Loading
Every factual claim in this course traces to something here, grouped by the module it serves. If you find a source that contradicts a lesson, that is worth raising with your teaching agent — the lesson should change, or it should say why it disagrees.
Deciding what deserves your evenings, before you spend any more of them.
WhyExperiment 2, the theater-ticket study: season-ticket buyers who paid full price attended significantly more plays in the first half of the season (an average of 4.11) than buyers given a discount (about 3.3), and the effect had faded by the second half. Paywalled with no reliable free copy — cited here for the finding itself, not as a link you can read in full. Primary source for the sunk-cost lesson.
WhyArgues a live, paying competitor is a good demand signal, not a warning sign — a crowded market means real demand, and a big market with visibly zero competitors is itself the red flag. Primary source for the “somebody is already charging for this” lesson.
WhyDefines toil by six criteria — manual, repetitive, automatable, tactical, devoid of enduring value, and scaling linearly with growth — the frame the maintenance-load lesson borrows to separate a support email that has to be answered right now from work that improves the product for good. Google describing its own operational practice, so read it as the source of the frame rather than as neutral research.
WhyThe $99/year recurring enrollment fee, one half of the maintenance-load arithmetic for anything shipped to the App Store.
WhyThe $25 one-time Play Console registration fee, the asymmetric counterpart to Apple’s annual charge: Google charges a door fee, Apple charges rent.
WhyThe real recurring maintenance tax is not app-store removal, it is the annual minimum-SDK bump — Apple has raised the required Xcode/SDK version every spring for at least three years running, and an app that misses the deadline cannot ship even a one-line fix. Two related Apple pages round out the maintenance picture: its App Review page claims 90% of submissions are reviewed in under 24 hours (Apple’s own self-reported figure, no published methodology), and its App Store Improvements policy removes apps that go three years without an update and draw near-zero downloads, with a 90-day cure window before removal actually happens.
WhyAndroid’s parallel to Apple’s SDK bump, on a deadline that recurs every August 31st: new and updated apps must target API level 36 by August 31, 2026, while existing apps hold at API level 35, and an extension to November 1, 2026 is available on request. Non-compliant apps are not removed — the consequence is delisting-by-degrees, so an app stays installed for everyone who already has it and simply stops being offered to people on newer Android versions.
WhySource of “kill criteria” — a stopping condition combining a measurable state and a date, set before you are in the moment that makes it hardest to think clearly. Primary source for the pursue/park/kill lesson.
WhyThe readable-in-full excerpt behind the “kill criteria” quotes used in the pursue/park/kill lesson. Duke’s own material describes exactly two outcomes, continue or quit — this course’s third verdict, “parked,” with its dated revisit, is the course’s own construct, not Duke’s, and is labelled as such.
How money actually arrives, and what it takes to arrive at your number.
WhySource of “hire” — when we buy a product we hire it to do a job — and, in its American Girl doll example, the one place it explicitly separates the job a child (the user) hires a doll to do from the different job a parent (the buyer) is doing when they pay for it. The article never generalizes that one example into a buyer/user framework; this course’s extension of it into a general principle is its own. Primary source for the “who pays is not who uses” lesson.
WhyGrounds the sharpest version of buyer-not-user: a consumer-style click-through signup creates a binding contract for a school district, and a vendor handling student data has to sit under FERPA’s “school official” exception, which requires the district’s direct control. Why the frictionless self-serve signup that works for a consumer product does not work for a school. Federal procurement rules compound it — 2 CFR § 200.320 requires a district spending federal education funds to use a competitive method (quotations, sealed bids, or proposals) above a set dollar threshold, which rules out “just take a credit card” entirely.
WhyThe worked state-law example on top of the federal rules: purchase contracts over $20,000 must go to the lowest responsible bidder after advertisement for sealed bids, and school districts are covered. The specific number above which selling to a district stops being a transaction and becomes a formal bid process.
WhyWhy a school buys on a calendar rather than in a moment: the district’s annual budget statement must be completed at least seven days before the budget hearing and made publicly available for the fourteen days preceding the annual vote. Miss the window and the next opportunity is next year’s budget.
