Good Practice Guide for Developers of Cultural Heritage Web Services
Good Practice Guide for Developers of Cultural Heritage Web ServicesAcknowledgements
This section was first published as part of the NOF-digi Technical Advisory Service Programme Manual in 2003.
Digitisation projects will inevitably be dealing with other companies, whether as a result of outsourcing particular tasks, such as web design, digitisation or programming, or in order to procure services supplies and equipment, such as systems, scanners or computer software. Working with limited budgets, and within the public-sector domain, it is essential that projects get value for money when involved in obtaining such services or equipment. This section of the guidebook offers advice on the process of procuring such equipment or services. While each project will have different purchasing plans, a generic set of guidelines offers some basic pointers in what can be a complex field.
Before starting any procurement process, it is worth establishing whether it is essential to purchase. There may be the possibility that your organisation or a related institution already has the piece of equipment required, and organising the temporary loan of that equipment would be enough to suit your purposes. Alternatively, it may be possible to hire the equipment in question. Hiring for a three year period, especially in the case of IT equipment, can often be cheaper than outright purchase, saving as it does on initial outlay and long-term maintenance costs. Equally, hiring temporary space via Web hosting services can be cheaper than purchasing dedicated system to run Web sites.
Solutions such as these are more suitable for short fixed terms, where project managers can be sure of the time for which the equipment or service will be required. The advantages of non-purchase can quickly evaporate should it be discovered that the goods in question are required for a considerably longer period than originally envisaged.
When participating in a large digitisation programme with many partner projects it makes sense to explore the possibility of joint purchase. Spreading expenses among several institutions or projects can greatly reduce the cost of equipment, or services. It is especially apposite for expensive technology that is essential for a small number of tasks, as opposed to cheaper equipment that will be used on a more routine basis.
While initial costs might be lowered by joint purchase, it can result in greater pressure on project managers to organise the use of the technology. If careful attention is not paid to the allocation of the equipment, there is greater possibility for workflow bottlenecks. There is also more likely to be a greater number of bureaucratic obstacles during purchase, and, in the long-term, possible disputes over final ownership of the equipment.
Competition is the essential factor in the procurement process. There are very few products available where no choice is available. Purchase of any goods or services will involve the comparison of possible suppliers followed by the considered selection of the option that offers the best value for money. Value for money, however, should not be taken to indicate the cheapest purchase price. Customers need to consider a range of criteria and choose a product that will work for the specific aims of their project. If these aims demand an expensive outlay then projects must be prepared to accept this. Ignoring these aims for the sake of finance will only cause problems later on, and may eventually lead to greater expense.
The time spent on a procurement process should be related to the expected outlay. Extended deliberation is required for the purchase of goods or services that are expensive and of essential strategic importance to the project. Project managers need spend less time on the process as expense and importance decrease. Indeed, in the case of cheaper goods, many institutions forgo procurement, as the resultant savings would not be worth the staff time spent on the process.
For some of the more standard IT goods (PCs, printers etc.), projects working within local authorities will already have specific suppliers nominated. The GCAT framework [1] supports those working within local authorities by providing a pre-tendered range of goods. Universities normally have their own purchasing offices or consortia that also provide a list of pre-arranged suppliers. When this is the case, the procurement process can become much simplified, individual purchasers dealing with one or two particular suppliers. However, when there is no such arrangement, it is essential that projects consider various possible suppliers to achieve value for money. The NOF-digi programme stipulated that for purchases estimated to be between £500 and £4999, two quotes were sought, while three quotes were sought for any amount of £5000 and above.
Most institutions specify that for a projected outlay over a certain cost, the procurement must involve formal tendering procedures. The essential element of any call for tenders is the composition of a detailed specification. The purpose of the specification is to describe clearly and accurately a project's essential requirements in terms of equipment or services. A specification will form the basis of all offers made by potential suppliers, and act as the foundation for any later contract. If a specification is not well prepared there is greater likelihood of vendors failing to deliver the aims expected by a project, or not even offering their services. Creating a clear specification also permits suppliers to formulate quicker, more accurate and less costly responses.
Specifications should be based around the following key sections: procurement aims/background, statement of requirements and the selection process. Depending on the nature and cost of the goods or services under consideration, project managers may wish to include other particular sections in their specifications. This might include a table of contents, security issues, training requirements, delivery terms, a glossary etc. However, such sections are often subsumed within the statement of requirements section.
