A library does not work because books are present.
It works because people keep thousands of small decisions aligned long enough for another person to find, trust and use what they need.
Someone selects the book. Someone orders it. Someone verifies the delivery. Someone creates or checks the record. Someone assigns the subject. Someone attaches the item identity. Someone shelves it. Someone helps the user find it. Someone repairs the damaged copy. Someone negotiates the database licence. Someone protects the patron’s privacy. Someone maintains the catalogue. Someone notices that a workflow has begun to fail. Someone trains the next person to keep the whole chain moving.
Library staffing is the mechanism that turns institutional intention into dependable human capability.
This article is part of eduKateSG’s How X Works programme and the How a Library Works series.
The shortest useful answer
Library staffing works by dividing the institution’s required capabilities among people and teams, giving those people the authority, tools, training and information needed to perform their roles, and then connecting the roles through reliable handoffs so the user experiences one coherent service instead of a collection of disconnected jobs.
The important word is capability.
A job title matters less than the capability the institution must preserve.
- Someone must be able to evaluate collections.
- Someone must be able to establish identity and metadata.
- Someone must be able to help users navigate uncertainty.
- Someone must maintain physical and digital access systems.
- Someone must understand rights, privacy and licensing.
- Someone must protect long-term preservation.
- Someone must manage buildings, logistics, technology and safety.
- Someone must coordinate the parts when the parts begin to conflict.
A library fails when an essential capability disappears even if the organisation chart still looks complete.
Staffing begins with the work, not the titles
It is easy to begin with familiar labels: librarian, assistant, technician, manager.
A stronger design begins with the reader job and institutional workload.
What must happen every day for this library to remain useful?
The answer may include selecting resources, processing deliveries, maintaining records, answering research questions, teaching database use, managing loans, protecting privacy, repairing books, scanning archives, negotiating licences, supporting accessibility, running programmes, troubleshooting systems, reporting metrics and preserving institutional memory.
Only after those capabilities are mapped does the staffing structure become meaningful.
This reflects the same logic described in How Library Governance Works: mission defines the desired state, and staffing supplies the people who can make that state real.
The library has a visible workforce and an invisible workforce
Users tend to see the people at service points.
They may not see acquisitions staff, cataloguers, systems librarians, metadata specialists, conservation staff, digital-repository teams, licensing staff, facilities teams, delivery staff or administrators.
Yet much of the quality a user experiences at the desk was produced earlier by people far from the desk.
A catalogue search works well because records were created carefully. A digital journal opens because licensing, authentication and systems were configured correctly. A rare book survives because storage and conservation were managed. A returned title appears in the correct branch because circulation and logistics staff maintained the route.
The public service layer is the visible edge of a much larger operating system.
Frontline staff: where the institution meets uncertainty
Frontline library work is not simply customer service.
Users arrive with incomplete information, uncertain expectations and tasks that do not always fit neatly into policy.
A person may ask where a book is when the real problem is that the book has been mis-shelved. Another may ask for help printing when the real difficulty is an inaccessible file format. A student may ask for “sources on war” when the real assignment requires primary evidence from one specific conflict.
Frontline staff therefore operate as sensors and translators.
They hear what users say, infer what is actually blocking the task, and either solve the problem directly or route it to a more specialised capability.
Reference librarians: converting vague questions into workable routes
Reference work depends on people who can move from uncertainty to evidence.
The reference librarian needs more than subject knowledge. They need question diagnosis, search strategy, source evaluation, citation reconstruction, database fluency, listening and the discipline to know when referral is the correct answer.
A strong reference interaction can involve:
- clarifying the user’s actual task;
- identifying the right source type;
- choosing the correct search system;
- generating useful terminology;
- testing results;
- evaluating authority and currency;
- helping the user judge when the evidence is enough.
Reference work is therefore a high-judgement capability. Automation can assist it, but good diagnosis remains central.
Collection specialists: deciding what deserves institutional attention
Collection development requires staff who can read the difference between popularity and importance.
Selectors may analyse circulation, course demand, publication trends, subject gaps, user requests, reviews, local relevance, research priorities and long-term value.
They also need to understand where the institution should deliberately not spend.
The collection specialist is therefore a portfolio manager of knowledge resources.