WhyModels income as separate ladders — time for money, your own services, productized services, selling products — with earning potential and difficulty rising together as you move rightward. Barry is the founder of Kit (formerly ConvertKit), an email-marketing company for creators, and the essay’s own worked example is Kit’s growth; its highest-leverage ladders are exactly what Kit sells tooling for. Read the ladder framing, and treat the audience-building advice as marketing for his own product category.
WhyThe bias disclosure behind the ladders essay, in Barry’s own words: he built his living on email marketing for creators and states Kit is at $45M+ in annual recurring revenue, bootstrapped. Read alongside the essay, because the ladders it ranks highest are the ones Kit sells tooling for.
WhyCurrent entry gate for Mediavine’s on-ramp tier: 1,000 premium sessions over 30 days, not the 10,000- or 50,000-session figures that predate the January 2026 restructure. “Premium sessions” is Mediavine’s own unit, not a plain analytics session.
WhyThe source of the tier gate the lessons rely on: effective January 2026, program eligibility runs on the previous calendar year’s ad revenue, starting at $5,000/year for the Official tier and rising through $100,000, $250,000, $500,000 and $1,000,000. Above the Journey on-ramp, Mediavine’s real gate is a revenue threshold, not a traffic one — and one only the network can see, so you cannot self-assess against it.
WhyThe current 25,000 monthly pageview minimum, plus the geography gate and six-month domain age that come with it — a number that goes stale fast, which is the point: check the vendor’s own page before planning around any threshold repeated here.
WhyThe announcement that cut the minimum from 100,000 to 25,000 monthly pageviews, and the source for dating that change. A network loosening its own entry requirement to a quarter of what it was is competing for supply — read it as a sales position, not a stable rule.
WhyThe two thresholds the derived $16.67 page RPM floor is computed from: a guarantee-eligible site must average at least 100,000 monthly pageviews and have earned $20,000+ in net ad revenue over twelve months. That floor is arithmetic from a vendor’s own eligibility rules, not a published average — no managed network publishes an average RPM at all — and it is what the 300,000-pageview figure in the revenue-shape lesson rests on. The same page also computes the guarantee on sessions while defining page RPM on pageviews, which is the unit confusion worth carrying into any ad forecast.
WhyNo published traffic minimum at all — eligibility is about content quality and policy compliance, not audience size. The lowest floor of the three ad networks named in this course, and correspondingly the lowest revenue per pageview. Google’s own revenue calculator prices that gap: per-category, per-region RPM estimates ranging $0.67–$13.33 (median $2.33 across all 75 category-and-region combinations), carrying Google’s own no-guarantee disclaimer.
WhyWhat the RPM figures are already net of: publishers receive 80% of revenue after the advertiser platform takes its fee, or about 68% when advertisers buy through Google Ads. Google describing the toll booth Google owns — accurate as far as it goes, and it stops exactly where it suits the publisher to stop.
WhyApple’s own published split: 30% commission in a subscriber’s first year, dropping to 15% after one year of paid service. Apple’s Small Business Program drops the rate to 15% on all paid apps and in-app purchases, regardless of subscription age, for developers under $1M in proceeds the prior calendar year. Both app stores’ commissions are under active court-ordered change as of August 2026 — treat this as the durable structure, not a fixed law, and re-check before relying on a specific rate.
WhyThe legacy 15%-then-30% split above and below $1M/year, still current outside the EEA, UK and US. Those three markets moved to a 10%/25% plus a 5% billing fee from June 2026, following the Epic v. Google settlement — a different number depending on where your customer is.
WhyThe arithmetic the “target ÷ price = customers” lesson borrows: 1,000 fans paying $100/year each is $100,000/year, provided you keep the full amount and have a direct relationship with every one of them. Kelly rewrote the essay himself and softened “1,000” to “thousands” — note which version is being quoted. Kelly tested his own claim twice more afterward: “The Case Against 1000 True Fans” (April 2008) went looking for creators actually living on the number and mostly failed to find them, and “The Stars of 1,000 True Fans” (2011) found real examples, but they typically needed closer to 100,000 fans, not 1,000. The sharpest of his revisits, with real numbers, is the entry below.