The first section of a formal tender should include a declaration of the project outcomes that the customer requires the vendor, via its equipment or services, to supply. Each outcome should be succinct and not more than a short paragraph, or even a single sentence. The specification may also articulate how the tasks required of the vendor will slot into the high-level goals of the project. So a project specification might begin by declaring that the project is looking to have a Web site developed as part of its remit to deliver learning packages associated with its collection of digitised images. Another project might indicate that, as part of a regional consortium, they are looking for a digitisation bureau to digitise its collection of local nineteenth-century newspapers.
The project aims should be placed in the context of a summary of the project's background, introducing readers to the purchasing organisation. As well as explaining more about the eventual goals of the project, this background information may provide a summary of what decisions led to the purchasers putting out this particular call for tenders.
This is the key section of the specification document. It identifies the performance requirements (for equipment and services) or outputs (for services) or functionality (for software) that the project requires, and gives details of how these requirements, outputs or functions will be measured once a contract has been signed. While it is likely that these elements will be refined during the course of the tendering process, specification writers should ensure the details are as precise as possible so vendors are clear about the parameters of the project's needs. Projects need to make a close examination of their workflow, pinpointing each section where work done by the suppliers will have an effect. Each of these sections should then be converted into a requirement or series of requirements. It is important not to include requirements that could be broken down into others. If a project fails to clarify its precise needs, then suppliers will be slower in offering effective responses. Any programme-level or funder-stipulated technical standards will have to be included into the document, such as, in the case of the NOF-digitise programme the nof-digitise Technical Standards and Guideline [2].
For outsourcing the digitisation of a set of images, some of the requirements will again be shaped to fit the technical standards (for example, the required format/s) whilst others will be dependent on the particular demands of the project (for example, delivery of a certain number of images per month). So that the performance of the digitisation unit can be examined, quality-control tests would also be included in the statement of requirements.
It is highly likely that some requirements will be of the utmost priority, whilst other will not be quite so important. Therefore, it is often worthwhile ranking each requirement as being mandatory, highly desirable or desirable. Being able to rank requirements will also allow for greater flexibility when it comes to negotiating and refining the specification with suppliers.
The selection process should offer the potential supplier details on how to respond to the call for tenders. It should outline what structure any replies will take, and the level of detail required. It is prudent for purchasers to ask for additional information such as general company information and reference to similar work that potential suppliers have done in other projects. This could include the contact details of previous customers, who might be able to provide their own assessment of the company in question.
Administrative details such as the project contact point, the relevant address, and the number of copies required should all be included. A timetable indicating deadlines for initial responses, negotiations and acceptance of the final tender should also be present.
Specifications often give a brief indication of how the tenders will be assessed, indicating any particular criteria that have been established. These criteria should be weighted by relative importance. Including such information allows the vendors to produce tenders more focused on the project's requirements.
Once a specification is complete it can be placed within a formal document, ready to be disseminated as a call for tenders.
If the contract is likely to be over a certain amount, it is necessary to follow European Union Directives on Tendering. At the moment the threshold is £153,376 for services or supplies, this level having been set in January 2004. All projects that are intending to make purchases over this amount must advertise their call for tenders for services or supplies in the Official Journal of the European Communities. Additionally, the EU directives stipulate that tenders cannot be received until a set number of days after the appearance of the advertisement in the Official Journal. This delay is in the region of five to seven weeks. More information on the directives is available from the EU Web site [3].
Should the projected outlay for procurement be below the levels prescribed by a project's host institution, then there is no formal need to construct a specification. Nevertheless, projects should still develop their own internal list of requirements for a particular service or piece of equipment. Creating such a document can serve many of the useful functions that a specification document serves in projects following formal tendering procedures.
Having completed a specification or list or requirements, purchasers must seek out potential suppliers in order to obtain a quote or notify them of the existence of a call for tenders. This can be tricky, especially as there are numerous competitors for IT goods and services. In some cases, previously-used suppliers might be able to offer their services or equipment. In other cases, however, it will be necessary to locate new suppliers. Initial research via the Internet, the relevant trade magazines etc. is a time-consuming process, but will give one a broad idea of the purchasing environment. Trade shows at professional conferences and commercial exhibitions, such as the Online Information show [4], are also helpful, allowing purchasers to compare quickly a large number of potential vendors. This should be accompanied by recommendations from colleagues, consortium partners, specialised review sources etc.