Their decisions are constrained by mission, budget, space, licensing, preservation capacity and the wider network of resources available through library consortia and interlibrary loan.
Acquisitions staff: turning intention into controlled custody
Once material is selected, acquisition converts a selection decision into a real institutional resource.
Acquisitions staff may manage orders, suppliers, invoices, subscriptions, gifts, licensing, renewals, delivery discrepancies and budget commitments.
They operate at the intersection of intellectual selection and procurement discipline.
The job is not merely “buy the book.” It is “obtain the right resource, in the right edition or format, under terms the institution can sustain, and hand it correctly into the next workflow.”
Cataloguers and metadata staff: preserving identity before discovery begins
A library user may never meet a cataloguer.
They meet the results of cataloguing constantly.
Catalogue records work because staff analyse what a resource is, record structured metadata, establish relationships, verify identifiers and connect objects to controlled names and subjects.
At scale, small metadata choices compound.
A wrong author identity can misattribute many works. An inconsistent subject can fragment discovery. A missing relation can hide a new edition from users who found the earlier one.
Metadata work is therefore quality control for the library’s knowledge representation.
Authority work: deciding when names are the same and when they are not
Authority control requires especially careful judgement.
Staff may need to decide whether two similar names refer to one person, whether one name belongs to two different people, how an organisation’s former and current names connect, or how older subject terminology should point toward a modern preferred term.
These decisions look small until machines begin using them.
In a digital library or AI-assisted discovery system, a wrong identity merge can propagate across many records and generated answers.
Authority work therefore provides stable identity underneath flexible language.
Circulation teams: maintaining the state of shared resources
Circulation can look routine because many transactions are repetitive.
The deeper job is maintaining truthful state.
Is this copy available? On loan? In transit? On hold? Missing? Returned but not yet shelved? Damaged? Under repair?
Circulation staff also manage exceptions: blocked accounts, disputed returns, lost items, damaged material, unusual borrowing rules and users who need accommodation.
Good circulation work makes a complicated policy-and-state system feel simple to the borrower.
Shelving teams: keeping the physical model aligned with the digital one
Shelving is one of the clearest examples of why apparently simple work can be structurally important.
A mis-shelved book can become functionally invisible even when every catalogue field is correct.
Shelving staff sort, reshelve, shelf-read, shift collections, identify damaged labels and maintain the physical sequence users rely on for browsing.
They are the reconciliation layer between database location and physical reality.
Preservation and conservation staff: representing the future user
Preservation often receives less immediate attention because its beneficiaries may not have arrived yet.
Conservators and preservation staff make decisions on behalf of future users.
They inspect materials, design storage, manage environmental risks, stabilise fragile objects, document treatment, plan disaster response and advise on safe use.
Digital preservation specialists perform a parallel role with checksums, storage redundancy, format monitoring, migration, preservation metadata and repository integrity.
The preservation team protects options that ordinary short-term metrics do not value well.
Systems librarians and technology staff: maintaining the hidden machine
Modern libraries depend on catalogue platforms, authentication, self-check, digital repositories, discovery systems, databases, Wi-Fi, APIs, identity services and vendor integrations.
Systems librarians and technology staff maintain the interfaces between library logic and software.
They may configure loan rules, manage integrations, investigate broken links, troubleshoot authentication, migrate records, monitor vendor changes, protect backups and translate between librarians and technical providers.
This role is especially important because technical defaults can quietly change library policy.
A system configuration can determine how long records are kept, what counts as available, which users can access a database and whether a search interface exposes the full collection.
Technology staff therefore need both technical competence and an understanding of library values.
Digital-library staff: where collections become computational
Digital libraries require staff who can work across content, metadata, repositories, access rights and preservation.
A digital project may involve scanning, quality control, OCR, metadata creation, file naming, repository ingest, rights review, web delivery and preservation storage.
No single step creates the digital library.
The capability lies in the coordinated workflow.
Digital-library staffing therefore often crosses traditional departmental boundaries. A project may need a curator, photographer, metadata specialist, software developer, rights adviser and preservation specialist working on the same object from different angles.
Accessibility specialists: removing barriers before users have to ask
Accessibility becomes stronger when expertise is built into ordinary library work rather than treated as a last-minute fix.
Accessibility specialists or trained staff may review websites, procurement, signage, document formats, event design, assistive technology, service policy and physical routes.
