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 |
|---|---|---|---|---|
| How do you handle an enquiry from a competitor? | Answer | Answer Production Engine | 2026-08-23 | The draft offered techniques for identifying competitor enquiries. Those techniques are applied to real customers on a suspicion that cannot be confirmed, which is the actual cost, so the page now argues against the identification project entirely and gives the disclosure position that removes the value of the exercise. |
| How do you handle an urgent request from a good customer? | Answer | Answer Production Engine | 2026-08-23 | The draft treated this as a yes-or-no decision. The failure is in accommodating without naming the cost, which sets the precedent and moves the disruption onto another customer, so the page now works from that. Added the deliberate absorbed favour, since a rule with no room for generosity would be wrong for the relationships it most concerns. |
| How do you handle seasonal demand swings? | Answer | Answer Production Engine | 2026-08-23 | The draft treated the trough as a marketing period and optimised the campaign. The peak-selling inversion is the stronger finding and is where the cheap conversions are, so it now leads. Added the self-inflicted seasonality point, which the first version missed and which explains why some troughs deepen year on year. |
| How do you keep quality consistent as you grow? | Answer | Answer Production Engine | 2026-08-23 | The draft proposed documenting the full process. That produces a manual nobody reads while leaving the moments customers judge undefined, so the scope was narrowed to those. Added the founder-absence failure, which grows fastest and which documentation does not address. |
| How do you know if a market is worth entering? | Answer | Answer Production Engine | 2026-08-23 | The draft treated the published counts as a demand estimate. They are structural rather than demand data and carry definitional limits, so both points were stated before the recommendation. Added the attention cost to the existing business, which the first version omitted and which is how new market entries damage the businesses financing them. |
| How do you keep customers informed during a long job? | Answer | Answer Production Engine | 2026-08-23 | The draft advised updating when there was something to report. That produces the missed update the customer reads as a signal, so the no-change update was made explicit and central. Added the requirement to state the completion date every time, since that is what customers open the update for. |
| How do you know if a marketing channel is working? | Answer | Answer Production Engine | 2026-08-23 | The draft recommended multi-touch attribution. At small-business volume the sample cannot support it and the added precision would not change any decision, so it was replaced with consistently recorded self-reported source. Added the intake check, since a handling failure is regularly misread as channel failure and causes working channels to be cancelled. |
| How do you know if you are charging enough? | Answer | Answer Production Engine | 2026-08-23 | The draft used competitor pricing as the primary reference. It cannot answer the question because their costs and scope differ, so it was demoted to a narrow use and the business's own three signals promoted. The unbilled-time section was expanded after the first version treated cost as materials and direct labour, which is where the error usually is not. |
| How do you know when to open a second location? | Answer | Answer Production Engine | 2026-08-23 | The draft led with demand analysis. Second locations fail on management far more often than on demand, so the without-the-owner test now comes first. Added the lower-commitment trial, which produces the same evidence as the location at a fraction of the cost and which the first version omitted entirely. |
| How do you measure visibility in AI answers? | Answer | Answer Production Engine | 2026-08-23 | The draft proposed tracking a visibility score. No defensible scoring method exists and inventing one would have been the failure this site objects to elsewhere, so the page now specifies a rate across a fixed panel and the conditions that make it comparable between runs. Added panel drift, which is how this measurement quietly becomes reassurance. |
| How do you measure whether an automation is working? | Answer | Answer Production Engine | 2026-08-23 | The draft recommended A/B testing as the rigorous option. At small-business volume that is misleading advice — the required sample exceeds what most see in a usable period — so it was demoted to a qualified answer with a channel-split caveat. The counter-metric section was added after the draft would have endorsed an automation improving its own number while degrading conversion. |
| How do you plan a year in a small business? | Answer | Answer Production Engine | 2026-08-23 | The draft was a planning template with objectives and metrics. Templates are what produce the unreferenced twelve-objective document, so the page now works from why plans fail in capacity-constrained businesses. Added written assumptions, since a plan that quietly stopped being achievable is the most common failure nobody notices. |
