Cogent Communications Holdings Inc (CCOI) 2007 Q3 法說會逐字稿

Good morning and welcome to the Cogent Communications Group Third Quarter 2007 Earnings Conference Call. As a reminder, today's conference is being recorded and will be available for replay at www.cogentco.com. I would now like to turn the call over to Mr. Dave Schaeffer, Chair and Chief Executive Officer of Cogent Communications. Please go ahead, sir.

Thank you, Melissa. Thank you and good morning. Welcome to our third quarter 2007 earnings call. I am Dave Schaeffer, Cogent's Chief Executive Officer. With me on today's call is Tad Weed, our Chief Financial Officer.

We are very pleased with the results from this quarter. Our third quarter revenue, EBITDA and loss per share were either within the range or better than the ranges that we had presented to you in our Q2 2007 earnings call when we were giving our forward estimates. We expect our 2007 year revenue, EBITDA and loss per share to either be better or within the ranges for fiscal year 2007 as we originally presented them to you back on our Q3 2006 earnings call.

Since our last earnings call, we have returned $4.5 million to shareholders as part of our second $50 million share buyback program, which was announced in August of this year. Under this program, we repurchased 194,000 shares with an average price per share of $22.85. I should point out that this is the $50 million buyback program that was put in place by our Board in addition to the initial $50 million buyback that was completed as part of and in conjunction with our 2007 $200 million convertible debt transaction.

During the quarter, we continued to increase our footprint. We added another 30 buildings to our network and expanded the network into three additional countries. During the quarter, we also significantly increased our sales resources and we added 23 net representatives to our sales team. As a result, we have 179 quota-bearing reps selling Cogent services in North America and in Eastern and Western Europe as of today.

We also achieved another significant corporate milestone in the third quarter by adding our 10,000 on-net customer connection to our network. Finally, we again exceeded our provisioning guarantees and our service level guarantees by delivering our services to customers with an average on-net provisioning cycle of only 12 business days. Again, I would personally like to thank the entire Cogent team, in particular the sales force, for their efforts in helping us to achieve these results. Involved in our provisioning and network expansion efforts, many of our team members also deserve a great pat on the back.

Throughout this discussion, as in the past, we will continue to focus on results and impacts of our on-net business, which is in fact Cogent's growth business. We will also highlight several other operational statistics, we believe, which will demonstrate our increasing market share, expanding scale and the operating leverage of our business. I will review certain operational highlights and also outline expansion plans and growth opportunities. Tad will provide additional details on our financial performance. Tad will also walk us through our expectations for fourth quarter 2007, updated and reaffirmed expectations for full-year 2007, and provide our initial guidance for fiscal year 2008. Not to steal Tad's thunder but, as Tad will present, we expect to achieve another significant corporate milestone in 2008. We anticipate having net income by the end of 2008, but this will not be taxable due to our significant NOL.

Following our remarks, we will open it up for questions and answers. Now, I would like to turn it over to Tad to read our Safe Harbor language.

Thank you, Dave and good morning everyone. This third quarter 2007 earnings report and this earnings conference call to discuss Cogent's business outlook will contain forward-looking statements within the meaning of Section 27A and Section 21E of the Securities Act. 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. You should also be aware that Cogent's expectations do not reflect the potential impact of mergers, acquisitions, other business combinations or financing transactions that may be completed after today. Cogent undertakes no obligation to release publicly any revision to any forward-looking statement made today, or otherwise update or supplement statements made on this call. Also during this call, if we use any non-GAAP financial measures, as defined by the SEC in Regulation G, you will find these reconciled to the GAAP measurement in our earnings release and on our Web site at cogentco.com. I would like to turn the call back over to Dave.

Thanks, Tad. Hopefully, you have had a chance to review our earnings press release. As with previous quarters, this press release includes a number of historical quarterly metrics. These metrics are also added to our Web site. Hopefully, you will find these metrics helpful and informative in understanding our financial results and trends within our operations. As always, if you have any suggestions on how we can either refine or add to these metrics, please let us know.

Our third quarter 2007 revenue of $47 million was at the mid-point of our guidance of between $46.5 million and $47.5 million. EBITDA as adjusted of $11.7 million was also within the range of our guidance which had been between $11.5 million and $12.5 million. Our loss per share at $0.12 was better than our range of guidance of $0.18 to $0.23 per share.

We continue to be pleased with the growth trends in our on-net business. We are continuing to increase our investment in our sales, marketing and network activities to increase the number of on-net customer connections and ultimately the on-net revenues from those connections. Our on-net revenues grew at 6.7% sequentially quarter-over-quarter from Q2 2007 to Q3 2007. This growth rate was greater than the increase of 6.5% we experienced from Q1 2007 to Q2 2007, but was slightly below our on-net target growth rate of 7% to 9%. Net additions of customer connections increased 7.4% sequentially quarter-over-quarter from Q2 2007 to Q3 2007. Our on-net revenue per subscriber or ARPU decelerated its rate of reduction to 3.5% from Q2 2007 to Q3 2007 from the historical decline rate of 5.1% in Q1 2007 to Q2 2007. During the quarter, we saw larger accounts leading to a deceleration in the rate of ARPU decline.

During the quarter, we again extended our average contract length. I think this represents a significant vote of confidence in Cogent by its customers. Our average contract length increased in the quarter by approximately 5% and as these customers take longer-term contracts, we are ensuring a more stable and long-term revenue growth trend. Overall, revenues for the quarter grew to $47 million. This represents a 4.1% sequential quarter-over-quarter increase from the $45.1 million of revenue in Q2 and a 23.8 % year-over-year increase from Q3 2006 when revenues were $38 million.

Traffic on our network grew by approximately 2% sequentially quarter-over-quarter and approximately 66% year-over-year Q3 2006 to Q3 2007. Our on-net revenue and customer connections continue to increase. On-net revenues increased by $2.4 million sequentially quarter-over-quarter or 6.7%. On-net revenues increased by $10.2 million year-over-year or 37.1% on an annual basis Q3 2006 to Q3 2007. On-net revenue continues to increase as a percentage of our revenues, growing from 78.2% of service revenues in Q2 to over 80% of our Q3 revenues. Over 90% of our new sales in Q3 were on-net services. This is a greater percentage of sales than we have historically been experiencing as a percentage of our on-net revenue growth, again demonstrating the acceleration in growth in that business. Our on-net customer connections increased by 7.4% for the quarter from a little over 9,700 customers to over 10,500 on-net customer connections by the end of Q3 2007.