Their role is not simply to respond to individual requests.
It is to reduce the number of unnecessary barriers built into the default system.
Privacy and data specialists: protecting the user’s intellectual space
Library privacy depends partly on staff who understand how data moves.
They may review logging, retention, authentication, vendor contracts, analytics, public-computer configurations and disclosure procedures.
The privacy question often appears inside another workflow rather than arriving as a separate project.
A new discovery platform is also a privacy decision. A new analytics tool is also a privacy decision. An AI reference interface is also a privacy decision.
Staffing must therefore ensure privacy expertise reaches the decisions where data collection is designed, not only where breaches are investigated.
Licensing and copyright expertise
Digital collections have made rights work more central.
Staff may need to interpret licence terms, copyright permissions, digitisation rights, educational use, document supply, text-and-data-mining conditions and restrictions on remote access.
This work protects both access and institutional risk.
Overly cautious interpretation can block legitimate use. Careless interpretation can expose the institution legally.
The job requires disciplined judgement between those extremes.
Programme and outreach staff: helping the collection enter community life
Libraries do not only wait for users to arrive.
Programme, outreach and engagement staff connect collections and services to schools, families, researchers, community groups and underserved users.
They may design reading programmes, exhibitions, workshops, author events, research clinics or partnerships.
The best outreach does not merely generate attendance. It creates a route from a community need into the library’s real capabilities.
Facilities and building operations staff: protecting the environment where knowledge work happens
Library buildings require continuous operational attention.
Facilities teams manage lighting, air-conditioning, lifts, water, cleaning, repairs, safety systems, furniture, environmental monitoring and emergency response.
A failed air-conditioning system is not merely a comfort problem if it threatens rare collections. A leaking pipe is not merely a facilities ticket if shelves sit beneath it. A lift outage is not merely inconvenience if it removes accessible access to an entire floor.
Facilities work becomes library work when building conditions affect access or preservation.
Security staff: protecting without turning the library into a fortress
Libraries are open institutions containing valuable property, vulnerable users and public space.
Security staff and procedures need to protect people and collections while preserving the welcoming character of the library.
The strongest approach is proportional.
Rare collections may require controlled access. Ordinary reading rooms should not feel like restricted zones. Incident response needs clear authority, but routine library use should remain ordinary.
Security staffing therefore balances prevention, response, dignity and public trust.
Managers: coordinating capability, not merely supervising people
A library manager’s deepest job is not attendance checking or meeting scheduling.
It is ensuring the institution has enough capability, capacity and coordination to meet its mission.
That means asking:
- Where is workload growing?
- Where are skills concentrated in one person?
- Which handoffs repeatedly fail?
- Which staff spend too much time repairing avoidable process problems?
- Where is user demand changing?
- Which capabilities will disappear if someone leaves?
- Which work should be automated, shared, simplified or stopped?
Management therefore turns observation into staffing decisions.
The handoff is where teams become a library
Specialisation creates expertise. It also creates boundaries.
A selected title moves to acquisition. Acquisition passes it to cataloguing. Cataloguing and classification prepare it for processing. Processing makes it ready for shelving. Circulation exposes it to users. Preservation may later intervene. Collection development eventually decides whether it remains.
Every transition is a handoff.
Handoffs fail when:
- the next team does not know the item is ready;
- required data is missing;
- ownership is ambiguous;
- different departments use incompatible terminology;
- exceptions are not documented;
- work waits in queues nobody monitors.
Staffing design is therefore partly interface design between teams.
The dangerous phrase: “everyone knows”
Institutional knowledge often hides inside informal habits.
Everyone knows which supplier to call. Everyone knows the temporary workaround. Everyone knows who approves the exception. Everyone knows where the rare uncatalogued box is kept.
Until the people who know leave.
Strong staffing turns critical tacit knowledge into documented institutional knowledge without attempting to document every trivial action.
The test is consequence: if forgetting this would seriously damage service, safety, evidence or continuity, it deserves a durable record.
Single points of human failure
A library may have redundancy in servers and none in expertise.
If only one employee knows how to restore the catalogue, negotiate the largest licence, maintain the digital repository or identify a rare collection, the institution has a human single point of failure.
Cross-training, documentation, shadowing and succession planning create resilience.
