaining services that we need to disconnect and we've been able to negotiate in some cases, customers agreeing to allow us to disconnect those services sooner than their contractual terms would allow us to. With that, we're comfortable that we'll be able to get through the vast majority of that intentional churn by mid-Q3 and then return to organic growth. We also now have higher confidence in the wavelength trajectory due to the realistic book-to-bill cycle and the quality of the funnel. Within the customer segments, we think that enterprise revenues will effectively be flat we think that corporate revenues should, on a consolidated basis, net of this intentional churn should be growing in kind of the mid single digits of 4% to 5%, and that represents both the on-net Cogent traditional corporate customer as well as the corporate customers that we have acquired from Sprint. Most of which that could be moved on-net have been moved on-net. And then the remainder are going to continue to be off-net as the locations are just not practical to bring on-net. And then finally, on the NetCentric segment, that is where the vast majority of the wavelength revenue will be a slide. And over 93% or 94% of the waves that have been sold to date, have been to NetCentric customers with the combination of the NetCentric IP growth and the wavelength growth, albeit a small percentage of that, we anticipate NetCentric aggregate growth with IP and wavelengths to be north of 10%. That combined growth rates and the fact that we have worked through the business that we want to exit should get us to an increased total revenue growth rate of 6% to 8%, compare that to Cogent's pre-acquisition of Sprint, where for an 18-year history, we had a compounded average growth rate of 10.2%. A big part of the reason why we required T-Mobile to enter into the transit agreement and subsidize us was both the losses and the realization that for a period of time while we were correcting the revenue mix in the business we acquired, we were going to suffer negative revenue growth. That is now clearly in our sights to turn positive. And then secondly, we are comfortable that we will also get better margin contribution than we had initially forecast. And quite honestly, from the day we've closed our margin contributions have actually exceeded our internal targets.

And then just do you have what the core growth rates were for corporate and NetCentric in the quarter ex all the grooming efforts?

So it's become harder and harder for us to kind of parse that out as we reprovisioned the customers. I believe the corporate segment grew between 3% and 4%, but that is not as a precise of a number as I would like to give you. And I think NetCentric probably grew at around 6% or 7% on a year-over-year basis.

Your next question comes from the line of Nick Del Deo from MoffetNathanson.

And Dave, I also appreciate the new call format. I thought that was helpful. So thanks for making those changes. First, going back to the SG&A line. I think you basically said that the entire sequential increase was due to normal seasonal items. It still feels like an awfully high increase even after taking those into account. I guess, like was Q4 SG&A depressed for some reason such that it wasn't a good jump-off point for things about Q1. And as we think about Q2 SG&A, how should we think about the roll-off of tax and audit costs and the sales meeting costs and those sorts of things?

I'm going to start then I'm going to pass it to Tad. Our sequential increase in expenses was greater this year than it was last year. And the primary reason for that is we have the entire Sprint employee base in Cogent summers for '24 and in '23, we only picked up those expenses on May 1. And all of the vacation accruals that those employees had were actually paid out in cash by T-Mobile as a condition prior to closing. So we did not assume those. Probably the biggest component of the sequential change in actually relates to the fact that people -- we have a use-or-lose policy and people user fiction in Q4 and then build an accrual going forward. But we can't give you a little more granularity on the components and kind of the variance between this year and other years.

So in the fourth quarter, there were no unusual items. I would say though bad debt expense for the fourth quarter of last year was unusually low. We're using it about 1% of our revenue and it was 1% of our revenues actually 0.8% this quarter. So that's about $1.5 million. The remainder of the $10.6 million increase is all these seasonal factors that Dave mentioned is 2.5% CPI on everyone's salary that had been here for a year, it's resetting of tax expenses, payroll taxes in the United States. That happens every year. It's annual audit fees that happens every year. Those -- and then the vacation accrual, which is not insignificant, sequentially, that's a $4 million change, and it's just a seasonal factor you -- people take vacation, of course, in the fourth quarter with the holidays. You're hitting the vacation of gold and not having to expense it because you've built that accrual over time, when you get back to the first quarter, you need to rebuild that again. So you have a swap in expense now building the accrual when you were charging the accrual in the fourth quarter. It's not insignificant. It's $4 million of the $10.6 million sequential increase, but it's normal.

It was just -- it's a normal activity. I would say it was just outsized because of the nature of seven months versus a full year Sprint employees. Does that help? Does that answer that?