Once suppliers have been notified about the existence of a call for tenders, placing the contents of the call on the Internet can facilitate responses.
When tenders and quotes do arrive from potential suppliers, it is important to scrutinise the documents with the same care that went into developing the list of requirements. In assessing suppliers' replies, the actual financial cost quoted should only be one variable under consideration. The quality of the service or equipment supplied, possible maintenance or backup costs and, most importantly, the supplier's ability to perform the task according to the project's exact specifications, are all vital factors that need to be considered. Purchasers might also contemplate having pre-trials to test the particular equipment or service under question. This might be particularly useful when assessing candidate digitisation bureaux, where projects could offer a small section of their collection for test digitisation.
Purchasers should also make sure that the price is stable, or, if there are to be price increases over a period of time, the mechanism for determining such increases is transparent. Purchasers should also be aware of possible additional costs such as delivery and installation. There might also be hidden sundry costs - the cost and availability of spare parts and consumables required for the equipment to function, accompanying handbooks and instruction manuals, and any possible need for staff training.
One should also consider the broader history of the suppliers in question. Are they economically healthy (it is difficult to get maintenance support from a bankrupt company)? Do they have sufficient quality controls? Do they have a history of sustained after-sales involvement? As always, it is worth asking colleagues and other professionals in the field, as well as the suppliers themselves, about all these issues.
Once purchasers and suppliers have exchanged specifications and tenders, it is important to engage in more detailed negotiation/discussion. This may follow a shortlisting process, where the most unsuitable quotes and tenders have been discarded. Suppliers will expect negotiation; their first quotes may reflect initial reactions, and they will be interested in exploring the details of each specification. An active dialogue may suggest alternative approaches and new ideas of how to fulfil a project's aims. It is worthwhile negotiating with each potential supplier rather than restricting oneself to just one option, although good practice in certain organisations, particularly within local authorities, calls for purchasers to concentrate on negotiating with a single, preferred supplier. Sometimes, negotiation will lead to suppliers making a better offer. This can be a straightforward reduction in cost, but could also mean slightly different equipment or (for the purchaser) more convenient payment terms. But vendors may also offer additional facilities for the original quoted price, such as extra equipment or training, or additional 'free' after-sales support. Beware, however, of any supplier that cannot meet your essential outputs or requirements.
As with writing specifications, negotiation can be a tricky art; it is well worth using someone with experience in the field. Within local authorities, universities or other organisations, there may be someone within the purchasing office designated for such tasks. Alternatively, attendance at purchasing workshops [5], consultation of various briefing papers and case studies will help improve the novice's negotiation skills.
If permitted by the project's host institution, it is important that suppliers are aware they are involved in a competitive process (although it is not considered good practice to reveal the identity of their competitors nor their quoted prices). Suppliers will be more inclined to offer attractive deals if they think they are under-cutting their rivals. This process should be continued even when only one supplier remains, making sure that the chosen supplier is unaware they are the final choice.
During an informal tendering process, it is important to draw up a supply agreement once a supplier and purchaser have agreed on a deal. For more routine equipment, however, this may simply be a standard contract already developed by the suppliers. For formal tendering, the updated specification should be used as the basis for a contract stipulating the expected outcomes or services that the vendor will provide. Again, this should be done with care. Signing a contract makes the relationship between purchaser and vendor legally binding. Should there be any disputes at a later date, it is to this contract that both parties, and any legal intermediaries, will return. Ambiguities within or omissions from the contract could catch up with projects that did not pay sufficient attention to the details.
Signing a contract can often take longer than expected. Where purchasers are unsure about the possible ramifications of certain sections of the contract, experts from organisation's legal departments will need to be drafted in to clarify the situation. Even where the wording seems clear, it may be better to consult legal experts to ensure contracts are watertight. Suggestions from legal advisors may necessitate further negotiation between supplier and purchaser. Both parties must also make sure that any updates from the original specification are agreed upon and incorporated into the final document.
The core of the contract will be the updated specification. Other issues previously discussed between the parties will also be incorporated into the document, including ubiquitous issues such as price, method of payment, timetable for delivery, and specific issues relating to aspects of the particular purchase in question.
