Method
Editorial log
Every published page on this site, who reviewed it, when, and what the review changed.
Scaled content is penalised when it is published without review. Rather than assert that review happens, the record is kept here and the build will not publish a page whose review block is missing or dated before the page was written.
One collection is reviewed differently and the table says so rather than hiding it. Market pages carry no third-party figure — every publishable fact on one is a named authority, a section number, an institution or a link — so each is checked by retrieving the source and searching it, and the URL, the status, the hash and the matched passage are recorded. A field that cannot be established that way is dropped; a page whose required fields cannot be is not published. Everything that carries a figure somebody else measured is still read by a person before it ships.
| Page | Type | Reviewed by | Reviewed | What the review changed |
|---|---|---|---|---|
| Should a law firm pay for per-lead legal marketplaces? | Answer | Answer Production Engine | 2026-08-23 | Removed a comparison of marketplace cost per signed case against paid search cost per signed case. Only one side of that comparison has a sourced figure and publishing the pair would have implied both were established. The page now compares what can be compared — the published lead costs — and directs the firm to compute the signed-case figure from its own records. |
| Should a law firm put an AI agent on first contact? | Answer | Answer Production Engine | 2026-08-23 | The draft compared signing rates before and after AI intake as though a figure existed. It does not — the question of whether AI first contact has changed signing rates is an open research need in this repository, not a settled result. Removed the comparison and replaced it with the denominator defect, which is the reason no such figure can currently be trusted from any firm's own reporting. |
| Should a firm run intake in-house or use a service? | Answer | Answer Production Engine | 2026-08-23 | The draft picked the answering service as the default recommendation. That was not supportable: which structure wins depends on the arrival shape, and the page's own argument is that the shape is unmeasured. Replaced the recommendation with the measurement and the crossing-point logic, and added the authorisation-scope section because it is the most common reason a correctly priced arrangement still fails. |
| Should a manufacturer pay for directory listings? | Answer | Answer Production Engine | 2026-08-23 | The draft argued against directories on cost-per-RFQ grounds and missed the stronger case for them, which is structured filterability at the elimination stage. Rebalanced, since the shop's own site usually carries no equivalent. Removed a claimed typical win rate on directory RFQs; no independent measurement of quote win rate by source exists and the page now tells the shop to produce its own. |
| Should a manufacturer publish prices or lead times? | Answer | Answer Production Engine | 2026-08-23 | The draft treated publication as a single choice. Splitting it into the disclosures that survive a shortlist versus the ones that price a job is what makes the answer actionable, and the minimum order value moved from a passing mention to its own section once it became clear it is the only item on the list that improves both sides of the ledger. |
| Should a practice add front-desk staff or automate? | Answer | Answer Production Engine | 2026-08-23 | The first pass argued that automation beats hiring, which the evidence does not support and which is not the finding. Rewrote it around the visibility asymmetry between the two queues, which is what actually decides these in practice, and added the wage-series caveat because the county figure is the number most often reached for and it averages clinicians into the front-desk cell. |
| Should a practice add overflow phone cover? | Answer | Answer Production Engine | 2026-08-23 | Cut an estimate of after-hours call share from the draft. There is no admissible source for it and inventing one would have been the exact failure the page argues against. The outside measurement now carries that weight instead, stated in enough detail to be repeated. Added the protected-information boundary, which belongs in the contract rather than in a later discovery. |
| Should a practice charge a no-show fee? | Answer | Answer Production Engine | 2026-08-23 | Removed a suggested fee range from the draft. Nothing in the evidence supports a number, and publishing one would have reproduced the exact defect the page is about — an unqualified figure detached from specialty and payer mix. Replaced it with the ordered sequence, and moved the payer-mix constraint ahead of the pricing discussion because it can remove the option entirely. |
| Should a small business be on social media? | Answer | Answer Production Engine | 2026-08-23 | The draft compared platforms by industry. That advice ages badly and is answerable more reliably by asking the business's own customers, so it was replaced with the one-week method. Added the unmonitored inbox, which is the concrete cost and which the first version treated as secondary to the posting question. |
| Should a practice open online self-scheduling? | Answer | Answer Production Engine | 2026-08-23 | The draft accepted the deflection frame and argued about its size. That was the wrong axis: the deflection claim is unfalsifiable in a saturated practice and the material exposure is on the schedule, not the phones. Restructured around that, and cut a paragraph estimating hours saved, which had no basis and would not have survived the number gate. |
| Should a manufacturer add machine capacity? | Answer | Answer Production Engine | 2026-08-23 | The draft treated the size-band table as a straightforward count. It is not: cells are suppressed for disclosure and an unclassified residual means the bands do not sum, and a shop that read it naively would draw a confident wrong conclusion about a nearly empty band. Both caveats are now in the body rather than in a footnote, and the quoting-bottleneck test was added because it can invalidate the whole case. |
| Should a manufacturer hire an estimator or buy quoting software? | Answer | Answer Production Engine | 2026-08-23 | The draft compared salary against licence directly, which is the comparison the page argues cannot be settled. Restructured around the missing benefit quantity and the log analysis that produces it. Added the within-part-class control after the first version would have overstated the speed effect, and added the cost-model ownership risk, which is where these implementations actually fail. |