WhyA working counterexample with real numbers: at $5–$10 per sale, 1,000 fans caps out around $10,000/year, an order of magnitude below the model’s $100,000. The sharpest available challenge to the price term in the arithmetic lesson, and one of Kelly’s own revisits.
WhyThe “raise the price” move in the target-arithmetic lesson, argued directly: invert Kelly’s numbers to 100 fans paying $1,000 a year and the same living needs an audience an order of magnitude smaller, backed with Patreon data on higher-tier pledges growing. Andreessen Horowitz invests in creator-economy platforms, including ones the piece cites, so this is thought leadership validating a category its publisher has money in. The mechanism holds; the enthusiasm is not neutral.
WhyBuilds pricing as arithmetic — a demand curve, revenue as price times quantity, objective value computed from what a customer’s time is worth — with psychology treated as a layer on top of the number, not a replacement for it. Davidson is co-founder and joint CEO of Red Gate Software, a commercial software vendor, and draws the book’s worked examples from selling Red Gate’s own tools; the old Red Gate-hosted copy is dead, so this is the only free download. Primary source for the pricing-is-positioning lesson.
Getting a stranger to pay before the product is finished, the plumbing between a price and money that actually lands, and the first ten who prove it.
WhyDefines commitment as giving up something of value — time, reputation risk, or cash — and is explicit that “commitment can be cash, but doesn’t have to be”: a pre-order or deposit proves intent the way a stated interest cannot, even against an unfinished “duct-tape prototype.” The chapter also closes by naming learning, not revenue, as the real goal of an early sale. Fitzpatrick sells the book in every format; there is no free authorised text. Primary source for “the smallest thing you can charge for.”
WhySource of “never throw more time or money at a problem, just scale back the scope” — the argument that scope, not price, is the variable a solo builder actually controls when something has to give. Used for the scope-versus-price distinction in “the smallest thing you can charge for.”
WhySource of Simple, Lovable, Complete — a narrower scope that is still “complete according to that scope,” distinguishing a smaller product from an unfinished one. Cohen is a two-time founder (SmartBear, WP Engine); the essay is argued opinion from that experience, not measured data. Used for the “smaller scope is not a smaller version of unfinished” distinction in the same lesson.
Why“Retool your expectations of Freemium: It’s a marketing cost” — the reframe that lets the free/paid line get judged by ROI against other acquisition channels without needing a conversion benchmark. Also the source of the “20:1 or even 100:1” free-to-paying ratio, offered as Cohen’s own observation from running two companies, not measured data. Primary source for “the free/paid line.”
WhyThe classic acquisition-only case for a free tier — give a taste away, hook people who upgrade — and it says nothing about what serving those free users costs. Cited in “the free/paid line” as the counterweight Cohen’s reframe answers: the shape of almost every pro-free-tier argument, minus the cost side.
WhyStates that B2C digital services are taxed where the customer resides (Article 58), and that the €10,000 relief in Article 59c is conditioned on the supplier being established “only in one Member State” — so it never reaches a non-EU seller, in the Commission’s own words, “not to suppliers outside the EU.” Primary source for “payment plumbing.”
WhyThe filing simplification open to “any taxable person, not established in the EU” selling digital services into the EU — one registration and one return instead of one per member state. Not an exemption, and not IOSS, which is the separate import scheme for goods under €150.
WhyDescribes verification, not incorporation, as the gate: Stripe’s KYC obligations require “information about your business, product, and relationship to the business,” with no requirement to be a registered company. Also states the business origin country “can’t be changed” after activation without opening a new account — one of the choices “payment plumbing” flags as expensive to get wrong.
WhyThe 2.9% + $0.30 domestic-card rate everyone quotes, plus the US surcharges (+1.5% international, +1% conversion, +0.5% manually entered) — and it is specifically the US rate, not a global one.
