The main other cost that must be borne is the time and money involved in fully determining system requirements and setting up a workable software/hardware solution. This should essentially be a fixed cost regardless of the number of sites. This will depend, however, on how much help is given to additional sites over and above provision of manuals. It would be wise to allow an additional overhead for each extra site. The consultant estimates that approximately ten to twelve weeks of professional time and the equivalent time for a skilled data entry person would be involved for the initial installation and a further week for each additional installation. This does not include travel time. Professional costs could amount to US$ 1 200–1 500 per five-day week if outside consultants are used. Data entry salaried staff may get the equivalent of US$ 200/week.
Each site would require the services of someone completely knowledgeable about the running of the selected software package for perhaps 10 percent of normal working duties. A global site may run as high as 20 percent due to increased demand on the system, more frequent reorganization, longer backup times, etc. To this one would add the services of a permanent data entry operator for, say, one-third of normal working hours for a regional site and two-thirds of normal working hours for a global site. Costs quoted below assume developed country rates. Developing country rates would obviously be much lower.
Further consultation - computer professional only | US$ 20 000 | |
| Data entry staff | $ 2 500 | |
| Software Cost | $ 5 000 | |
| Hardware Cost | ||
| - Startup only (to last for 2–3 years) | $15 000 | |
| - Single user | ||
| $42 500 | ||
This figure represents a maximum startup price. A fully blown system that would immediately be able to provide for all estimated source documents and give magnetic tape backup would add around US$ 10 000. It is recommended that a small scale version of the final system initially be adopted to provide experience without outlaying unnecessary money. Prices of hardware are continuing to drop so the longer the payment can be delayed the better.
| One-fifth professional computer time | US$ 6 000 | p.a. |
| Two-thirds data entry time | $10 000 | p.a. |
| Hardware maintenance contract at 10 percent purchase price | $ 1 500 | p.a. |
Further consultation - professional computer staff, applicable once only | $20 000 | |
| Data entry staff - once only | $ 2 500 | |
Further professional consultation - each additional site | $ 1 500 | |
Data entry staff - each additional site | $ 200 | |
| Software Cost | $ 2 000 | |
| - | if development aids for a very intelligent system were purchased for local development, they would need to be only purchased once. Their cost could be defrayed over all sites. Additional licenses for each site may still be required. | |
| Hardware Cost | ||
| - | Startup only | $ 8 000 |
| - | Single user | |
Comments to global site are applicable here. Thus for three regional sites, initial startup cost could be as high as:
$22 500 + (1700 x 2) + (2000 x 3) + (8000 x 3) = $55 900
This may be reduced by as much as $5 000 per site by selection of software with low memory requirements.
| One-tenth professional computer time | $ | 3 000 p.a. |
| One-third data entry time | $ | 5 000 p.a. |
| Hardware maintenance contract at 10 percent purchase price | $ | 800 p.a. |
On the figures shown above the extra initial cost of going to three regional sites rather than one global site is quite considerable. However, any savings in software and hardware on the prices given will be multiplied by three for regional sites and so a saving of even US$ 2 000 per site considerably reduces the margin. Ongoing costs, however, will almost certainly be dearer for three regional sites than for one global site, unless all duties (and salaries) can be incorporated within the normal duties (and salaries) of existing staff.
Over and above economic concerns, many other factors need to be considered before the final choice between regional and global sites can be made. These have been detailed earlier. Perhaps the one of user morale and feeling of involvement is one of the more significant factors in leaning towards a regional solution.
