Sensitivity Analysis Of Hosted vs. Premises IP-Telephony TCO/ROI
Ken Dolsky and Warren Williams
Does it pay to go with a hosted service? In many cases, the answer will be yes.
I
Ken Dolsky (kdolsky@ accessintel.com), a director of professional services at InfoTech, has 30 years of experience across a wide range of disciplines in the telecom industry. Warren Williams (wwilliams@ accessintel.com) is VP and senior program director at InfoTech. He also has more than 30 years’ experience in the industry. This is the second of a twopart article.
n last month’s issue, we presented some of the main reasons why enterprises might consider a hosted IP-telephony service instead of a customer premises equipment (CPE)-based solution (see BCR, November 2005, pp. 51–55). In the second part of our analysis on the hosted-vs.-CPE decision, we will present findings derived from a TCO/ROI tool that we have developed based on our research into hosted IP-telephony. We believe that the considerations described in last month’s article, together with the cost issues detailed below, will drive significant increases in enterprise adoption of hosted IP-telephony services over the next several years. The TCO/ROI tool compares TABLE 1—Key Assumptions the cash flows of premises-based Users (single location) IP-telephony systems with that of Voice mailboxes hosted IP-telephony services. Percent of users with soft terminals 1 Using this tool, it is possible to not Percent of users with digital phones only compare the financial effects of different premises systems and Percent of users with single line IP phones hosted services offers, but also Percent of users with multi-line IP phones determine the areas of greatest Cost of capital impact on ROI and cash flow. IP Telephony service charge per month per user 2 This article demonstrates a set Voice mail service charge per mailbox per month of results based on changes in costs Soft phone terminal charge per month which we have determined are the Single line IP phone charge per month drivers of significant differences in Multi-line IP phone charge per month the financial results of the two Charge per port for system equipment approaches.
Equipment/Service and Installation ■ System purchase costs for premises system ■ Hosted service monthly charges for service ■ Terminals (bought or included in service) ■ Software application upgrades ■ Training Operations and Administration ■ System and terminal maintenance (for all purchased equipment) ■ Remote monitoring for purchased system ■ MAC costs ■ Power, HVAC, insurance, floor space Access and Toll Costs ■ Access charges (T1s) ■ Toll traffic charges Human Resources (Initial planning and ongoing management) ■ Business/financial planning and project management
50 55 20% 25% 25% 50% 12% $30 $8 $1.50 $3 $5.50
Assumptions And Results The key assumptions that we used in this analysis are presented in Table 1. The financial results produced by the ROI tool when using these values in a 5-year analysis are shown in Table 2. The costs modeled in the tool are distributed in four categories as shown below:
(purchase price) (does not include terminals) $160 Charge per mailbox for voice mail system (purchase price) $36 Charge for IP Telephony software upgrade $1,000 Charge for voice mail/messaging software upgrade $250 Minutes of domestic toll use per month 3 33,000 Cost per minute of domestic toll use with hosted service 3 cents Cost per minute of domestic toll use w/o hosted service 3.8 cents
1 The tool allows users to have both hard phones and softphones. 2 Telephony service only—does not include toll or LD costs unless specifically included in bundle. 3 Assumes .5 hour per user per business day (local and LD)
Use BCR’s Acronym Directory at www.bcr.com/bcrmag
40 BUSINESS COMMUNICATIONS REVIEW / DEC 2005
TABLE 2 Results
Net Present Value—Premises System Costs $225,380 Net Present Value—Hosted IP Telephony Service Costs $137,002 Total (discounted) Savings From Hosted Service $88,378 Over 5 Year Planning Period ROI of Hosted IP Telephony Service 1 65%
1 This ROI was developed by dividing the total discounted savings by the NPV of the Hosted Service costs. Thus, this is the ratio of savings to “invested costs.” Another way to look at this is to say this is the percent of cash flow improvement for a Hosted Service.