WhyThe UK domestic rate, 1.5% + 20p — roughly half the US percentage — plus the EEA and international surcharges. Used to show that quoting “2.9% + 30¢” as a global constant is wrong by a factor of about two for a UK-based seller.
WhyA new account’s first payout typically lands 7–14 days after the first successful payment, longer depending on industry risk and country; the rolling schedule after that is T+2 in the US and Australia, T+3 in most other countries, T+4 in Japan, T+7 in Thailand.
WhyConfirms Stripe itself is the merchant of record under this product (a comparison table names it directly), claims sales tax/VAT/GST compliance in 80+ countries, and lists what it doesn’t support — Connect, embeddable components, third-party tax integrations, one-off invoices. The live case behind “what a platform decision costs later”: this is where Lemon Squeezy sellers are being migrated.
WhyStates that a sole proprietorship is the automatic US default for anyone doing business activity without registering otherwise, and hedges the headline LLC benefit itself: liability protection applies “in most instances,” not absolutely.
WhyDefines a sole proprietor as “someone who owns an unincorporated business by themselves” and routes them to Schedule C, Schedule SE and Form 1040-ES — the filing mechanics behind “no entity required to be taxed” in the US.
WhyThe UK’s equivalent default: self-employed status with unlimited personal liability for business debts, no company formation required.
WhyThe £90,000 UK VAT threshold — and the sentence most people miss on the same page: a business based outside the UK must register “regardless of taxable turnover” once it supplies anything to the UK. The friendly threshold is a domestic relief, mirroring the EU trap in the other direction.
WhySource of the automation test: a digital service must be delivered “automatically… where there’s minimal or no human intervention” — a pre-recorded course qualifies, the same material taught live does not. Page is over four years stale as of this course; treat every figure on it as a pointer to check, not a durable fact.
WhyCalifornia’s $500,000 economic nexus threshold, with no transaction-count trigger at all — unlike the 200-sale figure from South Dakota v. Wayfair that people assume is a US-wide rule.
WhyNew York’s nexus threshold is $500,000 in gross receipts and more than 100 sales over the preceding four sales tax quarters — both conditions required, a different measure from California’s despite the same headline number.
WhyTexas’s $500,000 threshold measures total Texas revenue over the preceding twelve months, including nontaxable sales and services — a third definition of “$500,000” that is not the same $500,000 as California’s or New York’s.
WhyElectronically transmitted digital products — software, eBooks, apps, digital images — are “generally not taxable” in California, but a printed copy or a flash-drive backup bundled with the sale makes “your entire sale usually taxable.” The clearest illustration that nexus and taxability are separate questions.
Why“The legal entity selling goods or services to an end customer,” responsible for collecting sales tax, processing refunds and chargebacks, and PCI compliance. Read with the vendor flag on: Lemon Squeezy sells exactly the service its own docs are describing.
Why“Paddle acts as a reseller of your product, and is, therefore, the ‘seller on record’” — a marketing formulation of a real contractual structure, not a neutral legal finding. Paddle sells exactly the service being described.
WhyStates Paddle is “registered in over 100 jurisdictions worldwide” and that “all the tax-related risk rests with Paddle, not with you” — again a vendor’s own claim about its own contract, cited as attributed marketing, not fact.
WhyThe headlined 5% + $0.50 per transaction, and the sales-team routing for anything under $10 — the same floor Lemon Squeezy sets, which rules out the merchant-of-record path for a $5 product.
WhyThe 5% + $0.50 base fee plus international, PayPal and subscription surcharges, and the vendor’s own worked example: a $20 sale into France nets a $2.06 fee — 10.3% of the product price, because the 5% is charged on the VAT-inclusive total, roughly double the headline rate.
WhyConfirms Stripe’s 2024 acquisition of Lemon Squeezy and the migration path to Stripe Managed Payments, and admits the acquisition “meant some tradeoffs for the Lemon Squeezy community: slower support responses and less frequent product updates.” The live, vendor-documented case behind “what a platform decision costs later.”
