Knowledge-Base FAQ¶
Reference repository — instantiates Strategic Caching
Stores approved answers to recurring questions where users can find them — each with an owner and a review date — so experts are not re-asked the same thing and everyone gets the same vetted response.
A knowledge-base FAQ caches the explanations and answers to questions that many people ask over and over, published where the askers already look, so the answer is retrieved rather than an expert re-deriving and re-typing it for the hundredth time. Its defining move is caching human-readable understanding for self-service: the cached item is an answer a reader consumes and applies themselves, not a decision made on their behalf and not machinery that runs. It earns the word strategic only when each answer is deliberately selected for reuse, owned by a named person, scoped to when it applies, and given a review date — otherwise it is just an unmaintained wiki that quietly rots into confident misinformation.
Example¶
A SaaS support team is drowning in the same handful of questions. "How do I change the card on file?" arrives dozens of times a week, and every ticket costs an agent about ten minutes of typing the same explanation. The support lead pulls the top recurring questions from ticket tags and search logs, writes a vetted answer for each, assigns an owner on the billing team, links each answer to the authoritative billing policy it summarizes, and stamps a review date on it. Published in the help center and surfaced inside the in-app support widget, the card-update article now answers most of those askers before they open a ticket.
The payoff is not only faster answers; it is the same answer every time, and agent attention reallocated to genuinely novel problems. The subtlety shows up when billing ships a new card-update flow: the article is now wrong. Because it has an owner, a review date, and a link back to the policy, the drift is caught and corrected rather than misleading readers indefinitely.
How it works¶
- Select from real demand. Mine recurring questions from ticket tags, search terms, and repeat contacts; write up the ones asked often enough to be worth maintaining.
- Answer, don't just point. Each entry is a self-contained, readable explanation scoped to the situations where it applies — not a bare link dump.
- Own it and anchor it. Every answer names an owner and links to the authoritative document it summarizes, so the FAQ stays a subordinate copy, not a rival source of truth.
- Keep it current. A review date and a "was this helpful?" signal drive revision before answers drift.
Tuning parameters¶
- Selection threshold — how often a question must recur before it earns an article. Lower means broader coverage but more entries to maintain.
- Answer granularity — one broad article versus several specific ones. Broad is cheaper to keep current; specific is easier to find and act on.
- Review cadence — how often owners recheck answers against the source. Tighter cadence keeps answers fresher at higher upkeep cost.
- Surfacing — a passive library the user must search versus proactive suggestion inside the support flow at the moment of need.
When it helps, and when it misleads¶
Its strength is collapsing expensive expert re-explanation into cheap self-service while standardizing the answer everyone receives — the common case gets cheaper and consistent, and scarce expertise is reserved for the genuinely new.
Its failure mode is documentation rot: answers drift out of date while still looking authoritative, and because the FAQ is easier to reach than the real policy, readers begin trusting the convenient copy over the source — the archetype's shadow-source-of-truth trap. The classic misuse is padding the base with entries nobody actually asks, or leaving orphaned articles with no owner and no review date. The discipline is to tie every answer to a source of truth and an accountable owner, and to judge the base by ticket deflection[n1] — the questions it actually resolves — rather than by article count.
How it implements the components¶
cache_item_selection— choosing which recurring questions, drawn from demand signals, deserve a maintained answer.reuse_value_model— weighing an answer's ask-frequency and the expert time it saves against the cost of keeping it current, so the base stays curated rather than bloated.freshness_rule— review dates and helpfulness signals set how current an answer must be before it is revised.source_of_truth_reference— each answer names an owner and links to the authoritative document, keeping it a subordinate summary rather than the authority.
It publishes an answer for a human to read and apply, not a binding prior ruling retired by review triggers — that judgment reuse, with its invalidation_rule, is Reusable Decision Precedent's — and it does not sit on the network path serving raw responses (cache_location, tiered_cache_structure), which is Web Cache.
Related¶
- Instantiates: Strategic Caching — the FAQ caches recurring human explanations near the point where people ask, governed by owners and review dates.
- Sibling mechanisms: Web Cache · Memoization · Local Inventory Cache · Prepared Template Library · Reusable Decision Precedent · Cached Approval
Editorial Notes¶
Form Classification¶
Form family: Record, Log & Register
Rationale: The repository durably accumulates approved recurring-question answers with ownership and review provenance.
Nearest alternative: Structure, Architecture & Configuration — A retrieval structure exposes answers, but the accumulating vetted content is the operative asset.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Library & Information Science
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Reference and information-service practice developed curated answer repositories organized around recurring user questions.
Related originating lineages:
- Communication & Media Studies — Audience-centered writing materially shaped concise, approved explanations for self-service use.
- Organizational & Management Science — Knowledge management contributed answer approval, ownership, and review cadence.
Review resolution: Both independent reviews place the primary lineage in library_information_science. The queued differences (alternate_origin_disagreement, domain_reach_disagreement, encyclopedia_synthesis_disagreement) concern secondary metadata rather than primary provenance. The final retains communication_media_studies, organizational_management only where a reviewer supplied a formative-lineage rationale; downstream application by itself is not treated as origin. origin_mode=cross_disciplinary_synthesis records the relationship among origin traditions, while domain_reach=universal records application breadth separately. encyclopedia_synthesis=true reflects whether either reviewer identified a corpus-specific synthesis, and confidence=high preserves the more cautious evidence assessment.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
The FAQ's nearest cousin is Reusable Decision Precedent: both cache human knowledge with an owner and a link to authority. The line is what the reader does with it — an FAQ hands you a readable answer to understand and act on yourself, whereas a precedent hands a decider a prior ruling to apply to a new case. When an FAQ starts being cited to settle disputes rather than to inform, it has quietly crossed into precedent territory and should be governed as such.
[n1] Ticket (or call) deflection rate — the share of would-be support contacts resolved by self-service content before a human is engaged. It is the standard measure of a knowledge base's value and the guard against padding it with articles no one reads. ↩