Revenues from our off-net business declined approximately 2.3% for the quarter. And non-core revenues -- these are revenues for services which we no longer sell, declined sequentially at approximately 16.5%. Pricing trends in the industry remain intact. Cogent's placing policy remains unchanged and we are committed to being the industry price leader. Our pricing of our most widely sold product, that is a $1000 a month 100 MB connection on a month-to-month contract, or $10 per MB remains unchanged since the Company's inception. We do continue to offer discounts for longer-term contracts as we have mentioned on previous calls and we have had an increase in the number of customers availing themselves of these contract discounts. In full, our average contract term extended as I mentioned earlier by approximately 5% within the quarter. We also guarantee customers that any on-net service will be provisioned in 17 business days or less.

In 2007, we delivered on this guarantee with an average on-net provisioning cycle of less than 12 days. On-net ARPU was approximately $1,238 in Q3 2007 as opposed to $1,283 in Q2, a decline of approximately 3.5 %. This decline was lower than the rate of decline that we experienced on a sequential basis in Q1, when that decline was about 5.1%. Off-net ARPU actually increased from $807 in Q2 to approximately $841 in Q3, and this was a result of selling larger off-net connections, an increase in both T3 connections as well as off-net Ethernet services, where we purchase the local [group] from generally an incumbent telephone operator.

EBITDA was approximately $11.7 million for Q3 2007. This represents an approximately 6.3% increase from the $11.1 million that we experienced in Q2 of '07. Our EBITDA and EBITDA margins continue to expand and increase due to the [gross margin] expansion of our business. Our EBITDA margins expanded 50 basis points from 24.5% in Q2 to 25% in Q3. Our gross margins decreased marginally from 52.5% in Q2 to 51.6% in Q3. Gross margin decline was primarily due to the increased fixed costs of operating our network and the additional costs that were assumed by acquiring six data centers at the end of last quarter, as we had mentioned. These data centers were predominantly empty and we picked up the cost of these centers as we begin to populate them with our customers.

Our direct incremental on-net gross margin continues to be nearly 100% and our direct incremental on-net EBITDA margin remains at about 95%. We continue to expand our network. We expect to meet or exceed our goal of adding 100 buildings to our network in 2007. In the third quarter, we added 30 buildings to the network and have added 82 buildings to the network year-to-date in the nine months ending September 30 of '07. We expect to add additional buildings and meet or exceed that goal. During the third quarter, we also fully integrated the six new data centers that we acquired at the end of the previous quarter and conformed those data centers to Cogent's standard configurations and operating procedures.

Since the last call, we have also extended our network by securing additional fiber in new markets and countries. We continue heading in Europe further east, extending our network significantly into Bucharest, Romania, into Helsinki in the North, Manchester in Britain, Edinborough, and Glasgow, and Scotland, and here domestically adding new cities to the network such as Stamford, Connecticut and Providence, Rhode Island. We continue to evaluate additional expansion opportunities, being very judicious about deploying that capital and focusing only on the highest traffic growth locations.

Tad would like to now cover some additional details on our Q3 operating results. Tad will also provide additional expectations for Q4 of 2007, reaffirming and tightening our full-year expectations and also providing our initial guidance for calendar and fiscal year 2008. I would now like to turn it over to Tad.

Thanks, Dave and again, good morning to everyone. And I would like to also personally thank and congratulate our team on their very hard work and performance this quarter. On loss per share, our loss per basic and diluted common share as Dave mentioned was $0.12 for the third quarter. We did have some non-recurring, below the line, non-operating gains, primarily related to a $ 2.1 million lease settlement that was acquired in our Verio acquisition back in 2004 and these gains reduced our net loss by $2.6 million in total during the quarter and as a result, reduced the loss per share by about $0.05. So without these gains, our basic and diluted common share loss for the quarter would have been $0.17. That recurring loss of $0.17 was a $0.01 better than our range of guidance issued to you of $0.18 to $0.23 a share.

Non-cash equity-based comp expense, including the impact of 123R, increased our Q3 2007 loss by about $3.1 million or about $0.07 per share. Regarding foreign currency, the euro to dollar conversion rate positively impacted our Q2 to Q3 comparable revenues by about $200,000. The average euro rate for Q2 was $1.35 and $1.37 for Q3, and I am sure as many of you know that euro rate has typically increased during this quarter. As a reminder, about 22% of our revenues are based in Europe and about 8% of our revenues are in Canada.

On capital expenditures, as we may have mentioned on prior calls, we are accelerating our rate of adding buildings to our network, which is impacting CapEx. CapEx was $9 million for the third quarter, slightly less than $9.5 million for the second quarter. We added the same number of buildings during Q2 and Q3, that being 30 per quarter. We have added a total 82 buildings for the network in the nine months ended September 30, so well on our way to our goal of 100.

On our balance sheet as of September 30, our cash and cash equivalent totaled $180.2 million. That is all invested in money market bonds. Our second $50 million approved stock buyback program, the Board approved in August 2007. Dave mentioned we purchased 194,000 shares for $4.4 million, so an average price of $22.85 per share. Capital lease IRU obligations were $90.9 million at September 30 and $7.3 million of that amount is a current liability and these lease obligations are being paid over a remaining weighted average life of approximately 12 years. Days sales outstanding on accounts receivable was 40 days at Saturday 30 which is our bogey rate. This compared to 32 days at June 30, which was well ahead of our bogey rate. We were impacted by the calendar where the last two days of the month fell on a weekend, so our -- some September cash fell into October and we are well back into below 40 today.

Operating cash flow increased by $1 million from the prior quarter, primarily from the increase in EBITDA. Positive cash flow from operations was $11.3 million for the third quarter compared to $10.3 million for the second quarter.

I am now going to provide guidance for three periods, for the fourth quarter, update 2007 based upon that fourth quarter guidance, and provide initial guidance for 2008, and these are based upon the current and expected run rates of our business, planned increases in sales and marketing programs, and anticipated network expansion projects. The guidance covers each line item on our P&L down to the net loss level, so hopefully you will find this very helpful in modeling.