■ Application software upgrades ■ Purchase vs. lease of terminals ■ Human resources involved in planning and managing systems ■ Decrease in price of hosted IPtelephony and voice messaging services ■ Decrease in price of premises IPtelephony and voice messaging systems
We assumed hosted service customers would get a better rate for toll usage
■ System administration ■ System technicians ■ Security management ■ Contingency planning management When reviewing the results presented in this article, it is important to understand that the ROI of the hosted solution is always a ratio of the savings from the hosted service vs. a premises system. The breakout of the four cost categories for the purchased system costs over 5 years is shown in Figure 1. The breakout of the same categories for the hosted service costs over 5 years is shown in Figure 2. Given that the major impacts on ROI are associated with these categories, six costs were varied in order to test the sensitivity of the ROI analysis: ■ Toll bundling
Sensitivity To Toll Bundling Toll bundling is defined as an offer by the hosted service provider to include the charges for all local toll calls in the monthly charge for the service. The ROI tool allows users to model bundling of any services (local, long distance, international and Internet) in any proportions. The inclusion of local service in this analysis was selected as being a common example of bundling that is offered by service providers. It is assumed that half the total domestic toll usage is local. The difference in price per minute between the hosted and non-hosted service is based on the assumption that the hosted provider offers its hosted IP-telephony customers a slightly better price as part of its hosted package. By including these advantages in the hosted service offer, its ROI advantage over a comparable premises system increased by 117 percent (Figure 3, p. 42). FIGURE 1 Purchased IP Telephony System Cost Elements (5-Year NPV = $225,380) Sensitivity To Software Application Upgrades Equipment The base case assumes that there are and Installation four application updates, two for IP15.8% telephony (in Years 3 and 5) and two Operation and for voice messaging (in Years 2 and 4). Human Administraton Resources In addition to the cost of the software 7.7% 41.4% upgrades themselves, the tool assumes that managers in a CPE system environment will incur corresponding 35.1% resource costs in evaluating (i.e., perAccess and forming a cost/benefit analysis), planToll Costs ning and managing these upgrades. In the hosted environment, it is assumed that there is no charge for the FIGURE 2 Hosted IP Telephony System Cost Elements application upgrades and there is virtu(5 Year NPV = $137,002) ally no need to evaluate, plan and manHuman age the process. By including these Resources advantages in the hosted service offer, 6.6% its ROI advantage over a comparable Access and premises system increased by 23 perToll Costs cent (Figure 4, p. 42). 17.1%
Operation and Administration 0.1% 76.3% Service and Installation
Sensitivity Of Terminal Leasing The base case assumes that all terminals are included in the hosted service and incur monthly charges as shown in Table 1 (which includes maintenance). The model shows that by including these terminals in the hosted service
BUSINESS COMMUNICATIONS REVIEW / DEC 2005
41
50%
10 % 80%
20 %
30 %
40 %
70 %
60 %
0%
10% 30% 65% 65% 30%
20%
30%
40%
50%
60%
70%
0%
FIGURE 3 ROI Of Hosted Service Increased By 117% With Toll Bundling
FIGURE 6 ROI Of Hosted Service Decreased By 54% With 50% Reduction In Customer Resources Needed For Premises System
R w/ OI o of H To o ll st B e RO und d S lin er w/ I g vic To of H e ll os Bu t nd ed lin S g er v ic e
42 BUSINESS COMMUNICATIONS REVIEW / DEC 2005
R C OI Pr ust of em om Ho is er ste e d RO s S Res Se ol ou r Re I ut r vic Fo du of i c r P ct Ho on es e w fo /F re ion ste r ul m o d l is f S es Cu er So sto vic lu m es tio er w n R /5 es 0 ou % rc es
10%
20%
30%
40%
50%
60%
70%
0%
100%
20%
40%
60%
80%
53%
FIGURE 4 ROI Of Hosted Service Increased By 23% With Software Application Upgrades
0
65% 83%
FIGURE 7 ROI Of Hosted Service Increased By 28% With 15% Reduction In Charges For Hosted Telephony and Voice Messaging
R w/ OI o Up o S f H gr of os ad tw te es ar d e Se Ap r v RO pl ic w/ I ic e Up So of at gr ftw Ho io n ad a st ed es re Ap Se pl r v ic ic at e io n
65%
R Pr OI ic of an e d s Ho Vo fo st ic r H ed e o S M st e es ed r v sa T ice RO gi ele w Re I ng p /B ho a Te du of le ct Ho ny se ph io st on n ed y in S an Pr er d ice vic Vo s e ic fo w e r /1 M Ho 5 es s % sa ted gi ng
10% 0%
20%
30%
40%
50%
60%
70%
10% 70% 65% 63%
20%
30%
40%
50%
60%
80%
0%
55%
FIGURE 5 ROI Of Hosted Service Increased By 18% With Terminal Leasing
FIGURE 8 ROI Of Hosted Service Decreased By 3% With 15% Reduction In Charges For Premises Telephony And Voice Messaging