| How do you migrate customer data without losing anything? | Answer | Answer Production Engine | 2026-08-23 | The draft treated verification as a final step. Staged migration with reconciliation at each stage is what makes a systematic mapping error findable before it affects the whole dataset, so verification was distributed through the process. Added attachments and free-text notes as the specific usual casualties, which a general verification pass tends to miss. |
| How do you stop an AI tool making things up? | Answer | Answer Production Engine | 2026-08-23 | The draft included prompt wording intended to reduce fabrication. That is the instructional approach the page shows does not work, and including it would have undercut the argument, so it was replaced with the structural controls. Added the erosion point, since improving reliability makes the checking habit harder to sustain exactly when the remaining errors are most likely to reach a customer. |
| How do you stop scope creep? | Answer | Answer Production Engine | 2026-08-23 | The draft framed this as needing firmness with clients. The causes are the quote wording and the first absorbed change, both of which are the supplier's, so the page was rebuilt around those. Added the announce-when-absorbing technique, which preserves the pricing norm without making generosity impossible. |
| How do you stop double booking? | Answer | Answer Production Engine | 2026-08-23 | The draft recommended a shared calendar and better discipline. Shared software does not prevent commitments made outside it, which is the actual cause, so the authority rule and the same-day entry expectation replaced it. Added optimistic durations, which produce the same symptom from a different cause. |
| How do you tell customers about a price increase? | Answer | Answer Production Engine | 2026-08-23 | The draft included a softening apology in its suggested wording. That invites the exception requests the page then advises refusing, so it was removed and the reasoning made explicit. Added the pre-decided exception policy, since inconsistency causes more departures than the increase and is entirely avoidable. |
| How do you tell if your marketing agency is working? | Answer | Answer Production Engine | 2026-08-23 | The draft was written as a case against agencies. The intake misattribution makes that unsafe — a substantial share of disappointing results are the client's own enquiry handling — so the page now gives the check that tells the two apart before a supplier is replaced. |
| How do you tell if an AI vendor is selling a wrapper? | Answer | Answer Production Engine | 2026-08-23 | The draft treated being a wrapper as the defect. That is wrong and would have misled a reader into rejecting good suppliers, since nearly all of them build on models they did not train. Reframed around what survives a model swap, and promoted fallback ahead of accuracy because it is both more decisive and almost never asked about. |
| How do you train someone to handle enquiries? | Answer | Answer Production Engine | 2026-08-23 | The draft was built around a call script and a tone guide. A script fails at the first unanticipated question and leaves the handler worse off than no script, so it was reduced to the opening and replaced with the answer sheet and bounded authority. Added the weekly review, which improves handling faster than repeat training. |
| How do you write an email that gets a reply? | Answer | Answer Production Engine | 2026-08-23 | The draft was organised around persuasion and structure. The effort estimate is what determines reply rates and it is formed before persuasion is read, so the page now works from that. Added the deliverability half, since a page about replies that ignores whether the message arrived is answering only part of the question. |
| What quote turnaround should a manufacturer commit to? | Answer | Answer Production Engine | 2026-08-23 | The draft used the first-responder figure as support. This repository has an open research question about exactly that claim's origin, and the honest position is to report the figure as circulating while stating that the citation chain does not reach a primary study. Rewrote the section accordingly and gave the independent reason speed matters, which is the buyer's procurement schedule rather than a race. |
| How long before marketing starts working? | Answer | Answer Production Engine | 2026-08-23 | The draft gave a single timeline with caveats, which is how this question is usually answered and why it stays unresolved. Splitting by mechanism makes it answerable. Repair was added as a third category after the first version left the fastest available return out of the comparison entirely. |
| How long does it take to close a lead? | Answer | Answer Production Engine | 2026-08-23 | The draft offered typical closing times by industry. No admissible source supports them and they would have been read as benchmarks, so they were cut in favour of the structural predictor, which a reader can apply to their own enquiries immediately. Added the two-pipeline consequence, which is the operational point the structural argument exists to support. |
| How long should you chase a shared lead? | Answer | Answer Production Engine | 2026-08-23 | The draft carried a shared-lead close rate as though it were established. It is not: the metric has no independent measurement and both circulating estimates come from interested parties. Removed the figure, made that absence the finding, and replaced the recommendation with a measurement the contractor can run on their own records. |
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.