What Makes a Document Delivery Service Efficient?

You request a research article. Within hours, a PDF lands in your inbox from a library hundreds of kilometres away. That single transaction hides a complex chain of identifying, locating, procuring, and supplying a document. But getting the document to you is only half the story. The real question is whether the service did this well. A document delivery service (DDS) can fulfil a request slowly, expensively, or in a way that leaves the user frustrated. So how do we judge whether a DDS is actually working? Library science measures its efficiency against three clear yardsticks.

The three pillars of DDS efficiency 🔗

The efficiency of a document delivery service rests on three interconnected factors: speed, cost, and satisfaction level. An ideal DDS should be cost-effective, deliver documents quickly, and satisfy every request it receives. These three factors are not independent of one another. Pushing hard on one often affects the others. The fastest delivery method, for example, is frequently the most expensive. A library aiming for very low costs may end up with slower turnaround times. The art of running an efficient DDS lies in balancing all three rather than maximising any single one.

Understanding why these three pillars matter requires looking at what a DDS is meant to achieve. No library can buy everything its users might need. Studies of interlibrary loan note that a typical parent library may hold only 10 to 20 percent of the references a researcher needs, while the rest must be arranged through interlibrary loan or document delivery. The whole point of DDS is to bridge that gap reliably. If the service is too slow, too costly, or too unreliable, users simply stop trusting it.

Speed: how fast the document reaches the user 🔗

Speed refers to the total time taken from the moment a user submits a request to the moment the document is in their hands. This is often called turnaround time. It covers three stages: receiving the request, processing it (identifying and locating the document), and delivering it. A delay at any one of these stages slows the entire process.

The mode of request and delivery has the biggest effect on speed. Older methods relied on postal mail to receive requests and physical photocopies sent by post to deliver documents. This could take weeks. Performance studies of research libraries once found average borrowing turnaround times of around 16 calendar days. For a student facing a submission deadline, that is far too slow to be useful.

How digital transmission transformed delivery time 🔗

The shift to electronic transmission is the single biggest reason document delivery has become faster. When requests arrive by email or through an online portal, and documents are delivered as scanned electronic copies, the postal delay disappears entirely. Online request and delivery is the fastest method available, although it tends to be more expensive than other modes.

The scale of this improvement is striking. Modern resource-sharing systems built around electronic delivery report an average delivery time of less than 12 hours for digital assets. Compare that to the 16-day average of older systems and the transformation is obvious. The technology removes the slowest part of the chain – the physical movement of paper.

In the Indian academic context, electronic delivery has become the norm. The JCCC@UGC-INFONET gateway run by INFLIBNET allows users affiliated with universities to search for articles and submit requests online, after which designated libraries supply the article electronically. A researcher can browse the table of contents of a journal, view abstracts, and request the full text without leaving their desk. This online workflow is what makes fast turnaround possible across a country as large as India.

Cost: making delivery affordable and sustainable 🔗

The second pillar is cost. Interlibrary loan and document delivery are notoriously labour-intensive, which makes them one of the more expensive services a library offers. Library staff spend significant time on multi-step processes: checking whether an item is held locally, locating potential lending libraries, reviewing lender policies and prices, and tracking the request to completion. Each of these steps carries a staff-time cost.

The numbers make this concrete. A landmark study of North American research libraries found the unit cost to borrow an item through interlibrary loan was around 18 US dollars, while lending an item cost about 9 dollars. These figures cover staff salaries, communication, photocopying or scanning, and delivery. Even though technology has reduced some of these costs since then, document delivery is never free to provide. Someone always pays – the requesting library, the supplying library, or the user.

Centralized versus decentralized cost structures 🔗

How a DDS organises itself has a major effect on cost. There are two broad models. In a centralized model, a single specialized document supply centre holds a large collection and fulfils requests from it. The British Library Document Supply Centre is the classic example. Because requests are routed to one well-stocked centre, processes are standardized and predictable. The trade-off is the heavy cost of building and maintaining that one comprehensive collection.

In a decentralized model, many libraries pool their resources and share with one another. India’s NISCAIR (formerly INSDOC) operates this way, drawing on a decentralized collection held in major libraries across the country. Requests are first met from NISCAIR’s own collection, then from Delhi-based libraries, then from libraries elsewhere in India, and finally from abroad if needed. This spreads the cost of collection-building across many institutions rather than concentrating it in one place.