R P OI an rice of d s Ho Vo fo st ic r P ed e re S M m e es is r v RO sa es ice Re I gi Te w ng le /B Te du of ph as le ct Ho ph io st on e on n ed y in S y an Pr er d ice vic Vo s e ic fo w e r /1 M Pr 5 es em % sa is gi es ng
R w/ OI oT of H er o m st i e RO nal d S Le e w/ I o of H as r vic in e Te o g rm st in ed al S Le er as vic in e g
65%
Affect On ROI
80%
Sensitivity To Resources The base case assumes that 60% the customer must commit considerably greater 40% resources in order to plan, implement and manage a 20% premises-based system, compared with a hosted system. The baseline assumptions for 0% the premises-based system include 0.2 man-years for planning (includes all financial and technical planning, business and project management) and 0.21 man-years of ongoing system/network planning and management each year. These figures represent a small potential implementation (about 50 users), and scale up as the system size increases. In order to determine the sensitivity of the ROI to these resource requirements, both the initial and ongoing planning resources for the premisesbased system were decreased by 50 percent (basically assuming that the CPE implementation could be made more resource-efficient). The result was that the ROI of the hosted service was reduced by 54 percent (Figure 6). While the 0.2 man-years involved in implementing a small system may seem negligible (and the above numbers may be conservative), there are two points to keep in mind: 1. In a small-system environment, some of the people needed to plan and manage the system have other functions, and reducing the amount of time needed to plan and manage the telecom system frees them up to perform other duties. While the customer will not see a direct difference in the bottom line associated with these costs (they are not going to lay off a fraction of a person), there is a hidden cost associated with diverting these resources to manage telecom, and this is largely removed in a hosted environment. 2. The numbers used in this analysis will scale up when larger systems are being implemented, and can start to affect the actual cost of resources (e.g., freeing up a full person can be used to offset the need for a new hire). Thus the analysis here should be viewed as directional and scalable. Sensitivity To Changes In Base Prices The model used in this analysis assumed (as shown in Table 1) that the charges for hosted IPtelephony and hosted voice messaging were $30 per user per month and $8 per user per month,
5 C 0% Pr ust Re em om d is er uct es R io Sy eso n In st u em rc es
instead of purchasing them separately (along with separate maintenance charges), the ROI advantage over a comparable premises system increased by 18 percent (Figure 5).
FIGURE 9 Effect Of Each Cost Factor On Hosted Service ROI
120% 100% 117%
S Up oftw gr ar ad e es Ap pl ic
28%
18%
Ho st 15 ed % Se Re r v du ic cti e o Ch n ar in ge
respectively. (These assumptions were based on InfoTech’s research with existing hosted IP-telephony services.) Reducing these charges by 15 percent resulted in a 28 percent improvement of ROI (Figure 7). By comparison, the model used in this analysis assumed (as shown in Table 1) that the purchased prices for the premises IP-telephony and voice messaging systems were $160 per user and $36 per user, respectively (based on InfoTech’s research with PBX vendors). Reducing these charges by 15 percent resulted in only a 3 percent decrease in the ROI of the hosted service (Figure 8). The reason this decrease was so small is the fact that, as shown in Figure 1, equipment and installation accounted for only 16 percent of the total CPE costs over 5 years. Other costs such as toll services, and operations and administration (maintenance charges) are significant factors in the total costs incurred by customers with premises-based systems, tending to “swamp” out the effect of a cut in equipment prices. On the other hand, the 15 percent reduction in hosted services charges is much more effective than a 15 percent reduction in premises system charges in improving the ROI of the solution, because, as shown in Figure 2, the hosted service charges account for more than 76 percent of the total costs over the 5-year period. Conclusion Figure 9 summarizes the impact of each cost factor on the overall ROI results. While toll bundling clearly has the largest effect, the other areas analyzed contribute significantly to the overall financial results. In the absence of a toll bundling offer, these other areas would represent some of the key decision factors. Even with toll bundling, their cumulative impact would be considerable
BUSINESS COMMUNICATIONS REVIEW / DEC 2005
T Le erm as in in al g
Bu nd Tol lin l g
54%
23%
15 % fo R r P ed re uc m tio is n es in Sy Ch st ar em ge
Price breaks on a service translate to greater savings than do price breaks on equipment
fo r at io n
3%
43