| Should a small business use an AI receptionist? | Answer | Answer Production Engine | 2026-08-23 | The draft compared the product against a human receptionist, which is the comparison the sales material invites and the wrong one. Rebuilt around the voicemail comparison. Added the pricing-unit section after noticing that per-call and subscription quotes invert with volume, which means most businesses are comparing two numbers that are not comparable. |
| Should an agent buy portal leads? | Answer | Answer Production Engine | 2026-08-23 | Cut a cost per closing figure. Nothing independent measures it and the page's own argument is that the only trustworthy version is the agent's own. Kept the response-time evidence, which is sourced, and added the distribution-count term, which changes the arithmetic materially and sits in the product description rather than the price list. |
| Should an agent join a team or stay solo? | Answer | Answer Production Engine | 2026-08-23 | The draft compared split percentages against typical team splits. No admissible source establishes a typical split and the figure would have read as a benchmark. Replaced with the per-closing comparison the agent can compute from their own records, and promoted the capacity constraint to its own section because it is what makes the lead flow worth anything or nothing. |
| Should an agent stay licensed? | Answer | Answer Production Engine | 2026-08-23 | The draft framed this as stay or leave. The inactive and referral statuses are a real third option that most states publish and that fits the position many agents facing this invoice are actually in, so it now has its own section. Also added the net-of-split correction after the first version compared gross commission against fixed costs, which overstates the case for renewal. |
| Should you answer calls outside working hours? | Answer | Answer Production Engine | 2026-08-23 | The draft assumed after-hours cover should be continuous and argued for it. The hour-band framing replaced that, because the measurement it recommends usually shows a narrow loss window and buying the wide contract is how firms wrongly conclude cover does not pay. Added the licensing constraint, which is the reason the on-call alternative does not scale and which no answering supplier will raise. |
| Should you answer the phone during a job? | Answer | Answer Production Engine | 2026-08-23 | The draft treated this as a customer service etiquette question. Reframing it as what answers instead makes it actionable, and it exposes voicemail as the default that most businesses have chosen without deciding. Added the explicit exception list, since a rule with an undefined exception does not survive a fortnight. |
| Should you ask customers for referrals? | Answer | Answer Production Engine | 2026-08-23 | The draft proposed a referral programme with an incentive. Incentives are restricted in several of the industries this site serves, and the administration exceeds the return for most small businesses, so both were demoted and the compliance question stated plainly. The describability argument was promoted, since it determines whether any request can work at all. |
| Should you block AI crawlers from your site? | Answer | Answer Production Engine | 2026-08-23 | The draft recommended a specific set of agents to allow and block. Agent names and operator behaviour change often enough that the list would have become wrong while reading as authoritative, so it was replaced with the durable policy shape and an instruction to read current operator documentation. Added the partial-blocking failure, which is common and produces the worst outcome available. |
| Should you charge a service call fee? | Answer | Answer Production Engine | 2026-08-23 | The draft recommended a fee level. There is no admissible basis for one and it would have been read as a benchmark, which is the defect this page objects to in competitor-watching. Replaced with the channel-split rule, and promoted the credit-against-the-job structure from an aside to its own section because it is what changes the customer's reading of the charge. |
| Should you charge for a consultation? | Answer | Answer Production Engine | 2026-08-23 | The draft answered on the value of professional time, which is unresolvable and produces no decision. Rebuilt around what the meeting produces, which is observable. Added the regulated-profession note, since in those contexts the fee is a smaller question than what was said and recorded. |
| Should you cold call? | Answer | Answer Production Engine | 2026-08-23 | The draft assessed cold calling as a single practice and concluded against it. Separating the individually reasoned version from volume dialling is the correction, since the first is legitimate in several of the markets this site serves. The consumer registry rules were moved ahead of the economics because they can settle the question before any assessment. |
| Should you cut marketing spend in a slow quarter? | Answer | Answer Production Engine | 2026-08-23 | The draft advised against cutting. That is not defensible without knowing what the spend produces, and for unmeasured spend the cut is often correct, so the page now sorts by measurement instead. Added the three diagnostic patterns, since a slow quarter is frequently not a marketing problem at all. |
| Should you give a price over the phone? | Answer | Answer Production Engine | 2026-08-23 | The draft recommended deflecting to a site visit. That is the habit the page is about and it loses buyers who were ready, so it was replaced with the range-plus-variables answer. Added the honest-floor section, because a range that no job has landed at is the version of this advice that does damage. |
What the build refuses to publish.
Four checks run before any page in this system is generated, and each one stops the build rather than producing a warning nobody reads.
A page must name, in writing, what is on it that could not appear unchanged on a sibling page — and no two pages may give the same answer. Independently of that, every page is reduced to a fingerprint with its own subject and place names masked out; if two fingerprints match, or come close, both pages are named and the build stops. That is the check that catches a page whose only distinguishing feature was the city in the heading.
A page must also carry at least two datapoints that are specific to its own subject rather than to its category, and if the same figure is claimed as specific on two different pages, it was specific to neither. Pages ship in cohorts with a size cap so indexation can be observed between them, and every page carries the review block that produces the table above.
All four are properties of the schema, not of anyone remembering. Reviewed by Siddharth Sharma, except the 0 market page(s) above, which are checked against their sources and carry the record of it.
GATES
A gate enforced by a person reading pages stops being enforced somewhere around page 60.