Yes, it does. Are you willing to share anything regarding where that's going to land in Q2? Or is this a run rate to think about?

Slightly ticked down because more people will hit their FICA capacity and the payroll taxes typically decline.

On the IPv4 addresses, you said it was a large number that you took back. Are you willing to share the exact number?

Yes. Well, it was from more than one customer. So it can't be attributed to just one. But it was in the order of about 600,000 to 700,000 addresses and it's in aggregate that were taken back.

So your underlying trends are much better than they appear?

We've had this in the past, it's rare that you get this much in a quarter, but we don't predict when people do bad things. And Walt's trying to say it's going to go away. I can't answer that question because I can't tell you no one's going to highlight Turkish security laws and we get a takedown notice from the government of Turkey for a big block of addresses. I mean, stuff like that happens, and it is episodically, it was just more extreme this quarter than not. And fortunately, for us, we have such a good tailwind from the price increases, but the revenue still grew sequentially at 14.4%.

And then maybe one last quick one if it's okay. I think you said that a majority of your Q4 of the backlog and funnel from waves that you showed in Q4 fell out as part of the cleanup process. Maybe can you share what that number is? I'm just trying to get a sense of what your gross adds to the backlog and funnel were in Q1.

Yes. So between the stuff that fell out and stuff that installed starting in -- at the end of Q3. So we have visibility to starting to tell people we could get them firm delivery dates starting in January. Nearly 90% of the total funnel that existed at the end of Q3 '24 fell out. Only about 10% of that funnel which was about also 3,500 ended up either installing or carrying over and more of the installs that occurred, for example, in Q1 were things that were sold in Q4 and went into the funnel and the funnel is continuing to grow. And until we could actually install, I was extremely reluctant to give people kind of a book-to-bill kind of cadence and really even an ARPU. Now that we've got at least a couple of quarters where we could get people actual install dates with SLA commitments associated with it. We've got a lot more visibility. And I can look at the IP funnel and have great deal of clarity around this conversion rate on a monthly basis. I think what we've said 4% to 5% is conservative. And I'm hoping that conversion rate actually accelerates as the funnel grows and we demonstrate to customers we can really deliver. But the fact that we've actually delivered services now in 329 sites, I think just has earned us a lot of credibility and it's helped us build the funnel at a much faster pace than we were building it before.

But it sounds like, based on your commentary, you had at least a few thousand on a clean basis, a few thousand additions to the backlog and funnel in the quarter, which would I guess, in tandem with your expected install kind of gets you that 10,000 by year-end.

That is correct.

You have a question from Michael Rollins at Citi.

Two, if I could. So first, curious if you could discuss the slower Internet traffic growth year-over-year and what you're seeing also with regards to pricing and the implications of all of that as you look at Internet transit revenue performance within the NetCentric revenue going forward? And then secondly, are you seeing any changes in customer behavior in terms of sales cycles, decision-making since the beginning of April after the tariff announcement? And can you just remind us how a slower macro could impact your business performance?

Yes. Let me start with the traffic growth number. So if you look at open vault data, which is looking at traffic on the other side, which is end user total downloads. That has slowed to about 8%, which is in alignment with what we are seeing kind of on the supply side or upstream component. I think there are really three things going on concurrently. One, the rate of broadband adoption in countries that have decent access network capabilities has slowed. Two, the number of minutes of use per day has also moderated and third, the adoption of video has slowed. So at the beginning of the pandemic, we were at about 18% of end-user video consumption being streamed, that in five years accelerated to about 54%. It is going to continue to go up from here, and in particular, the pivot to more live event availability is helpful. But I do think with a larger video base, that application is maturing. I also think we're in a period when most of the network load for AI is directed at wavelengths because most of that network load is for training and not inference. But as the results of those large language models get distributed, that should present a new use case where users will use Internet connectivity more and bit volumes will go up. So we have seen this pattern of kind of oscillations in aggregate demand has applications change historically. And I think that will probably continue to be the case. But I think we're going to see over the next year or so, a reacceleration at least in bid intensity per user. Now in terms of pricing, we have seen the rate of price declines pretty consistent now for 25 years. That's at about 22%, 23% a year. And while there is always some short-term variability, that long-term trend line is pretty consistent, and I don't think that's going to change, even though there is less competition, the technology associated with manufacturing those bit miles is continuing to improve pretty significantly. So I think we'll see some more price declines for the industry and kind of more of a return to historical traffic growth rates. To your last question around tariffs, I'm going to actually answer it with kind of two different views. One is for our NetCentric customers, and this could be either wave or IP. Most of the time, they need equipment to accept those services. They're not happy. That equipment is more expensive. It does have some tariff load on it as at least a portion of it is coming from high tariff locations, but they still need it to deliver a service. So I think that's probably an initial shop. And in terms of materiality to the overall cost structure, I think it's much like Cogent. It's not material, but there's just a shock when that happens. But I would say the only thing on the tariff front that could potentially impact NetCentric business is if there is an effective content tariff, i.e., the movie tariff which doesn't exist. I have no visibility to how that's going to affect end-user demand. And then on the Corporate side, I do believe that a number of corporate users are just concerned with the overall macro situation and are we entering a period of reduced or negative growth and higher inflation. And for that reason, they're just being more cautious on long-term commitments. But again, because we sell a utility, I don't think that's going to have a material impact. I don't think we're in the kind of 2008-2009 great financial crisis level of paralysis. And our underlying corporate growth continues as we complete this blooming we feel confident that aggregate growth and corporate growth will be positive later this year.