WhyNames the “Collison installation,” argues nearly every startup has to recruit its first users manually (“you can’t wait for users to come to you”), and endorses starting narrow on purpose (Facebook at Harvard). The essay argues toward a compounding growth curve for venture-backed startups, not toward ten happy customers — this course borrows the tactic and drops the growth obligation it was built to serve. Primary source for “the first ten are hand-sold.”
Getting anyone to find what you built, choosing where to spend attention, and telling the work that keeps paying from the work that resets to zero.
Why“Poor sales rather than bad product is the most common cause of failure” — Thiel’s diagnosis, aimed at venture-funded startups, that distribution is a design input decided before the build, not a phase to figure out once the product is done. Primary source for “shipping is not launching.”
WhySource of the Bullseye framework — brainstorm all nineteen traction channels, cheaply test the most promising few in parallel, then commit to the one that is working — cited directly: “targeted experimentation with a few traction channels, followed by laser focus on the one that is working.” Weinberg is DuckDuckGo’s CEO and uses DuckDuckGo as one of the book’s own case studies; the evidence base is roughly forty founder interviews, self-reported by survivors. Primary source for “pick one channel and stay on it.”
WhyGoogle’s own vocabulary for what it pays raters to reward: the Know/Do/Website/Visit-in-person intent taxonomy and the dominant/common/minor-interpretation ladder (§12.7), plus “there is no length requirement for Highly Meets results” and different-not-lower expectations for small hobbyist sites (§7.3). Written to shape ratings that train the system Google sells advertising against — a statement of intent, not a description of the algorithm. Primary source for “SEO for a small site.”
Why“We first need to establish what you’re looking for — the intent behind your query” — Google’s public framing of ranking as intent-first, the plain-language counterpart to the rater guidelines’ operational taxonomy.
WhySource of Google’s actual timeline language — “some changes might take effect in a few hours, others could take several months,” wait a few weeks before reassessing — used to displace the “6–12 months” figure no Google page actually states. Also states a small site “usually doesn’t need to do anything except publish,” and that site organisation only matters past a few thousand URLs.
WhyThe bias note behind the SEO lesson’s Google citations: Search & Other, YouTube ads and Google Network together were $82.3bn of $113.8bn in Q4 2025 revenue — about 72% — the quantified basis for “Google is not a neutral narrator about the system it owns.”
Why“Google doesn’t accept payment to crawl a site more frequently, or rank it higher,” paired with “indexing isn’t guaranteed; not every page that Google processes will be indexed” — Google stating plainly that it won’t specify the mechanism closely enough to check the advice.
WhyThe Who/How/Why self-assessment (authorship, disclosure of automation and AI use, and content existing “primarily to help people”) and the E-E-A-T framing, with trust named as “most important” — the current-tense name for guidance that used to be called the Helpful Content Update.
WhyLists the Helpful Content Update as a retired system — folded into core ranking in March 2024, alongside Panda (2015) and Penguin (2016) — the source for why “passes the helpful content system” dates any article that still says it.
WhyThe page the old “four months to a year” Search Central video redirects to; it now states no timeline at all, and instead tells readers to ask a prospective SEO “what kind of results do you expect to see, and in what timeframe” — Google moving from answering the timeline question to telling you to ask it. Also the source for “no one can guarantee a #1 ranking on Google.”
Why“There’s no guarantee that changes you make to your website will result in noticeable impact in search results,” plus the “wait a few weeks” re-check cadence and the one place Google commits to “months” — how long its systems take to decide a site now produces helpful content.
WhyStates that recovery from a core update can be gated on the next core update rather than a calendar date — “that could mean waiting until the next core update” — repeating the same no-guarantee disclaimer found on the traffic-drops page.
Why“We cannot make predictions or guarantees about when or if your URLs will be crawled or indexed,” and that the most common reason a site isn’t indexed is “because it’s just too new” — about indexing, not a ranking penalty, and the source for why the “sandbox”/domain-age claim has no primary backing.