Updates to our 2007 guidance are mostly a narrowing of previous ranges and our projected 2007 revenue EBITDA and loss per share are all expected to be within or better than our standing 2007 guidance. So here we go on the fourth quarter, we expect the fourth quarter total revenue to be between $49 million and $50 million. The components of that revenue, we expect on-net revenue to increase quarter-over-quarter by 7% to 8%. We expect off-net revenues to be flat or to an increase of approximately 5%, so off-net to be flat to an increase of approximately 5%. And non-core revenues to decline quarter-over-quarter by approximately 15% to 20%.

We expect our gross margin percentage to increase and to expand by approximately 200 basis points to about 54% for the quarter. We expect EBITDA as adjusted to grow to between $13 million and $14 million. That will put our EBITDA margin expanding by another 200 basis points to about 47%. We expect SG&A, excluding the non-cash equity-based comp expense that's classified as SG&A in the 10-Q format, to be 28% to 29% of our fourth quarter revenues. We expect that non-cash equity-based comp expense in total to be between $3.5 million and $4.0 million for the quarter. We expect depreciation and amortization to be between $17.5 and $18.5 million. And we expect net interest expense, so interest expanse net of interest income to be between $1.0 million and $1.5 million of an expense. These results are expected to result in a loss per share of between $0.17 and $0.22 for the quarter. Net loss per share assumes $47 million weighted average outstanding. As a reminder, if we purchase additional shares, that will lower the shares outstanding and has the impact of raising our loss per share.

2007 guidance that I am refining and updating is as follows. Our total revenue now is expected to be between $184.5 million and $ 185.5 million for the year and that will result in the following growth rate for the components of revenue -- on-net revenue to grow to between 38% and 40%, and I think our guidance there was 35% to 40%. Off-net revenues to decline by between 5% and 10%. Non-core revenues to decline by between 25% and 30%.

We expect our full-year 2007 gross margins to be approximately 52.5%. We expect EBITDA for 2007 to be between $46 million and $47 million and we expect SG&A to be between 28% and 29 % of our revenue, again excluding comp expense classified that way in the 10-Q. And non-cash equity-based comp expense to be between $10.5 million and $11.5 million for the year. Depreciation and amortization to be between $66.5 million and $67.5 million, and net interest expense to be an expense of approximately $4 million to $5 million. These results are expected to result in a loss per share for the year of between $0.55 and $0.75. Net loss per share assumes a weighted average shares outstanding of 47.7 million -- again, if we buy additional shares between now and the end of the year, that will reduce the shares and could increase the loss per share. Total CapEx is still expected to be under $30 million for 2007.

Now, our initial guidance for 2008. Total revenue is expected to be between $225 million and $235 million. If we take the midpoints of '07 and '08, this is an increase of about 25% from '07 to '08 in total revenue. And the following growth rates for the components of revenue -- on-net revenue to grow between 30% and 33%. Off-net revenue to be flat to an increase of 5%. And non-core revenues to decline by between 35% and 40%.

We expect our 2008 gross margin to increase to approximately 60% for the full year and we expect EBITDA as adjusted to growth between $75 million and $80 million. At the mid-points of the ranges, this is an approximately 67% increase in EBITDA from '07 to '08. We anticipate our 2008 EBITDA margin to expand to approximately 34% for the year. We expect SG&A to be between 26% and 27% of total revenues for the year, again excluding equity-based comp. We expect that equity-based comp to be between $14.5 million and $15.5 million. We expect depreciation and amortization to be between $62 million and $63 million, and we expect net interest expense to be approximately $11 million to $12 million.

Those of you who have seen our press release, our 2008 interest expense guidance includes about $9 million of additional non-cash interest expense that we expect to have to record under a proposed accounting standard that Cogent and companies with similar convertible notes are expected to have to adopt on January 1, 2008, and the pronouncements entitled APB 14-a called "Accounting for Convertible Debt Instruments That May Be Settled in Cash Upon Conversion" and the pronouncement essentially requires that our convertible feature of our notes be bifurcated from the notes and separately valued. It looks like that value is about $80 million. Well, that will result in an increase in the debt discount and a corresponding increase in additional paid-in capital for the balance sheet entry, if you will and due to that increase in debt discount which needs to be amortized, there will be additional non-cash interest expense from amortizing that through June 2014, which is the earliest put date. There is obviously no change here in cash flows. So the pronouncement is currently proposed. We will also require that we and all those impacted by this restate the 2007 results as well and show additional interest expense from the amortization of the discount. That amount for us is about $5 million and the notes, as a reminder, were issued in June 2007. So, that's the additional interest expense from June '07 through the end of the year.

We are including this because in consulting with our auditors, the approval of this pronouncement appears imminent. So, we thought it was appropriate to include this in our 2008 guidance. I apologize for that long explanation, but that is a very material item should it be approved.

We expect our 2008 loss per share to be between $0.15 and $0.30 per share. As Dave mentioned, we anticipate achieving another corporate milestone in 2008 and we expect to have positive net income by the fourth quarter of 2008. We do have significant NOLs that will offset this positive net income. Our EPS estimate assumes $47 million weighted average shares outstanding for the year and any buybacks would reduce those shares. And we expect our 2008 CapEx to be similar to our 2007 rate and adding another 100 buildings and to be approximately $30 million. I will now turn the call back over to Dave.

Hey thanks, Tad. As we mentioned earlier, we expect to continue to expand our footprint. We did add 30 buildings to our network in the most recent quarter as well as the inter-city route extensions that we discussed. We have added 82 buildings during the first nine months of 2007 and are on pace to exceed our goal of adding 100 buildings in 2007, and we anticipate adding another 100 buildings to our network in 2008. As of September 30, 2007, we had 1,189 buildings directly connected to the network, representing over 520 million square feet of rentable office space out of an addressable inventory in North America of about 6.2 billion square feet. We also have approximately 25 buildings that are under construction and in the process of being connected to our network as we speak.

In terms of our sales force, as of today we have 179 quota-bearing reps selling our services in North America and Europe. These 179 reps equate to 152 full-time equivalents based under our ramping protocol. And just to remind people on this call, basically when a rep is hired, in the first month, they have no quota; one-third, two-thirds, and then full quota responsibility on a sequential monthly basis. We began Q3 2007 with 143 representatives and ended the quarter with 166, for a record net quarter add of 23 sales reps. We began Q3 2007 with 122 full-time equivalents and ended the quarter with 134 full-time equivalent reps, representing a increase of 12 full-time equivalents over the quarter. Our sales reps continue to maintain their productivity. While productivity per FTE for the quarter was below our target of 4, we also have been focusing our reps on some larger and more lucrative opportunities and also building [a funnel] to ensure their continued success. We continue to be very disciplined about monitoring the productivity of our reps and removing those reps who do not meet our productivity requirements.

