Reference
Library
Place

TL;DR

Reference Library Place is a human-curated daily reference portal with 205 hand-selected resources across twelve subject departments — no advertising, no algorithmic ranking, no content farms. Each morning it publishes a fresh editorial edition: a featured site, a word of the day, a historical chronicle, a quotation, live news, and market data, all organized with the discipline of a print almanac and the reach of the open web. Access to the full catalog and the AI-assisted Reference Desk is reserved for card-holding readers admitted individually by the editor.

From our perspective at Aaron Cohen Associates, this is exactly the kind of digital library infrastructure that institutions have been failing to build for three decades: a reference environment designed around human judgment, not traffic optimization. The lessons it embeds — curation as editorial responsibility, source authority as a selection criterion, the reference desk as a service model rather than a search bar — are the same lessons we bring to every library planning engagement we undertake.

When the Search Engine Gives You Ten Answers and None of Them Are Right

Consider a scenario we encounter repeatedly in our planning work: a research professional at a mid-sized academic institution needs the controlling legal authority for a specific federal procurement rule. She opens her browser and runs three searches. The first two results pages are law firm blog posts optimized for the query but containing no primary citation. The third is a content aggregator with no date, no author, and no source link. Forty minutes later she has not found the Federal Register entry she needs, though it has been publicly available online since the regulation was promulgated.

This is not a failure of her research skill. It is a failure of the discovery environment. The web's surface has become so saturated with algorithmically generated content that the primary sources — the ones a reference librarian would have handed across the desk in thirty seconds — are buried under layers of paraphrase and optimization. Reference Library Place is built on the proposition that the problem is solvable by the same method that solved it before the internet existed: a trained editor who knows the sources and maintains a list of the trustworthy ones. The digital transformation of library services does not require abandoning the reference librarian's judgment. It requires encoding that judgment into a system people can actually reach.

Refdesk.com and the Case for Human Curation

The site openly acknowledges its debt to refdesk.com, which Bob Drudge began publishing in June 1995. Drudge, the son of a librarian, built Refdesk as an early attempt to organize the chaotic early web as a human-edited reference directory, eventually growing it into what one journalist called "a virtual emporium of utility" that former Secretary of State Colin Powell reportedly used as his browser homepage. Drudge ran the site for over two decades before retiring in 2017.

Reference Library Place describes its own initial catalog as seeded from refdesk.com, calling the original site "a living fossil of 1998 HTML" and positioning itself as what an editorially rigorous, typographically considered descendant would look like. The comparison is meant charitably. The site's About page expresses genuine gratitude to Drudge for three decades of patient curation that the internet otherwise undervalues. The distinction is not disdain for the predecessor but a recognition that its aesthetic had not kept pace with either the web's capabilities or readers' expectations.

From a planning perspective, the lineage matters. The web directories of the mid-1990s — Yahoo, Refdesk, the Open Directory Project — were built on the assumption that a trained human editor could maintain a usable map of the internet's most valuable destinations. That assumption held until the web's growth outpaced the editorial capacity to track it. What Reference Library Place demonstrates is that the assumption still holds when the scope is disciplined: 205 entries across twelve subject departments is a territory a small editorial team can actually maintain with integrity. The lesson for institutional library planning is that focused depth beats shallow breadth every time.

The Catalog: The Site's Core Product

The catalog is the site's core product. At the time of this writing it holds 205 entries distributed across twelve departments: General Reference, News, Science and Nature, Government and Law, Business and Finance, Arts and Literature, History and Biography, Geography and Travel, Language and Learning, Everyday Tools, Technology and Web, and Diversions. Each entry has been selected because it is a primary or authoritative source in its domain, or a genuinely useful tool with no obvious substitute.

The selection criteria are stated plainly on the About page: commercial content farms, SEO spam, and link walls are excluded on principle. The practical effect is a catalog that feels nothing like the results page of a search engine. Under Dictionaries and Thesauri, starred entries include Etymonline for detailed word histories, the Oxford English Dictionary for historical authority, and Roget's International Thesaurus in its classical topical arrangement. Under Government and Law, the Cornell Legal Information Institute sits alongside the Library of Congress and the Federal Register. Under Health and Medicine, the Mayo Clinic and MedlinePlus appear alongside the Cochrane Library's systematic reviews.

The catalog is not trying to be comprehensive in the manner of a web index. It is trying to be trustworthy in the manner of a good reference librarian. This distinction — between comprehensive and trustworthy — is one we make explicitly when planning digital library services from scratch. A collection of two hundred vetted sources that researchers can rely on completely is worth more to any institution than a database of two million records where the signal-to-noise ratio is indeterminate.

What Readers Find Each Morning

Beyond the catalog, Reference Library Place publishes a fresh daily edition each morning that gives the site its almanac character. The centrepiece is the Site of the Day, a single chosen resource accompanied by an editor's note that explains the site's particular merits, its limitations, and when a reader would be best served by it. These notes are written with an editorial voice that is specific and candid rather than promotional.

Surrounding that feature are several standing elements that change daily. The Day's Coinage presents a single word with etymology, pronunciation, and definition, with a preference for terms that have genuine history behind them. The Day's Chronicle offers three historical events on the current date, chosen for substance rather than the kind of trivia that fills most "this day in history" feeds. The Day's Meditation is a single quotation. "Truth Will Out" offers a single verifiable and usually surprising fact drawn from cited sources. The "Born This Day" entry gives a brief biographical note on a notable figure born on the date. Live news headlines from major wire services are refreshed every half hour, and market data is displayed with a link to live quotes. The issue numbering in the masthead reflects the number of days and years since publication began, making the daily character of the publication legible at a glance.