WhyDefines scaled content abuse — “many pages generated for the primary purpose of manipulating search rankings and not helping users” — the named policy violation behind “don’t bother competing on volume,” which targets scale without value rather than the tool used to write.
WhyOpens by telling most readers to stop: “if your site doesn’t have a large number of pages that change rapidly… you don’t need to read this guide,” scoped to sites with a million or more pages — the source for telling a one-person site not to fall into the technical-SEO rabbit hole.
Why“Please don’t use HN primarily for promotion… It’s ok to post your own stuff part of the time” — a qualitative, never-quantified test, plus the flat “don’t solicit upvotes, comments, or submissions” bright line. Written by Y Combinator, a firm with a direct commercial interest in which startups get attention on its own forum. Primary source for “posting in public without being a spammer.”
Why“Can I ask people to upvote my submission? No” — and the penalty reaches “submissions, accounts, and sites,” meaning a solicitation penalty outlives the account and follows the domain.
WhyThe sanctioned channel for posting your own work, with hard conditions: something people can try out (excludes landing pages, sign-up pages, newsletters), non-trivial, and something you’re around to discuss.
WhyDefines spam behaviorally — “repeated or unsolicited actions… that negatively affect redditors” — with no numeric ratio anywhere on the page, and tells a promoter to “post authentic content into communities where you have a personal interest” or else advertise.
WhyThe only place the “10% rule” still lives on Reddit’s own site — presented to moderators as one option some communities choose, sandwiched between “promotional content is not inherently considered to be spam” and “it is ultimately up to you and your team to decide.” Not a sitewide rule.
WhyStates that “each community on Reddit may set its own rules, which are defined and enforced by the moderators of that community” — the source for “local rules govern, and every platform says so itself.”
WhyThe counterexample that kills any universal ratio: “self-promo should be less than a quarter of one’s stories and comments” — a quarter, not Reddit’s retired tenth — plus the structural anti-spam that bites first, an invitation tree and a new-user block on posting unseen domains.
WhyThe only policy document Indie Hackers actually links; its one spam clause is purely technical (auto-responders, scraping) and says nothing about self-promotion, link-dropping, or ratios — the source for the lesson’s negative finding that no sitewide promotion policy exists.
WhyReports that 14% of HubSpot’s own posts had compounded twelve months after publication — “just slightly over the 10% average” from the wider study cited — but publishes neither the denominator nor the methodology behind either figure, which sits behind a gated download. HubSpot sells marketing software; an article arguing content marketing works is closer to a case study for the product than a disinterested finding. Primary source for “compounding versus churning.”
Cost against profit, and how to let go of something you built.
WhySource of the LTV:CAC ratio everyone quotes second-hand — Skok’s actual sentence is “the best SaaS businesses have a LTV to CAC ratio that is higher than 3, sometimes as high as 7 or 8,” a description of top performers, not a passing grade, paired with a payback-period test. Both instruments answer whether it is safe to pour more money into acquisition, which assumes a capital budget a solo builder does not have. Skok is a partner at Matrix Partners, a venture capital firm, and forEntrepreneurs is that firm’s publication — the canonical source for this heuristic, and the firm bias to name whenever it is cited. Primary source for “unit economics at your scale.”
WhyThe standard formula — sales and marketing expenses divided by new customers — and its numerator: “advertising and marketing spend, commissions and bonuses paid, salaries of marketers and sales managers, and overhead costs related to sales and marketing.” Every line item assumes a department; a shelf project with none of them reports a CAC of zero, which the unit-economics lesson uses as the reason to restate CAC in hours instead of dollars.
WhyLists “founder’s time spent on sales” as a cost component and warns that omitting it understates CAC, but folds it back into the ordinary dollar formula at an assumed $50/hour rather than treating hours as CAC’s native unit. Corroboration that the omission is a recognised problem inside vendor content, not a source for the hours-native version, which is this course’s own construct. HubSpot sells the sales and marketing software its own worked example counts as a cost line.