The purpose is not to make everybody interchangeable.
It is to ensure that one departure does not erase a critical capability overnight.
Succession planning: preserving capability through turnover
People retire, resign, change roles and take leave.
A mature staffing system assumes change.
Succession planning identifies roles where departure would create disproportionate risk and builds a transfer path before the departure occurs.
This can include updated documentation, paired work, role rotation, mentoring, recorded decision histories and explicit transition periods.
Succession is not merely replacing a name in the organisation chart. It is transferring enough context that the new person can understand why the system looks the way it does.
Training is maintenance
Training is often treated as optional development that occurs after “real work” is complete.
In knowledge institutions, training is part of maintenance.
Cataloguing standards change. Databases change interfaces. Privacy law changes. Assistive technologies evolve. AI tools appear. Vendor platforms change workflows. New forms of research data enter the collection.
Without continuing learning, staff competence decays even when individual effort remains high.
Professional development therefore preserves institutional capability against a changing external environment.
Cross-training improves resilience but has limits
Cross-training allows staff to cover nearby functions when demand spikes or colleagues are absent.
A circulation assistant may learn basic reference triage. A reference librarian may understand enough systems logic to diagnose common access failures. A metadata specialist may learn basic repository ingest.
But cross-training should not pretend specialist work requires no specialist expertise.
The correct design is layered competence: broad shared understanding plus deep expertise where consequences justify it.
Workload is not the same as headcount
Two libraries with the same number of employees can have very different capacity.
Workload depends on opening hours, collection size, user volume, number of branches, digital systems, event programmes, research complexity, language coverage, preservation responsibility and service expectations.
Staffing therefore needs workload models rather than simple employee ratios.
A small specialist library may need highly skilled staff despite modest footfall. A busy public library may require large frontline coverage because service demand is continuous and unpredictable.
Peak demand and base demand
Libraries do not experience uniform demand throughout the day or year.
After-school periods, examination seasons, weekends, programme times and major assignments can create peaks.
Staffing only for average demand produces queues and degraded service during peaks. Staffing only for the highest peak can waste capacity at quieter times.
Flexible scheduling, cross-trained staff, self-service systems and temporary programme support help match capacity to changing demand.
Queues reveal staffing mismatch
Long queues are not automatically evidence of insufficient staff.
They may reveal a badly designed process.
If one staff member must perform six unnecessary checks because two systems do not integrate, adding more people can hide the design flaw without repairing it.
Good staffing analysis therefore separates:
- true demand growth;
- temporary demand spikes;
- avoidable rework;
- system friction;
- policy complexity;
- missing automation;
- understaffing.
The repair should target the actual constraint.
Service desks should not become error buffers for every broken system
Frontline staff often absorb failures created elsewhere.
A confusing website creates more desk questions. A broken authentication system creates more password help. Poor signage creates more directional questions. Inconsistent catalogue records create more failed searches.
If management measures only how efficiently staff handle these contacts, the institution can accidentally optimise the repair layer instead of fixing the source problem.
Good staffing systems feed repeated frontline problems back to the teams that own the underlying mechanism.
Automation should remove low-value repetition before it removes judgement
Self-check machines, automated sorting, batch metadata tools, scheduled reports and digital workflows can reduce repetitive labour.
The strongest use of automation is to remove predictable mechanical steps while preserving human attention for ambiguity, exception and judgement.
A self-check machine can record a loan. It cannot necessarily interpret why a user cannot access a restricted resource. An automated metadata tool can suggest fields. It may not safely resolve a historically ambiguous identity. A chatbot can answer routine opening-hour questions. It should not silently replace specialist research support where evidence quality matters.
Automation works best when the institution knows which part of the job is rule-bound and which part is interpretive.
AI changes the shape of library work
AI can help with search, transcription, translation, summarisation, metadata suggestion, image description, demand analysis, licence review and question triage.
This does not remove the need for staffing.
It moves some work upward.
Staff spend less time generating first-pass outputs and more time deciding whether those outputs are correct, appropriate, lawful, accessible and trustworthy.
The emerging capability is supervision of machine-assisted work.
That requires source discipline, error detection, privacy awareness and clear escalation rules.
AI should not create invisible labour
Automation can appear to save time while shifting work into review, correction and exception handling.