We continue to be quite comfortable with our 2007 and 2008 revenue plans. Our business remains entirely focused on the Internet and therefore, a beneficiary of the huge trends in our economy that are migrating traffic to the Internet from other forms of communication and also proliferating the amount of information that is available.

Due to the successful recruiting, we expect now to increase the number of sales reps, quota-bearing reps that we have, by the end of the year to 190 from our previously announced goal of 180. We see significant opportunities to add these reps and still have ample addressable market. We also expect to continue to add sales reps in 2008 to address both our existing footprint and our expanding footprint. As announced on the previous call, we expect to increase the number of reps by 50 in 2008. As a result, this will take our sales rep count from the previously announced 230 reps at end of '08 to now 240 reps. We continue to be very judicious about hiring only the best candidates who apply for jobs at Cogent and then to train and incent these people to keep them productive and to manage out those individuals that don't meet our strict requirements for productivity.

We continue to expand our footprint. We now have over 10,000 miles of metro fiber, 249 operational rings and over 100 markets. We have over 26,000 miles for terrestrial intercity fiber. We are today the most interconnected network in the world, today directly connected to over 2,270 other networks. Over 350 of these networks are settlement-free peering partners and over 1,920 networks that connect to Cogent are Cogent customers who pay us for connectivity. We believe our network has substantial capacity to accommodate our future growth plans. We are currently utilizing a little bit less than 22% of the width capacity in our network.

So in summary, our on-net revenues, our growth business, continues to grow as an increasing percentage of our total revenue and traffic continues to increase on our network. We are surpassing our provisioning goals of 17 days for on-net services, allowing customers to experience 12 days from order placement to actually receiving Cogent services in any of our on-net footprint. We are well on track to exceed our footprint expansion plans and our rate of adding buildings to the network. We continue to have a strong balance sheet, with less than $20 million in true net debt. We continue to be focused on judiciously expanding our network into the highest traffic locations, both domestically in existing markets and new markets as well as internationally. We will continue to invest in our sales force and continue to remain focused in our marketing efforts. And then finally and maybe most importantly of all to our shareholders, under our 2008 plan, we expect to generate in excess of $1 per share of free cash flow for the benefit of our equity shareholders. And we will continue to look to return that value to our shareholders.

Now, I would like to open the floor to questions.

Thank you. (OPERATOR INSTRUCTIONS) We will go first to Tom Watts with Cowen & Company.

Hi, Dave, how are you?

Good, Tom.

Congratulations on the quarter. Could you -- couple of things, one, could you just give a little bit more color, you had hoped for the 7% to 9% on-net revenue growth in the quarter and why didn't we quite get there even though things are going well? And then second, clearly you are having phenomenal expansion on your sales force, are you seeing different products being sold or how is it changing what is coming in the door? Particularly you mentioned the productivity was lower this quarter and then -- is that going to rebound and how did that change product mix?

Sure. First of all, let's start with the sequential on-net revenue growth. You are correct. While we were hoping for a little better sequential growth, we did see acceleration from the previous quarter. I think probably the biggest contributor to our traffic growth -- to our revenue growth not hitting our target of 7% to 9 % was the fact that sequential traffic growth was only a couple percent for the quarter. So we are getting less of an uplift from existing customers buying bigger connections, but we were continuing with a great deal of success signing brand new customers and new logos. And I think this is a result of some of the seasonality associated with Internet traffic growth, particularly as it becomes more of a kind of content delivery and entertainment media, there tends to be a little bit more seasonality with slower growth in summer months. Also as a backdrop, several of our competitors have recently released results in which they have experienced sequential quarter-over-quarter declines in their Internet businesses, where we are experiencing 6.7% sequential growth and as Tad indicated in our guidance for next quarter, as we have seen traffic increase just in the month of October at a sequential rate of about 5% month-over-month, we are quite comfortable with the 7% to 8% on-net revenue growth number that we are forecasting for fourth quarter and the full-year growth number in on-net revenue of 30% to 33% for 2008 also is a direct result of these long-term trends that we are seeing. To your second point about sales force product mix, we continue to hire a sales force at an unprecedented rate at least for Cogent and make sure that we continue to manage out those individuals that are less productive. Our product suite remains very focused though. We sell no other products other than Internet connectivity and the co-location space associated with that Internet connectivity. So, we are a direct beneficiary of the growth in Internet traffic and migration of traffic from other types of networks to the Internet.

What we do not sell however are value-added services in competition with our customers, so the sales force product mix that they have in their bag remains constant. What we have instructed sales people to do is spend time building those funnels, getting out there and talking to enough customers, so they can selectively pick some of the larger accounts to be working on and close. That did have some impact on productivity. We have seen productivity bounce last quarter in Q2, it was actually the highest in the Company, sales grade of 5.8, but that was coming off of a quarter where in Q1, it was 4.5%. So, we have used these four units per rep as somewhat indicative of this kind of fluctuation on a quarterly basis. It also is a direct result of these reps continuing to focus on building their funnels.

Perhaps I should have been more accurate. I think the last quarter you said a lot of the newer reps were selling smaller contracts because those were easier. Have you seen any -- it sounded like you try to refocus in this quarter, have you seen progress on that front? Should we expect that going forward?

Yes, we did. We sit down with the sales force on a weekly basis and monitor their productivity, we look at their funnels and also try to give them council to make sure that they meet their quota requirements and remain employed to Cogent. And in doing that we really emphasize that people needed to work on building bigger funnels and having the luxury of being able to work on bigger accounts and we actually saw that in the deceleration of the rate of ARPU decline, coupled with the fact that contract length for the quarter across the entire base increased by 5% resulting in lower ARPU because of the discount. So, we absolutely are seeing a pick up in larger sales and we expect to see that even accelerate more in the upcoming quarter, particularly as some existing customers will come back and increase their traffic purchases as their traffic requirements increase.

Okay. Thanks very much.

We'll go next to Jonathan Schildkraut with Jefferies & Company.