Your next question comes from the line of Tim Horan from Oppenheimer.

Can you give us maybe just some timing on the data center sale and maybe expectations on price if you've received any? And then on the wavelength side, have you seen any competitive response?

Yes. I'm going to take those in reverse order, Tim. Probably the only competitive response was aimed at the hyperscaler segment of the wave market and the decision by at least one of our competitors after 20 years to agree to sell dark fiber, which is a potential substitute for wavelengths where the customer buys a dark fiber and then produces their own wavelengths on that. I think on the kind of delivery of wavelength services, we have not seen any change in pricing or delivery schedules and what we have heard from customers that have put orders into our funnel is that our pricing is good. We have not had to be maybe as aggressive as we thought we would have to be once we had the network fully configured, again, five months doesn't make a permanent trend. But we feel pretty good that the pricing that we're going to market has been well received by customers and viewed as adequately competitive to win share. I'm going to put it now to your data center question, and as I stated earlier, we have a handful of situations where we are moving from a letter of intent to contract. There is nothing that we can announce today. We are also continuing to do the work necessary to complete that data center conversion, but we think that will be completed in the next two months. We were pretty clear that we'll have that done by the end of Q2, and we are on track to do that. In terms of pricing, I would say the couple of parties that are negotiating leases are similar to our asset price. On the parties that are negotiating for outright purchase, there is a much wider dispersion at least one of the contracts is at the gas price, but the others are below that. We need to vet the ability of each of the parties to perform, and that's part of what is going on while we are refining economic terms and an LOI into a contract. Because we've never done this before, I remain reluctant to give a firm date on when we can do this. I think we're making good progress, and we'll get this -- we will monetize some of these. But it really does take two parties and the parties have to have the wherewithal to perform. So we're going through that process. And I know early in the process, you have the ability to tour one of the facilities that was a work in progress. And I can assure you, if you go back to that facility and Merchantville today, it would look very different than when you toured it.

But is something you think you can get done in like three months? Or is it more of a six-month type of negotiation?

It's really hard for me to answer. To be conservative, I would take the longer view, Tim, not the shorter view, just because we've got parties at the table, but we've got to flush through what are the conditions they need met. What is their time line that could close, how much is their earnest money deposit, what are the outs that they're looking to negotiate? And again, there's a fairly broad spectrum from sophisticated private equity to existing data center operators to more new business models and I just want to maximize value, and I think it's going to be more than three months. I don't know how much more, but I think that's probably aggressive to say that there's an actual closed sale. I think we are through our questions. The last topic I'm going to touch on just quickly is something that affects me personally, and that is that I have, due to Cogent stock volatility had to substantially increase some of the shares that I have pledged to pay taxes over the years. I have not increased any borrowing, but I do want shareholders to be aware that I'm trying to be as transparent as possible. I've been forced to inject money into my real estate portfolio, and that is continuing. I want to personally thank everyone. Hopefully, this new format was more efficient. We did get everybody's questions got answered and got it down to 1 hour 15 minutes, and I look forward to seeing you all in person soon. Take care. Thanks. Bye-bye.

This concludes today's conference call. Thank you all for joining us. You may now disconnect.