WhySkok’s own follow-up on the defect in the standard LTV formula: revenue divided by churn can return infinity as churn approaches zero. “The old formula … ceases to work properly when you have very long customer lifetimes and negative churn … LTV can become infinite.” Their fix discounts future periods rather than truncating the horizon, and remains an infinite-horizon model, just a weighted one — cutting the horizon off at a stated number of months, which the unit-economics lesson does, is this course’s own move, not this paper’s. Same Matrix Partners bias as the SaaS Metrics piece above.
WhyRestates Skok’s ratio and adds the caveat that it only becomes meaningful once acquisition is repeatable and scalable — calculating it on deals closed through founder relationships produces a CAC with no predictive value. Sleeper is also a Matrix Partners investor: both of this course’s LTV:CAC sources are the same venture firm restating its own figure, which is why the bias note names Matrix Partners rather than either man individually. The unit-economics lesson’s primary citation for the ratio itself is the SaaS Metrics 2.0 piece above; this is the corroborating caveat, not the source of the heuristic.
WhyGrants the right to receive personal data “provided to a controller, in a structured, commonly used and machine-readable format,” conditioned on processing by consent or contract and by automated means, and “shall not adversely affect the rights and freedoms of others.” Article 20 says nothing about closure or wind-down — the shutdown obligation is identical to the day-one obligation. Quoted from gdpr-info.eu rather than EUR-Lex because EUR-Lex returned an empty body across four URL formats and every tool path tried; gdpr-info.eu is a mirror describing its own provenance, not the Official Journal, and the lesson discloses the substitution in its prose. Primary source for “kill criteria written in advance.”
WhyThe CCPA’s data-portability requirement, in the subdivision people miss (§1798.130, not §1798.100): a consumer must be able to receive data “in a readily useable format that allows the consumer to transmit this information from one entity to another entity without hindrance,” and specific personal information “to the extent technically feasible” in a structured, machine-readable format. Cited opposite Article 20 to show the two statutes diverge — California’s applicability gates are numeric and checkable; GDPR’s are not.
WhyThe three bright-line thresholds that gate the Act — over $25 million in gross annual revenue, buying/selling/sharing 100,000+ California residents’ personal information, or deriving 50%+ of revenue from selling it — the only jurisdiction in the kill-criteria lesson whose applicability a reader can check without a lawyer. Says nothing about any other state’s privacy law.
WhyThe worked example of a shutdown done properly: staggered notice (30 days free, 60 days paid), charging stopped at the announcement rather than at close, a stated refund turnaround, an itemised export by named format, and an explicit deletion date. A first-hand record of what was promised, not evidence of what was delivered — not a one-person project, so read as a template rather than a peer, but every mechanism in it scales down. Cited in the kill-criteria lesson’s sunset-mechanics section.
WhyThird-party coverage of the Relay.app wind-down, useful for independently timestamping the announcement date. Inbox Zero sells one of the migration destinations it recommends in the same article — read as interested coverage, not neutral reporting.
WhyThe policy behind “let the registration lapse”: registrars may delete an expired registration at any time (¶2.2.1), followed by a 30-day Redemption Grace Period during which the name can be restored and DNS resolution must be disabled (¶3.1–3.2). Read together, lapsing has no predictable date — the site stops resolving whenever the registrar acts, with no redirect. Cited in the kill-criteria lesson’s domain-is-not-a-footnote section.
WhyThe proof’s author, precise about what kind of system the law describes: “different averages with different dimensions,” computed over a defined period, in a system with a stream of arrivals and a stream of departures. A live shipped product has no departure stream, so the theorem does not describe a portfolio of things you shipped in order that they would stay — a discovery the WIP-limit lesson treats as more useful than the formula itself. Also records that the popular “WIP = throughput × cycle time” form is credited to Hopp and Spearman’s Factory Physics (2000), not derived here. Primary source for “the WIP limit.”