Thank you for talking the questions. Just two questions, first, if you could tell us what some of the costs associated with the launch of the new data centers, the six new data centers was, could we get a sense of what the incremental margin might have looked like without that? And secondly, what percent of your customer base is now under long-term contract?

Sure. So, let me start with the data centers. Adding six new data centers to our footprint that were virtually vacant and we did complete that acquisition at the end of second quarter in late June. So, we picked up all of the expense associated with those centers. The direct cost of carrying those centers was approximately $200,000 a month or about $600,000 for the quarter. We also picked up probably another $300,000 in increased operating cost of existing centers due to increases in power costs and rental costs within those centers. We also experienced a slight uptick in capital expending as a result of bringing those centers up to Cogent's standards. Now, while these were completely built out centers, they had been virtually vacant at the point of acquisition and did not have, say, the same security and environmental monitoring systems that our other 28 centers have. We did complete the integration and the standardization of those facilities, but probably most importantly, we are beginning to sell customers into those facilities and beginning to increase the occupancy. If you look across our entire footprint, we are still running below 40% occupancy within our data centers and also rack and power as a percentage of revenues did increase for the quarter from about 3.4% of revenues to a total of about 3.8%, so we are starting to see some of the uptick from the growth that we are getting in these date centers. I am going to let Tad then take the margin question.

Well, on the contract length -- the contract length question, so if you look at 2007 installed orders through the end of the second quarter, the weighted average was about 10.7 months, and that increased to about 11.3 months in the third quarter. Another way of looking at that is contracts that are one year or greater, so we do have some two year contracts and longer, where about 75% through the second quarter and then that increased to 80% of contracts in the third quarter. So, again, the trend is more longer term contracts.

Thank you.

We'll go next to Jonathan Atkin with RBC Capital Markets.

Yes, I was wondering if you could refresh us on the mix between corporate and net-centric revenues, the portion of corporate versus net-centric among your incremental revenues, and then the new buildings that you are adding, are they predominantly data centers or office buildings?

Well, John, first of all, in Q2 our mix was roughly 56% net-centric revenue and 44% corporate. That mix actually changed in the third quarter where we are about almost 57.6%, I think, net-centric and just about 43% -- 42.3% corporate in our revenues. And in terms of new buildings, here in North America, the majority of the buildings we are adding are corporate. In Europe, virtually all of the buildings that we are adding are net-centric or data centers. That mix remains at -- for incremental buildings about 75% of the incremental buildings are corporate, about 25% of the incremental buildings are data centers, with Europe kind of leading the way there.

And then, Tad mentioned 22% Europe, 8% Asia, and maybe if we fast forward one year, if the mixes of domestic versus international -- 8% Canada, I am sorry, is the mix of domestic versus international likely to be similar as we exit next year?

Yes, we don't anticipate a difference in the mix and the growth rates are relatively similar across the region. So, don't expect that 22% and 8% to vary significantly.

And then, when you connect to a data center, how many other bandwidth providers are also connecting?

It is very lightweight, but data centers by their nature have multiple providers. I would say that on the typical data center that we connect to, there are seven or eight other unique networks connected to that facility. Whereas, when we connect to a corporate building, there is generally only one or two other networks typically connected to that corporate building.

Thanks very much.

We'll go next to James Breen with Thomas Weisel.

Thanks. Given where your guidance is for revenue for next year, can you talk about the trends in the off-net side and non-core revenue businesses to give us a better picture of the growth in on-net business? Thanks.

Yes, hey, sure Jim, it's Dave. As we pointed out, we expect our on-net revenue growth to be between 30% and 33% of revenues, sequentially that business to grow. We also indicated that for the first time incremental sales are, at cost, 90% being on-net whereas our kind of historical run rate was only about 80% of new sales being on-net where 20% were off-net and now we are more like 90-10. We do expect to see a continued deceleration in the rate of decline in the off-net business. That will be helped by the availability of higher capacity in the local groups from the incumbent operators and the increased availability of Ethernet services. I think with these trends we expect to see relatively flat. So, I think Tad said, kind of flat to down 5%, our off-net revenue growth. We do expect to see the non-core continue to dissipate quickly. Remember included in non-core is our managed modem service. This is really the provisioning of wholesale dial-up services to companies like EarthLink and United Online, which are continuing to see continued defections from dial-up to broadband. And then, finally, or the most important part of our business, that being on-net, we will see growth from both new logos buying from Cogent as well as existing customers particularly net-centric customers taking more capacity. I think all of these factors together give us a great deal of comfort in seeing our total top line revenue grow from about $185 million this year to about $230 million at the mid points of our guidance for next year.

Great and just one follow-up on the traffic. You said that the same store sales -- the customer growth slowed a little bit this quarter, was that broken evenly between the net centric versus your corporate customers?

Actually, the corporate customers continue to grow at a pretty fast pace, but they account for only about 4% of our traffic. 96% of the traffic on Cogent's network comes from that 57% of our revenue that is coming from the net-centric customers. Those customers typically buy on a per mega bit and their additional spend comes from consuming more mega bits, whereas corporate customers, even though they may grow don't generate more revenue for Cogent, simply because we sell in increments that are typically larger than they can consume. So traffic growth disproportionately impacts the net-centric part of our business in terms of revenue growth.

Great, thank you.

We will go next to Tim Horan with CIBC World Markets.

Good morning guys. Thanks a lot, good quarter.

Hey, thanks Tim.

Dave, the stock looks like it is weak this morning and when you look at your guidance, it looks like your looking for more like 8% sequential growth going forward and that's just on the surface seems aggressive, particularly where the stock is. Why wouldn't you guide to more or like 6%, 7% like you are seeing now and if the upside comes, that's fine. And in that vein, could you maybe talk about what you are expecting in ARPU the next year, you might have touched on it, but how does that kind of revenue growth break down on the on-net front and then I just have one follow-up. Thanks.