Decentralized cooperation lowers the burden on any single library, but it brings its own complications. Library cooperation systems have historically faced problems such as keeping a union catalogue updated, an extra inter-lending burden falling on large libraries, and participating libraries withdrawing. Resource-sharing systems that work best are those that standardize and automate the process. When automated routing reduces staff intervention, teams can focus on the difficult requests, and costs fall. Groups where members agree to lend free of charge – such as reciprocal communities – cut transaction costs even further.

Satisfaction level: did the service actually meet the need? 🔗

The third pillar is satisfaction, and it is the most user-centred of the three. A DDS can be fast and cheap and still fail if it does not deliver the document the user asked for, in usable condition, when they needed it. Satisfaction is usually measured through the fill rate (also called the fulfillment rate) – the percentage of requests that the service successfully completes.

Why 90 to 95 percent is the benchmark for success 🔗

A well-run document delivery service is expected to fulfil a very high share of the requests it receives. Modern automated systems demonstrate what is achievable. One major resource-sharing community reports a request fulfillment rate of 95 percent, achieved through a mutual commitment to fast turnaround and free transactions. The National Library of Medicine’s DOCLINE system, which routes biomedical document requests, recorded a fill rate of 92.9 percent across more than 500,000 requests in a recent year.

This is why a fulfillment rate in the 90 to 95 percent range is treated as the benchmark for DDS success. Reaching it signals that the service has the right collection coverage, efficient location and routing processes, and reliable delivery. Older or less automated systems fall short of this mark – performance studies of research libraries found borrowing fill rates closer to 85 percent and lending fill rates as low as 58 percent. The gap between 58 percent and 95 percent is the gap between a struggling service and an efficient one.

Satisfaction, however, is more than a single percentage. It also depends on the quality of the delivered copy, the accuracy of the document supplied, and how the turnaround time matches the user’s actual deadline. Research on academic users has even questioned long-held assumptions about delivery speed, suggesting that beyond a certain point, investing more resources in being faster may not improve satisfaction. A document delivered in eight hours instead of twelve may make little difference if the user did not need it that urgently. Understanding what users genuinely value prevents libraries from spending money chasing speed that nobody asked for.

How the three factors work together 🔗

The reason DDS efficiency is interesting is that the three pillars constantly pull against one another. Maximising speed by using the fastest online delivery raises costs. Minimising cost by relying on slower, cheaper channels can lower satisfaction. Pushing fulfillment rates toward 100 percent may require sourcing rare items from expensive foreign suppliers, which drives up both cost and turnaround time.

An efficient document delivery service is therefore not the one that wins on any single measure. It is the one that finds the right balance for its users. An academic library serving doctoral researchers may prioritise a high fill rate and accept higher costs, because a missing reference can stall a thesis. A service with a tight budget may accept slightly slower delivery in exchange for sustainability. The skill lies in tuning the three factors to the community being served.

This balancing act explains why automation and resource-sharing networks have become so important. Tools that show holdings, policies, and fees before a request is placed, and that automatically route a request to the next library when one cannot fulfil it, improve all three pillars at once. They cut staff time (cost), speed up routing (speed), and increase the chance that some library in the network has the item (satisfaction). For Indian universities, consortium-based systems like UGC-INFONET have played exactly this role, saving the time of research scholars and academicians by making shared collections accessible online.

What do you think? If you were running a document delivery service for your own college library with a limited budget, which of the three factors – speed, cost, or satisfaction – would you protect first, and what would you be willing to trade off to do it? And do you agree that a faster service is always a better one, or are there situations where chasing speed simply wastes money?

References

1. https://ebooks.inflibnet.ac.in/lisp4/chapter/referral-document-delivery-and-translation-services/
2. https://en.wikipedia.org/wiki/Interlibrary_loan
3. https://www.ala.org/tools/atoz/cost-analysis
4. https://exlibrisgroup.com/products/rapidill-interlibrary-loan/
5. https://www.inflibnet.ac.in/services/dds.php
6. https://www.oclc.org/research/areas/systemwide-library/interlibrary-cost-calculator.html
7. https://exlibrisgroup.com/interlibrary-loan-system/
8. https://www.nlm.nih.gov/pubs/techbull/nd25/nd25_docline_40_beta.html
9. https://www.routledge.com/Interlibrary-LoanDocument-Delivery-and-Customer-Satisfaction-Strategies-for-Redesigning-Services/Weaver-Meyers-Stolt-Fong/p/book/9780789003041