The almanac form deserves a moment's consideration as a planning model. The almanac has been a household reference object since the fifteenth century precisely because it combines a stable structural framework — the calendar — with content that changes on a predictable schedule. Readers know what they will find and where to find it. The surprise is in the content, not the structure. This is exactly the architecture a curated research access portal should follow: predictable navigation, fresh content, authoritative sources.

A Catalog-Grounded AI Assistant

Readers who hold a card have access to the Reference Desk, an AI-assisted tool that answers research questions using the site's own catalog as its primary grounding. The site describes it as running on the most capable models the team can afford, and the design intention is to give the desk access not just to general language model knowledge but to the specific, vetted sources the catalog contains. The desk also maintains a question archive for card-holding readers.

The Reference Desk is not presented as a replacement for the catalog itself but as a complement to it, a way to put a real question to a resource that knows what the catalog holds and how to navigate it. The absence of a password-based login is a small but deliberate detail: access is managed through a six-digit code sent to a registered email address, removing one common friction point without compromising the application-based admission process.

The technical architecture here is instructive for any institution considering AI-assisted reference services. The critical design decision is not which language model to use but what corpus to ground it in. A model grounded in a vetted catalog produces answers that are traceable to sources the editors have already approved. A model grounded in the open web produces answers that are traceable to whatever the training corpus captured, including the content farms and SEO-optimized paraphrase the catalog explicitly excludes. We discuss this distinction at length in our work on digitization lab planning and library technology roadmaps, and it applies with equal force to reference desk design.

Card-Holding Readers, Individually Admitted

The front page of Reference Library Place is open to any reader with no account required. The full catalog, the Reference Desk, the ability to pin entries, and the question archive are reserved for readers who hold a card. Admission is not automated: applications are reviewed and granted individually by the editor, which is consistent with the site's general position that human judgment should sit at the center of its operations.

Once admitted, new readers receive a seven-day free trial. Ongoing access is priced at nine dollars per month or ninety dollars per year, the annual rate representing two months free against the monthly price. Readers who join during the first season are enrolled as Founding Patrons and marked as such in the site's Roll, a form of recognition that is permanent for as long as the publication runs rather than tied to a subscription tier. The site carries no display advertising and states plainly that it never has. Its operating costs are covered by membership and by a separate sponsorship model.

The membership architecture reflects a considered position on what sustainable reference publishing looks like. The site's revenue comes entirely from readers who value it enough to pay, and from sponsors whose products are consistent with editorial character. There is no third-party advertising network whose interests could diverge from the reader's. For institutions exploring digital library development, this model is worth studying: a small, deeply engaged membership community with a willingness to pay for curatorial quality is a more stable foundation than a large anonymous audience monetized through advertising.

One Sponsor Per Edition, One Opportunity Per Institution

Reference Library Place offers a deliberately limited sponsorship model: one sponsor per edition, one placement on the front page and in the day's archive. A sponsored edition takes the form of an identified Site of the Day placement, with the sponsor's name and a tagline set in the publication's own typefaces. There are no banners, no pop-ups, and no retargeting pixels. The rate card lists a single edition at five hundred dollars, monthly patronage at fifteen hundred dollars per month, and annual patronage at fifteen thousand dollars per year. The site works with one sponsor per edition and declines inquiries that are not a fit for the publication's editorial character.

A separate offering is aimed at libraries and other institutions: a branded edition that runs the Reference Library Place catalog and Reference Desk infrastructure under the institution's own name and masthead. The institution controls its catalog, its departments, and its admission policy, while Reference Library Place provides the infrastructure, link checking, and model upkeep. The site describes this as filling a gap that most institutional libraries face: they lack the staff to build and maintain a reference web portal to this standard.

This institutional partnership model deserves particular attention from library directors and academic administrators. The gap the site identifies — insufficient staff to build and maintain a high-standard reference web portal — is exactly the gap we document in service point planning engagements across academic and public library systems. The Reference Library Place partnership model is essentially a staffing solution delivered as a platform: the editorial infrastructure already exists, the institutional client adds its own subject-matter knowledge and admission policy, and the combined product serves its community at a standard that neither party could achieve alone.

The Reference Library Place Model as a Planning Template

Fifty years of library planning has taught us that the reference function is the most underbuilt part of most institutional library programs. Physical reference desks are consolidated or eliminated as a cost measure; the digital reference environment is treated as a byproduct of the institution's existing database subscriptions rather than as a service designed from patron need outward. Reference Library Place is a useful corrective precisely because it was not built by an institution trying to rationalize existing costs. It was built from the reference function itself, asking what the ideal discovery environment for a knowledge-seeking reader would look like and then building toward it.

The planning implications for academic, public, and specialized library institutions are several. First, the catalog model — scope discipline producing trustworthiness — applies directly to institutional database selection: a smaller set of fully integrated, deeply supported resources almost always produces better patron outcomes than a large portfolio of underutilized subscriptions. Second, the daily editorial model demonstrates that a fresh publication schedule is achievable with a small team when the infrastructure is right; institutions building digital library services should ask what their equivalent of the daily edition looks like — what content will give patrons a reason to return to the portal rather than defaulting to a search engine. Third, the institutional partnership model offers a specific procurement option: rather than building a reference web portal from scratch, an institution can license the Reference Library Place infrastructure and focus its own staff resources on catalog curation and community admission.

We bring these same analytical frames to every library master planning engagement we undertake. The reference function does not live only in the physical building. It lives in every discovery environment the institution operates, digital and physical, and the quality of those environments is a direct function of the editorial decisions embedded in their design. Reference Library Place has made those decisions with unusual clarity. The institutions that study them carefully will build better reference services as a result.

For institutions ready to begin that conversation, we are prepared to bring both the planning methodology and the network of digital library infrastructure partners to the table. The first step, as always, is understanding what your community actually needs to find — and then building the discovery environment that gets them there.