If an AI tool produces thousands of metadata suggestions that staff must verify manually, the institution needs to measure review cost, not only generation speed.
If a conversational interface gives more users access to complex research, specialist staff may receive fewer routine questions and more difficult escalations.
Staffing plans therefore need to observe the whole workflow before declaring automation a capacity gain.
Human review should scale with consequence
Not every machine output needs the same level of scrutiny.
A suggested shelf-display title can tolerate more error than a generated answer about legal research, medical evidence or archival identity.
Staffing should therefore connect review intensity to consequence.
Low-risk work can move quickly. High-impact work requires stronger verification.
This is the staffing expression of the governance principle described in How Library Governance Works: decision process should scale with consequence.
Outsourcing changes who holds capability
Libraries routinely purchase services from vendors.
They may outsource cleaning, security, digitisation, cataloguing support, software hosting, delivery, conservation treatments or analytics.
Outsourcing can increase efficiency and provide specialist capability.
It can also hollow out institutional knowledge if the library no longer understands the function well enough to specify, evaluate or replace the vendor.
A good rule is to distinguish between outsourcing execution and outsourcing understanding.
The institution does not need to perform every task internally. It should retain enough knowledge to govern the task responsibly.
Consortia extend the staffing model beyond one institution
Library consortia allow institutions to share specialist capability.
A small library may not need its own full-time digital-preservation specialist, licensing lawyer or systems architect if a trusted consortium provides that expertise.
This changes the staffing boundary.
The library’s effective workforce becomes larger than the people on its payroll.
But shared capability requires clear service expectations, communication routes and fallback plans. A borrowed capability is still a dependency.
Part-time, contract and volunteer labour
Libraries may use part-time staff, contractors, students or volunteers for appropriate functions.
The design question is whether the work requires continuity, professional judgement, access to sensitive data or long-term institutional knowledge.
Volunteers can strengthen programmes, community connection and support work. They should not become an invisible substitute for essential professional capability that the institution has chosen not to fund.
Temporary labour is useful where the work is genuinely temporary or bounded.
Permanent dependency on temporary labour can be a signal that the staffing model no longer matches the service model.
Shift design and opening hours
Long opening hours create a staffing geometry.
The library must maintain sufficient coverage for safety, public service, supervision and specialist support across the day.
Not every specialist needs to be present at every hour.
This creates layered service:
- core frontline help is available throughout opening hours;
- specialist services may operate during defined windows;
- complex questions can be logged and escalated;
- digital channels can extend service beyond physical presence.
The key is making the service promise clear. A user should know what can be solved now and what requires follow-up.
Remote and hybrid staffing
Some library work depends heavily on physical collections and spaces. Other work can be performed remotely.
Metadata, licensing, digital reference, systems administration, data analysis and some collection work may support hybrid arrangements.
The correct design depends on handoffs.
If remote staff cannot inspect the physical item when necessary, the workflow needs a reliable local partner. If front-line staff cannot reach remote specialists quickly, users experience the organisational boundary as delay.
Hybrid work succeeds when the information and escalation routes are designed, not assumed.
Emotional labour is real library work
Public libraries in particular serve people under stress.
Staff may encounter frustration, loneliness, homelessness, conflict, technology anxiety, grief, language barriers or users navigating government and social services.
This creates emotional labour beyond ordinary transactions.
Training, clear boundaries, escalation routes, security support and realistic expectations protect both staff and users.
Empathy is a capability. It should not require unlimited personal exposure.
Burnout is a systems signal
Individual resilience matters, but persistent burnout can indicate structural mismatch.
If staff are constantly working around broken tools, covering unfilled roles, handling unclear policies or absorbing escalating demand without additional capacity, fatigue is not simply a personal wellness issue.
It is evidence about the operating system.
Good management looks for recurring causes: workload, role ambiguity, poor tooling, insufficient authority, emotional exposure, weak training or unstable priorities.
The repair should change the condition creating the strain, not merely ask individuals to tolerate it better.
Psychological safety improves error detection
Libraries depend on staff noticing problems early.
A staff member may notice that a catalogue migration dropped fields, a vendor is retaining too much user data, a shelf range is dangerously overloaded or a policy is producing inequitable outcomes.