WhyDefines WIP limits as “an enabling constraint” that establishes a pull system and argues from flow and slack — “when resources are fully utilized there is no slack in the system and the result is very poor flow” — never from Little’s Law, which the guide does not mention. Written for teams moving shared work through a process, not for one person carrying standing upkeep. Kanban University sells training, consulting, conferences and certification in the method it is describing.
WhyAnderson, who adapted Kanban for knowledge work, argues the same way as Kanban University’s guide and does not mention Little’s Law either: “it is the WIP limit that ultimately stimulates conversations about process problems.” The David J. Anderson School of Management sells training, Kanban University credentials and a book shop — a practitioner’s confidence in the method he certifies people in.
WhyThe one strand of this literature aimed at an individual: “There are only two real rules with Personal Kanban: 1. Visualize your work 2. Limit your work-in-progress.” Still not a match for a shelf of shipped products — the work it limits is a task list (“people, tasks, responsibilities, deadlines, and even recreation”), and a task list empties the way a shelf of live products does not. Modus Cooperandi sells coaching, consulting, tooling and training in the method.
WhyThe only concrete number the WIP-limit lesson could find for how many live products one person runs — three, at $28K/month combined — and it is exactly what it looks like: a first-person case study, not research, describing one outcome rather than a limit, with the body cut off partway by the site’s own subscriber wall. Even the anecdote does not say how its subject decided what he could carry. Cited as the state of the evidence, not as a source for the lesson’s own construct.
WhySource of the fixed re-decision point the quarterly review borrows: “the betting table is a meeting held during cool-down where stakeholders decide what to do in the next cycle,” and under “Keep the slate clean,” old work never carries over “without first shaping and considering them as a new potential bet.” The six-week cycle itself is reported as “after years of experimentation we arrived at six weeks” — a report of what worked for one company, not a measurement, which is why this course borrows the shape and not the interval. 37signals gives the digital edition away, sells a print edition of it from its own shop, and sells the tool the chapter itself says it built to run the method — “we built Basecamp to execute the techniques in this book.” Primary source for “the quarterly review.”
Running the whole cycle forwards, on something you have not built yet.
WhyArgues an order, not a way of judging a candidate: start with something simple attached to an ecosystem that already has buyers, repeat until it covers your income, only then attempt a standalone subscription business — on runway grounds, since “the biggest pitfall that trips up first-time product people is trying to create something too complex.” Walling co-founded TinySeed, an accelerator that takes equity in companies built on advice like this, and runs MicroConf, a paid conference and community whose members are this framework’s exact audience — both a vendor and an investor interest, worth naming together rather than either alone. Primary source for the “right shape for where you are” question in “choosing project #8.”
WhyCohen’s test for a real idea before you commit: not whether people say yes, but whether independent answers converge on the same problem. His WP Engine research is the convergent case — “the more people I spoke with, the more agreement there was over the pain they had” — set against a different, abandoned idea he researched first, where the “truth” instead diverged into unrelated markets, prices and purposes as he pressed each person. Cohen founded WP Engine and Smart Bear Software; the page links to a book and a course he sells.
WhyThe whole method in one sentence — “if you built this thing, who would buy it?” — write down five real names, and treat failing to reach five as a red flag. Jackson later built Transistor.fm, a podcast-hosting business, so he is an operator too, but the page itself sells nothing, which is unusual enough among this lesson’s sources to say plainly rather than imply a bias that is not there.
WhyDefines reference-class forecasting — “the outside view … is based on knowledge about actual performance in a reference class of comparable projects,” placing a project in a statistical distribution rather than forecasting its specific risks — and states the method’s own entry requirements: a class “broad enough to be statistically meaningful” with “credible, empirical data for a sufficient number of projects.” Seven of your own projects fail both requirements plainly, which the your-own-reference-class lesson says outright rather than papering over. Also splits inaccurate forecasts by cause: honest forecasters face low barriers to the method, while forecasters whose numbers serve someone else’s incentive face “no incentive to debias” and the method’s potential is low. Primary source for “your own reference class.”
Every claim on these pages links to its source. If a source looks wrong or out of date, check the resource list and tell your teaching agent — the course is meant to be corrected.