Yes, sure, let me start with the growth. So we did see an acceleration in our sequential quarter-over-quarter growth to 6.7%. If you compounded that out, that would put us in the high 20s, probably 28%, 29%. We do see acceleration in that sequential growth to the kind of 7% to 8% this quarter. And if you look at kind of the long-term trends, we have the average 9% to 10% sequential on-net revenue growth over the past couple of years. We continue to believe that the selling environment and our competitive advantage is as strong now as it's ever been. So, we feel very comfortable that we will eventually return to that 9% to 10% long-term trend. We did guide to 30% to 33% in a sense of conservancy in the sense of giving us some ramp to do that. You will see some fluctuations quarter-over-quarter particularly as traffic growth fluctuates quarter-over-quarter. And we continue to increase as a percentage of the world's Internet traffic, so we become much more of a proxy for the entire Internet, and therefore our ability to kind of outperform the Internet tends to compress. We do believe however that there are a number of significant social and business trends that are going to continue to drive traffic growth at the same historic rates that it has experienced for the past 15, 18 years and from that, we feel very comfortable with the revenue guidance. And again, we feel that the 7% to 8% that we guided to in Q4 is extremely achievable based, particularly on early results that we have seen from October traffic growth.

Thanks for that. On a longer term, on note, one of your main benefits versus your peers is you have not had to kind of re-price your base. What we are hearing from data center price is that you can buy service in the $20 megabit range, kind of declining 25% a year. I know in the corporate environment, the office building environment, it's more like $40 a megabit. But if you extrapolate those trends out, at some points in the next three years you are probably going to start re-pricing your base. Can you -- and that could at that point slightly impact the revenues. I know it ways out, but can you talk about your plans, what you are going to do when that comes to fruition? Thanks.

Hey, sure Tim. I'll take that and I'll let Tad touch on your ARPU question, which we still have yet to answer on. If you look at our effective price per megabit, with today, 80% of our contracts being one year or longer in length. Our effective price per megabit today is about $9 per megabit based on the pricing discounts that we get for contract term. Remember $10 a meg is month to month, $9 a megabit is for one year contract and $8 megabit is for two year or longer contract. Also, the price event you referred to in the 20s in some data centers is by a few more aggressive carriers that are probably losing money and are also only offering those prices to extremely large customers. We have a very different model in which we offer flat pricing, not based on volume commitments but rather on contract terms because our belief is that there is a huge advantage for Cogent in having a lot of new business models choose Cogent where a few of those models would become very successful and others may not. So, we think in the long-term our pricing strategy is the right one. In terms of eventually re-pricing our base, again, we have a policy that says any customer can present us an invoice from a competitor and they can either take the standard Cogent price or undercut it by 50%. Now, I will admit we did see a slight increase in that, take grade on that program last quarter. And as I mentioned in Q1, I think we had a grand total of one instance and we had -- where we had to do that. I think we actually had two instances in these most recent quarters. So, it's still relatively insignificant. But if you really put your crystal ball in front of you and look far enough out into the future, the Internet is very deflationary, prices will fall. We are uniquely positioned to be the price leader. Remember, we are utilizing less than 22% of lit capacity. And while we do not anticipate any material impact from re-pricing our existing base for the foreseeable future, eventually the market average price will get to Cogent's price, but there will be a whole lot less players in the market and that's probably a good segue to Tad's ARPU response.

Yes. The overall trend on ARPU basically, what we experienced this quarter versus last quarter we see continuing through the fourth quarter and into '08, so that being an increase in off-net ARPU and a decline in the on-net ARPU, however, the decline is expected to occur at a slower rate. We don't and haven't in the past given the exact rates because we can essentially hit the revenue target with a different mix of units, but I do think the overall impact will be and the trends will continue with off-net ARPU increasing and on-net decreasing but at a slower rate.

Thanks a lot guys.

We will go next to Tom Seitz with Lehman Brothers.

Thanks for taking the question. Couple, I guess, follow-ups. If you have got 80% of your customers on long-term contracts today, can you tell us what's baked into the guidance for next year for your expectation for the percentage of customers on long-term contracts next year? And then, second question you have talked about how the European opportunity is mainly a data center NetCentric customer opportunity, which I think implies a higher ARPU customer. And I was wondering if you could tell us of the sales force additions that you are adding. What number or what, how -- what's the growth like in the European sales force verses the U.S. sales force. Thanks.

Okay, hey, great, Tom, thanks for the questions. First of all, we do anticipate continued lengthening of contracts. We were actually a bit surprised in Q2 at the rate of acceleration and customers taking us up on our discount, our offer and longer-term contract. We continue to see that trend continuing. Our modeling methodology, as we prepare guidance is, that Tad and his team, basically take all of the trend lines from key drivers in our business or from the previous six months, we look at any outliers that are not particularly within the range. And then we take kind of the average of that range and project it going forward. With that, we would expect that the percentage of customers taking advantage of the discount and therefore on longer term contracts in month to month, will probably by the end of next year hit close to 90% of our base as opposed to the 80% today. So, we'll expect to see that continue.

Now, to your European question, while we would love to have a corporate business in Europe the building footprint just doesn't fit our demographic requirements. There are few buildings in Europe of the scale of buildings here in the U.S. and secondly those very larger buildings tend to be more single-tenanted than multi-tenanted. So, we do have a few buildings in Europe. We are actually in the process of adding a few buildings in the Canary Wharf region in London to our network. So, there are some corporate opportunities, but it is true that most of the opportunity in Europe is selling to either content companies or many cases PTTs within Europe and we continue to do quite well in that market space. That does generally mean higher ARPUs meaning that they are generally paying more than a corporate customer per connection. However, some of these countries have relatively small install broadband bases, there is an opportunity for significant growth in their broadband penetration rates, whereas in Western Europe and the U.S., that's much more of a mature market. So, I think we'll see some smaller ARPUs there and we'll see some acceleration in growth that comes from someone who buys a smaller connection, but in a market that has only 10% or 15% broadband penetration, maybe growing to the 60% to 70% that we are seeing in the Western World. And then the final impact on non-U.S. ARPU is quite honestly the FX impact and as Tad said, we are obviously expecting to see some pickup this quarter at an accelerative rate than we have seen in the past. But at the end of the day, we are not currency traders, we are running Internet Company and we are trying to the best of our ability, not to create cross-border arbitrage opportunities with our products.

Great, thank you very much.

Thanks, Tom.

The next is Jurgan Usman with Wachovia Securities.

All right, thank you very much, at the end of the line here, couple of questions. First of all, on the revenue churn, can you maybe give that out?

Sure. Essentially, on-net has held steady at 2% for quite some period of time, so it was 2% for the quarter, basically identical to what it is has been -- was for the first half and for the prior two quarters of 2007. We did see that off-net, ARPU or churn rather did decline, and I will give you the exact numbers here, it was 6% for the second quarter and 3.5% for the third quarter, so the churn for off-net did decline and for on--net essentially dead on at 2%.

