The whole issue of trying to attach quantifiable value to archive holdings is a difficult one. It is far more common for archives to talk in terms of the ‘cost of loss’ if they could no longer access their holdings, than it is for archives to predict the positive return on investment from a preservation project that maintains or improves access.

It may be possible to identify some specific items that have known high value by assessing the programme material in the archive, for example major historical events or a popular TV series. But what about the rest of the archive? This is usually the majority of the content!

In general it is not possible to analyse the programme material of every item in the archive to predict. This typically leads to several approaches.

• 1. A ‘mass preservation’ preservation strategy is adopted whereby as much of the archive as possible is preserved. The ‘cost per item’ for this approach is low since economies of scale can be created, and hence the ‘cost per use’ of preserved items is also as low as possible.
• 2. Preservation is done ‘on demand’, i.e. items are only transferred when they are requested for use. This approach can appear attractive, however, the cost per item of digitisation on demand is higher than a ‘batch’ approach.
• 3. General areas of content are identified that have immediate business value or are expected to have business potential. Preservation then targets these areas in preference to other parts of the archive.
• 4. Other measures such as ‘archive usage’ can be used to target preservation at the parts of the archive that are statistically more likely to be requested in the future.

Our report primarily considers the last two cases, i.e. where preservation is targeted as opposed to the first two cases where either a blanket strategy is used or preservation is done on-demand. The cost models in the report can be used in all four cases.

We use the notion that the content of an archive can be grouped into content categories, and some measure of business value or importance can be attached to each of these categories. The report includes a detailed description for a method based on usage statistics, but there are other techniques possible, depending upon expert opinion for example.

Defining content categories and then ordering them in terms of importance is a useful first step in determining what to preserve and what not to.

In summary:

• We assume that a measure of business importance can be attached to items in an archive, which in turn can be used (along with other factors) to prioritise the items when they enter one or more preservation chains.
• Classifying content into categories, e.g. genre, and using statistical measures such as historical usage can be used to make it easier to apply this approach across a large number of items.
• Further refinements are made to take into account the likelihood (or knowledge) of exceptional items in the archive that need special consideration.
• The result is an estimation of the number of items that are targeted for preservation.