If employees fear punishment for reporting concerns, the institution loses sensors.
Psychological safety does not mean freedom from accountability.
It means people can surface uncertainty, mistakes and risk before those problems become larger failures.
Performance metrics can distort staffing
What gets measured can change behaviour.
If reference staff are judged only by number of questions answered, they may be pushed toward speed rather than research quality. If shelving staff are judged only by items shelved per hour, error rates may rise. If selectors are judged only by circulation, long-term or specialist materials may be neglected.
Staff metrics should therefore reflect the real job rather than the easiest count.
Useful measures can combine volume, quality, error, response time, user outcome and strategic contribution.
The staffing dashboard should watch capability, not only productivity
An institution can appear productive while becoming fragile.
Work gets completed because one expert works excessive hours. A temporary contractor holds critical knowledge. Training is cancelled to protect output. Backlogs stay low because quality checks are skipped.
A better staffing view includes:
- workload and backlog;
- error and rework;
- coverage by skill;
- single-person dependencies;
- training debt;
- vacancy duration;
- turnover;
- overtime;
- service response;
- user outcomes.
These signals reveal whether the workforce is merely busy or actually sustainable.
Staffing debt
Libraries can accumulate staffing debt in the same way systems accumulate maintenance debt.
Positions stay vacant. Documentation ages. Training is postponed. Senior staff absorb more exceptions. New services are launched without retiring old ones.
For a time, the institution still functions.
Then one departure, crisis or technology change exposes how little spare capability remains.
Staffing debt is dangerous because it can be invisible during normal conditions and catastrophic during stress.
Redundancy is not always waste
Efficiency culture can treat overlapping capability as duplication.
Some duplication is resilience.
If two people understand a critical system, the second person may appear unnecessary during ordinary operation. During absence, emergency or succession, that overlap becomes continuity.
The correct question is not “Can we remove every duplicate skill?”
It is “Which capabilities are important enough that losing one person must not eliminate them?”
Specialists and generalists need each other
Generalists help the institution respond flexibly across ordinary needs.
Specialists carry depth where mistakes are costly or knowledge is difficult to acquire quickly.
A library composed only of specialists can become fragmented and expensive. A library composed only of generalists can lack depth when difficult cases arrive.
The strongest model combines broad service competence with specialist escalation paths.
The generalist knows enough to recognise the shape of the problem. The specialist knows enough to solve the difficult version safely.
The staffing boundary changes with library type
A school library, neighbourhood public library, national library and specialist research library should not have identical staff structures.
The necessary capabilities follow mission.
A school library may concentrate on reading development, curriculum support and student guidance. A research library may need subject specialists, data services, scholarly communication and advanced metadata. A national library may require preservation, legal deposit, large-scale digitisation and heritage expertise.
Staffing is therefore a mission-specific architecture, not a universal template.
Recruitment should test the work that actually matters
Hiring becomes stronger when recruitment evaluates the capabilities the role will actually use.
A reference role may need question diagnosis and evidence judgement. A systems role may need troubleshooting and translation between technical and nontechnical teams. A preservation role may need material assessment and documentation discipline.
Credentials can signal preparation. Work samples, scenarios and structured interviews can reveal applied capability.
The aim is not to make hiring theatrical. It is to reduce the gap between the role described and the work actually performed.
Onboarding is the first handoff
A new employee enters an existing institution full of unwritten assumptions.
Good onboarding explains more than procedures.
It teaches mission, service expectations, privacy, escalation, systems, collection structure, decision rights and the reasons behind important rules.
The goal is to shorten the period during which the new staff member can perform tasks but cannot yet recognise when a task has become unusual.
Expertise begins when the person can identify exceptions, not merely repeat normal steps.
Role clarity reduces hidden conflict
Many workplace conflicts are actually ownership conflicts.
Who owns a broken database link? Who decides whether a donation is accepted? Who contacts the vendor? Who can change the loan rule? Who updates the public notice? Who closes the incident?
If responsibilities are unclear, work stalls or is duplicated.
Role clarity defines:
- who decides;
- who executes;
- who must be consulted;
- who must be informed;
- who owns the final outcome.
Clear ownership reduces coordination cost.
Meetings are useful only when coordination requires them
Libraries are collaborative institutions, but collaboration can become expensive if every decision requires everyone.