Once the relevant parties have signed the contract, the suppliers can commence delivery of their product or services.
1. Procurement aims
2. Statement of requirements - a list of detailed requirements, indicating each section of project where suppliers' work will have an effect. This will include:
3. Selection process
Links to two calls for tenders are listed here, providing some examples of how others have set out their specifications. Each call is for a different service or product and therefore has a different level of specification. Nevertheless the key elements outlined above are present - a list of procurement aims, a statement of requirements and details of the selection process.
The AHDS portal specification is available at <http://ahds.ac.uk/ahds_portal_specification.html>. This is a very detailed specification, indicating the precise technical requirements of the Arts and Humanities Data Service in the construction of its new portal.
A call for tenders put out by the Macauly Institute for the development of a national archive of aerial photography is available at <http://www.mluri.sari.ac.uk/landcover/tenders.doc>.
Further guidance and help on procurement and tendering can be found via the Web site of the Office of Government Commerce: <http://www.ogc.gov.uk/ogc/procurement.nsf/pages/CUPGuidance.html>.
This section gives an anonymous account of the process followed by one organisation in attempting to locate a Web site designer. It shows what steps were taken in creating a specification, developing a shortlist and then making a final selection.
1 Brief
1.1 The Project Team in consultation with the webmaster and Library and Information Services staff prepared a brief.
2 Invitation to tender
2.1 The project team prepared a long list of potential designers. Team members compiled a list of web sites with design and content features relevant to our project. Each institution was then contacted and asked for details of their designer, together with their comments on the design process. A further list of designers was compiled through responses to enquiries to a number of email discussion lists. In the end a long list of nine companies was prepared (which are referred to as ONE to NINE in this document). Of these one (FOUR) was selected on the basis of past association with the Institution. The remaining eight were selected against the following criteria:
2.2 Each was sent a copy of the design brief, together with more detailed information on the content, a sample pack of primary resource materials and a critical review of sites examined by the project team. Each designer was asked to submit a costed proposal to meet the requirements set out in the brief. Companies were also invited to contact the project team by email or telephone if further information were required.
3 Shortlist
3.1 Of the nine designers contacted, eight submitted costed proposals. One company (FIVE) was not able to submit a full proposal, but made a preliminary submission.
3.2 The project team assessed the proposals. Each proposal was rated against the following criteria:
3.3 The Table below summarises the main points considered for each designer.:
| Company | Comments | Shortlist |
|---|---|---|
| ONE | Addressed brief; concerned over navigation; costs £25,000 but not including project management | Y |
| TWO | Could not guarantee to meet timetable. Cost £12,400 | N |
| THREE | Misunderstood brief : overemphasis on database integration. Cost £28,200 | N |
| FOUR | Did not respond adequately to particular requirements of brief : more concerned with a 'generic' solution. Cost £22,300 | N |
| FIVE | No submission | N |
| SIX | Cost low; doubts over management and ability to deliver project of this scale based. Costs £6000 | N |
| SEVEN | Extensive relevant experience; sound technical approach; costs £24,200 | Y |
| EIGHT | Cost £42,000 | N |
| NINE | Relevant experience; addressed brief; Cost £28,900 | Y |
3.4 The designers had not been supplied with an outline project budget. Although the project team had allocated £20,000 for design many of the proposals were in excess of this. One proposal was ruled out on grounds of cost alone. The others were assessed against the remaining criteria and three companies were shortlisted. Before being invited in the companies were contacted and asked whether their costs could be reduced, and to revise their proposals to meet the outline budget. All three companies did this.
4 Selection
4.1 The shortlisted companies were invited to give a short presentation to the project team, followed by a discussion with the team and with the Webmaster.
4.2 The outcome of the selection process was as follows:
| Company | Comments | Selected |
|---|---|---|
| SEVEN | Liked the team; demonstrated obvious understanding of our material; own experience directly relevant; able to meet schedule and to meet outline budget. | Y |
| NINE | Good track record, but doubts over personality of team; lacked technical expertise and couldn't answer some questions; more used to working with larger institutions | N |
| ONE | Disappointing presentation; had not understood technical issues; did demonstrate any appreciation of the kinds of subject matter worked with. | N |
The company selected for the project is SEVEN. The team has directly relevant experience of producing web products based on similar materials. After negotiation, they were able to meet the outline budget cost with minimal changes to their original proposal.