All right. And the next one is can you give me an update on your largest customer and how did they, I guess in quarter regarding bandwidth commitment?

Sure. Our largest customer represents about 2.7% of revenue. That particular customer did increase their commitments in the quarter to Cogent primarily by expanding their footprint into Europe. So, they have taken multiple connections throughout Europe as well as North America. We did see our quite honestly their North American traffic was slow while still positive not growing at the same rate, kind of entertaining maybe, a bit of a maturity in their business model. But we have a very diverse customer base with a number of different business models. And overall we are seeing good pickup across all of our customers, represented by the fact that our monthly traffic growth in October was about 5% sequentially just month over month.

All right. And then, the next one is on your productivity, seems to be fluctuating a lot more volatile, I guess, in the past couple of quarters or so, which I guess is kind of (inaudible) given that you have a lager number of sales people. Has there been an increase in terms of sales force or is it just because of they are still building their funnels or what's going on there?

Big part of it is funnel building here again. Sales force churn has been relatively constant. What has happened though is the average tenure of the sales force continues to decrease. As we indicated, we had record hiring in the most recent quarter in Q3. We added more reps than we had ever added in a given quarter. So, that kind of brings down the average tenure of the sales rep with less tenured people, they tend to have less mature funnels and therefore, are more subject to these volatility. And again, we feel very comfortable that the long-term trend remains intact and you are just going to get volatility quarter-over-quarter based on lumpiness of specific customer contract and timing. But the long-term trend are for these reps would remain productive at about five units per ramp per month.

Okay. And then, last one I promise. Are you --

That's okay.

On that revenue growth guidance for 2008 kind of implies about 7% sequential growth, obviously lower than your current guidance of 7% to 9%, 7% to 8% in the fourth quarter. Now, you did say in the past that, I guess longer term you expect on-net to grow 9% to 10% sequentially. I mean is this something that you probably see somewhere longer in the future, maybe 2009 or so. Or I was just wondering, there seems to be a little bit of disconnect here with what you said and the guidance, what the guidance imply. Thanks.

Sure, Jurgan. As I said, I think the long-term drivers of our growth remain intact. We do believe that with adequate number of sales people and that's why we would increase the number of sales people that we plan to add, we can continue to achieve that type of growth. As we look at, coming up with projections for the next quarter and the next year, I am going to again revert to the methodology we used. And what we do is take actual results form the previous six months and straight-line down, because we experience lower growth sequentially in Q2 and Q3. In our long-term historical run rate we're projecting that out in to Q4 and into next year. But, we do believe that the underlying drivers of growth are as strong as ever and we do believe that we will be able to return to that 7% to 9% or ultimately 9% to 10% growth rate in the long-term, and maybe even in 2008. We just try to be as conservative as responsible. And hopefully, investors get comfort in the fact that we gave our guidance in November of 2006, and are literally at the midpoints of those guidance numbers a year later in an industry that is growing rapidly. So, I think our methodology of forecasting has hopefully earned us a little bit of creditability.

All right, finally last on CapEx, 2008 seems to be, I guess, flat versus 2007 about $30 million give or take?

That's correct, Jurgan. This year CapEx went up just marginally because of the data centers that we didn't anticipate when we put that guidance out and having to bring them up to standards. We do expect to continue to expand the network and report on a series of new markets and obviously new buildings. We intend to exceed the goals that we have. We absolutely anticipate CapEx to be flat and in that kind of $30 million range or slightly under, next year as well as this year.

All right. Thank you very much.

Thanks.

We'll go next to Scott Goldman with Bear Stearns.

Hi, good morning guys.

Hi, Scott.

Just wanted to go back on the gross margins; I guess, back in the second quarter you added the data centers late in the quarter. You talked about being able to grow gross margins 100 basis points in the third, obviously that did not materialize and there had been a decline. So, just wondering what may have changed from those original expectations that resulted in that decline? Is that also what brought incremental margins appear to be below your historic run rate in this quarter as well? It's the same factor that impacted gross margins at work there. Secondly, you talked about seasonality of traffic, just wondering if you can make any comment as far as what you may have seen kind of late in the quarter, September or what you saw already in October as far as traffic growth.

Sure. First of all, with regard to the new data centers, they did have a direct cost on network operations, as I reflected earlier, of about $600,000. As we took over centers that were baked in, were virtually baked in. Just a few costs connects them down with virtually no revenue and then had to both bring them up to standards and to begin to populate them.

We did indicate that our gross margins for Q4 should expand 200 basis points. As I have shared with investors over the years, there is typically some volatility in any given quarter due to lumpiness of payments, various service contracts and so for. But, that 100 basis point margin expansion on a sequential basis, both gross and EBITDA trend remains intact. If you look at our margins for next year, we are actually forecasting better than a 100 basis points sequentially of gross margin and EBITDA margin expansion for the year. So we are looking at full year 2008 gross margins of about 60% up from the roughly 54.5% this year, and EBITDA margins continuing to expand from the roughly 27.5% that we are going to do this year up to the roughly 34% that we are forecasting for next year. We do see continuing operating leverage as the on-net portion of our business continues to increase, as we increase the utilization of our existing footprint more effectively. So, while we did see some gross margin degradation in the quarter as a result of picking up these empty data centers, we see that as a one time anomaly and not as any kind of long term trend. I fell very comfortable about the guidance we have given. In terms of seasonality of traffic, as I mentioned earlier, we saw significant pick up in the month of October where traffic on a month over month basis picked up at about 5%. That will compound to better that a 75% annual rate. We also saw significant pickup in September over August. Basically July and August were relatively flat months as we have seen in previous summers, but we think that the proliferation of media migrating to the Internet will be the continuing driver of growth, maybe a little bit more pessimistic than others in the industry, we don't see an acceleration, but we do see kind of a consistent rate of traffic growth.

If I could just add one thing to the gross margin comment, in our cost of goods sold, we did have some costs associated with our expansion and our expansion frankly is kind of ahead of our schedule. So, we did incur some cost of goods sold expenses kind of ahead of plan which contributed to the margin decline Q2 to Q3.