A meeting is justified when information must be integrated, trade-offs require several functions, or a decision cannot safely be made by one role.
Routine status can often be handled through visible queues, dashboards, written updates or shared systems.
Good staffing protects attention as carefully as it protects headcount.
The escalation ladder
Not every problem deserves senior leadership.
A mature library defines escalation levels.
- routine questions are solved at first contact;
- specialist questions route to specialist staff;
- policy exceptions route to authorised managers;
- legal, privacy, safety or reputational risks escalate further;
- institution-wide issues reach governance.
This prevents two failures at once: frontline paralysis and leadership overload.
The right problem reaches the right level.
When staffing fails
Staffing failures rarely appear first as “staffing failures.”
They appear as:
- longer queues;
- slower cataloguing;
- larger backlogs;
- more broken links;
- less shelf reading;
- delayed preservation;
- shorter reference interactions;
- more reliance on workarounds;
- cancelled training;
- greater dependency on one experienced employee;
- rising errors and complaints.
The organisation can continue moving while capability quietly erodes.
This is why staffing needs leading indicators, not only crisis response.
Repairing a staffing problem
The first instinct is often to hire.
Sometimes that is correct.
Sometimes the better repair is to simplify a process, automate a repetitive step, stop a low-value service, repair a broken integration, clarify authority, redistribute work, build cross-training or purchase shared capability through a consortium.
The repair sequence is:
- Identify the user-visible failure.
- Trace the failure back through the workflow.
- Find the true constraint.
- Decide whether the constraint is capacity, capability, coordination, tooling or policy.
- Apply the smallest repair that restores reliable service.
- Measure whether the failure actually decreases.
Hiring should solve a staffing constraint, not become the default answer to every organisational problem.
How to measure whether staffing works
There is no single staffing metric.
A useful measurement set can include:
- service response time;
- cataloguing and processing backlog;
- error and correction rates;
- reference outcomes;
- system uptime and incident recovery;
- shelf accuracy;
- training completion;
- vacancy and turnover;
- cross-coverage of critical skills;
- user satisfaction;
- staff workload and overtime;
- unresolved risk.
The important question is whether these measurements explain capability rather than simply reward visible activity.
The staffing return path
Every staffing decision should eventually return to the user and collection.
Did the additional role reduce waiting? Did training reduce error? Did automation free staff for more complex work? Did a new specialist increase access to a previously underused collection? Did cross-training reduce service interruption?
If the staffing change does not improve a meaningful library outcome, the institution should inspect the assumption behind it.
Headcount is an input. Capability is the intermediate state. User and collection outcomes are the return.
What AI should not erase
As library work becomes more automated, several human roles become more valuable, not less.
- judging ambiguous evidence;
- understanding local context;
- protecting privacy;
- designing accessible routes;
- recognising when a user’s stated question is not the real question;
- deciding when machine confidence is unwarranted;
- preserving institutional memory;
- taking responsibility for consequential decisions.
These are not leftovers after automation.
They are the governance and judgement layer that allows automation to be used safely.
The complete mechanism
- The library defines the capabilities required by its mission.
- Work is divided among frontline staff, specialists, managers, operations teams, shared services and vendors.
- Roles receive clear authority, tools, training and information.
- Specialists provide depth where ambiguity or consequence is high.
- Generalists provide flexible first-line coverage and routing.
- Handoffs connect selection, acquisition, cataloguing, shelving, circulation, reference, preservation and digital services.
- Documentation and cross-training reduce dangerous single-person dependencies.
- Scheduling matches staff capacity to changing demand.
- Automation removes repetitive steps while preserving human review for consequential judgement.
- Workload, errors, queues, training debt and user outcomes reveal staffing stress.
- Management repairs the real constraint rather than treating every problem as a headcount problem.
- Governance ensures the workforce continues to reflect mission, risk and public value.
That is how library staffing works.
A library’s people are not simply labour applied to books.
They are the living layer that keeps the institution coherent when the collection changes, the technology changes, the building changes, the user changes and the question arrives in a form nobody predicted.
The strongest library is not the one where every employee knows everything.
It is the one where the right capability exists somewhere, the route to that capability is clear, the handoff is reliable, the knowledge survives staff turnover, and the user does not have to understand the organisation chart before the library can help.