Okay, great. One follow-up and one additional question is, I guess as long as I stay under Jurgan's number of questions, I should be okay, just on that -- anyway Tad you can quantify those costs that you were just referring to and then just quickly on customer adds on-net ramping sales force, we have talked a bit about the productivity on the call already, churn relatively flat, is there anything about the sales cycle of trying to sell to higher bandwidth that would maybe have brought that on-net customer add number below maybe where I was expecting relative to past quarters?

On the quantification, I would put it in the total cost of goods sold dollar amounts in the $700,000 to $800,000 range. As Dave mentioned, the way we build our forecast is looking at the recurrent run rate so -- and anything there that is recurring as supposed to non-recurring is baked into the plan for the fourth quarter and for 2008. On the sales cycle, I would say there is nothing in particular there to point out, at least in my end, but Dave might have something to say.

Yes, I think on the sales cycle, there really is no difference in selling to someone, if they are a large customer or a small customer, they are basically going to buy what they need for service and when they are out of contract with their existing provider. Where the variation comes from is really having a bigger funnel. So, if a rep has more things that are becoming actionable in a given month, they are able to pick and choose and go after the bigger deal and we are really trying to train and incent the sales force to continue to build that funnel. The limiting factor in Cogent's growth for the foreseeable future is actually still sales people in not contacting enough of the opportunities in our footprint, but again within the footprint, you want to go after the better opportunities first and that's probably the bigger fact.

Great, thanks for taking the time guys.

We'll go next to Andrew Beale with Arete Research.

Hi, it's Andrew Beale from Arete Research. Can I just go back to the sales force churn question first of all and just ask you how many gross sales force additions did you make to get to your 23 net? Secondly, on -- in terms of the 100 buildings that you're planning to add in '08, is there any measure you could give on the revenue potential per building of those relative to your existing buildings? I mean just presumably there -- they are getting slightly smaller in terms of average size. And then thirdly, I don't know if there is any way you can think about breaking down the revenue growth in a slightly different way to talk about how much of it comes from increasing share within buildings, how much of it comes from adding new buildings and how much is just from the underline demand, which I think you probably said was 3% sequentially, but any thoughts you have on those areas, I appreciate it, thanks?

I'll take the first question there. I can kind of give you roll forward, if you will, of sales rep that we started the third quarter with 143 and then added 53 and then churn 30, [so 166] again.

In terms of the building demographics, actually the corporate buildings that we are adding today are actually slightly larger than our existing average. So, that number continues to increase for about nearly at 580,000 square feet up from roughly the 570,000. We are very judicious about adding those buildings. It's in those buildings we have increased our penetration work today across the entire footprint, we have about 6.3 customers per corporate building out of an addressable market of about 50 in each of those buildings. For those corporate customers, it is almost certainty that they are not going to buy a second connection in the same building or increase their connection. Now, that potion of our revenue stream is about 23% of revenue, so roughly 43% being corporate, 23% of that 43% is the net corporate customers.

Now, for the NetCentric customers, a little bit different demographic. Within the data centers, we have just under 15 unique customers per center across the entire footprint. That represents about a 7.5% unit penetration, but still about an 18% penetration by bandwidth consumed within those facilities. In any given month, roughly 70% of our sales come for brand new logos, someone who never bought a connection from Cogent before, roughly 30% of sales comes from existing customers. So, then is you look at split between NetCentric and corporate, that kind of 70/30 ratio applies almost equality to the two bases. So, with 30% of NetCentric incremental revenue coming from existing customers, there is a high correlation between traffic growth and revenue growth for that 30% of the 57%. So, hopefully that was enough detail to answer your question.

That's great, thank you.

We'll go next to Andrew Tuttle with Crow Point Partners.

Can you hear me?

Yes, sure.

Okay thanks. Forgive me if you have answered these questions before, but could you give the traffic growth stats sequentially and year-over-year?

Yes, year-over-year traffic grew by a little over 66% from Q3 2006 to Q3 2007.

Okay, I've also noticed that in about the last 18 months or so, the average revenue per on-net connection is probably down 18% or 19%. Is that decline going to abate do you think going forward or where does that go?

Yes, I think Tad touched on that, we see that ARPU continues to decline at least for the next year, eventually, at some point, it will actually start back up. But the decline is driven by kind of two key factors. The first is on the NetCentric portion of our business which is 57% of revenues. That portion of our business is selling to smaller and smaller customers within those data centers. So their average buy is less not that they are paying less per megabit, but they are just buying less megabit. On the corporate side, we see ARPUs virtually flat. And then, the second factor is this increasing contract length. So in the quarter we saw the entire base of average contract duration extend about 5%. So we saw it going up probably 10.7 months to 11.3 months that has the impact of reducing our price per megabit because we do offer discounts.

So our effective price per megabit today is just around $9 a megabit. That is down from probably, maybe $9.40 or $9.50 a year ago on a weighted average basis of customers' contract length. So ultimately, what will happen is, we'll get to a saturation point of people that take longer term contracts, I think 90% probably close to that, where we think we'll be by the end of next year. And then secondly, we will see that even the smaller NetCentric customers' buys are as big as our ARPU, and at that point it will see ARPU for on-net business begin to actually increase.

Okay. And if you get to 90% of your customers on long term contract by the end of next year like you're targeting, I imagine that churns going to have to tick up a little bit at least after that point, if you kind of map it out on your contracted customer basis.

Well, not really, because remember, churn is measured when a customers goes away. As we had mentioned on the previous call, approximately 94% of our contracts have auto-renewed provisions on them. So even for those customers that take a one year contract, at the end of that one year, they automatically renew for one year unless they proactively cancel. Now, we do not count that renewal in our new sales numbers. But, I think as the percentage of business that goes under longer-term contract increases, that should actually have a beneficial impact on churn, and we've seen no increase in churn, if anything, we are seeing a slight decrease in the rate of which revenue is churned.

All right, great, thanks guys.

Okay.

And it appears that we have no further questions at this time. I'd like to turn the call back over to our speakers for an additional or and closing remarks.

Well, again, thank you all very much for your support for listening to today's call. The management team will be available, if you have any follow-up questions. I personally, again, would like to thank the entire Cogent team for the efforts of growing our revenues, of installing those revenues in timely manner and generating the incremental cash flow that's allowing us to buy back shares in the market. So with that, I would like to thank everyone, and we'll talk next quarter. Take care, bye- bye.

And once again that has concluded today's call. We do appreciate your participation, you may disconnect at this time.