Everything the adaptive question bank can teach and test in Marketing, from foundations through advanced practice. Work through it in order, or start practicing and let the questions find your level.
The test of a segment is whether it would respond differently to the same offer. Groups that merely look different — but buy for the same reason at the same price — are a classification exercise rather than a marketing one.
Cheap to obtain and weakly predictive on its own: two households with identical income and age buy entirely differently. It earns its place as a targeting filter once a behavioral segment has been found.
It matters where geography changes the need — climate, regulation, density, distance to a store — and not merely where it changes the map. Applied without that mechanism it produces regions that behave identically.
Values and lifestyle explain choices demographics cannot, and they are expensive and unstable to measure. The usual failure is inventing them from intuition and then finding no way to reach the segment.
It groups by what people have actually done rather than by who they are, which makes it both the most predictive basis and the one requiring data you may not have before launch.
Asking what the purchase is being hired to accomplish cuts across demographics — the same person hires different solutions in different situations, so the SITUATION is often the segment.
All four have to hold at once, and reachable is the one that quietly fails: a segment you can describe precisely but cannot address through any channel is an analysis, not a target.
The choice is about resources against competition. Concentrating suits a small budget facing large incumbents; differentiating costs more and defends more ground; treating the market as one works only where needs genuinely converge.
Owning a small segment produces referral, pricing power and word of mouth that a thin presence across a large one never generates. The risk is a niche that cannot grow, which is why the next segment matters at the outset.
A message written for everyone has to drop every specific, and specifics are what make a claim believable and memorable. Broad copy is not broadly appealing — it is weakly appealing to each reader.
It lives in the customer's head, not in the brief, so it is something you influence and never control. What competitors do and what customers already believe both move it without your involvement.
An internal document rather than a tagline — it exists so that every later decision has a reference to be checked against. Written vaguely it approves everything and therefore constrains nothing.
These are the table stakes that get you considered, and matching them earns no preference at all. Money spent perfecting a point of parity buys entry; money spent on a point of difference buys the choice.
A real one must matter to the segment, be deliverable, and be hard to copy. Fail the third and it becomes a point of parity as soon as a competitor notices it works.
Its value is in choosing the two axes, because those are a claim about what buyers actually trade off. Mapping the dimensions you find convenient produces a tidy chart that describes no decision anyone makes.
You are asking people to unlearn something, which takes longer and costs more than establishing a position from nothing. It usually fails when the product has not changed enough to justify the new claim.
Defining a new frame avoids a comparison you would lose, and it transfers the cost of educating the market to you. Most claimed category creation is positioning inside an existing frame with new vocabulary.
Buyers frequently compare against something outside your category — a spreadsheet, an intern, doing nothing. Naming the wrong frame produces a campaign that argues against a rival the customer never considered.
It is specific on both sides: a named benefit for a named segment. A proposition that would suit any customer of any competitor is a description of the category rather than a reason to choose you.
One claim, held long enough to be remembered. The discipline is refusing the second and third claims, because each one added divides the attention available for the first.
Structuring around what the customer is trying to get done separates the pains they will pay to remove from the gains they would merely enjoy. Most feature roadmaps over-serve the second.
The evidence is behavioral — retention, unprompted referral, reluctance to give it up — rather than expressed enthusiasm. Interest is easy to obtain and says nothing about whether the offer was needed.
Winning one segment completely produces references, case material and a repeatable motion that expansion can use. Expanding before the beachhead holds spends the credibility that was supposed to fund it.
A segment can want the offer, be reachable, and still cost more to serve than it pays. That calculation belongs in the targeting decision, not in a later surprise about margins.
Recognition compounds only when the claim is repeated identically — packaging, support, pricing and sales all state a position whether or not anyone intended them to. Contradiction between them is what erodes it.
B. The marketing mix
The value of the frame is that it forces all four to be decided together: a premium product at a discount price through a bargain channel contradicts itself, and each element is a claim about the others.
Services are produced and consumed at once, so the people delivering them, the process they follow, and the tangible cues a customer can judge ARE the product. They were added because a service has nothing to inspect beforehand.
The core benefit is the reason for buying, the actual product is what is delivered, and the augmented layer — support, warranty, onboarding — is usually where competitors differentiate once the middle layer has converged.
Depth serves one need at several price points; breadth covers several needs. Confusing them produces a catalog that is wide and shallow — present in many categories and credible in none.
The difference has to be perceptible to the buyer before purchase, which is a much narrower set than the differences that exist. An advantage nobody can detect while choosing is an engineering fact, not a marketing one.
Separate names contain risk and cost more to build; a shared name transfers equity in both directions, including the bad direction. The decision is largely about how correlated you want the reputations to be.
It is the last communication before the decision and often the first after it, so it does work no advertising can: it is present at the moment of choice and again at the moment of use.
It means availability — being purchasable at the moment the need occurs — rather than a location. Most place failures are a buyer who wanted the product and could not conveniently get it.
Each layer takes margin and gives reach, control or credit terms in exchange. The decision is what you are buying with that margin, and whether you could buy it more cheaply another way.
Selling direct at a better price than your own retailers undercuts the people carrying your inventory, and they respond by promoting a competitor. Pricing parity across channels is usually the cheapest peace.
Convenience goods need to be everywhere; considered purchases benefit from fewer, better outlets; exclusivity buys retailer commitment with scarcity. Mismatching intensity to purchase type wastes either reach or margin.
Each element does something the others cannot — advertising builds recognition at scale, personal selling handles objections, PR lends third-party credibility. The mistake is treating them as interchangeable spend.
Consistency is not repetition of one asset; it is every channel making the same claim in its own idiom. Customers experience the channels as one brand and notice the contradictions long before the marketer does.
Price cannot maximize share, margin and prestige at once. Naming the objective first is what makes the later argument about a discount resolvable rather than a matter of taste.
It is simple and ignores the only two things that decide whether a price works — what the buyer thinks it is worth, and what alternatives cost. It can leave money on the table and price you out at the same time.
Pricing to perceived worth requires knowing what the buyer compares against, which means it depends on positioning. It captures the most value and takes the most research, which is why it is recommended more than practiced.
Anchoring to the market is safe and cedes the pricing decision to a competitor whose costs and strategy you do not know. It is most defensible where the offers genuinely are substitutes.
Buying share quickly works where scale lowers cost or where switching later is hard. Where neither holds, you have bought customers who came for the price and will leave for the next one.
Capturing early adopters first funds development and segments the market by time rather than by feature. It requires a real reason to buy now, and it trains later buyers to wait.
Charm endings, anchors and decoys work on perception rather than arithmetic, which is exactly why they need care: a tactic that depends on the buyer not examining it degrades trust when examined.
Both let one product serve several willingness-to-pay levels without separate products. The design problem is making the lower tier genuinely useful while leaving a clear reason to move up.
The free tier is an acquisition channel whose cost is real serving expense, so the model works only when conversion is high enough or the free users create value — referral, content, network effects.
Recurring revenue means acquisition is an investment recovered over time, so churn is not a retention metric but the thing that decides whether the acquisition was profitable at all.
It is measured by changing the price and observing volume, not by asking. Stated willingness to pay is systematically wrong in both directions, which is why the test beats the survey.
A discount moves volume now and teaches the market what the product is really worth. Frequent promotion converts a price into a ceiling, and customers simply wait for the next one.
C. Customer journey, funnel, and lifecycle
Two different jobs hide here: making the PROBLEM known and making the BRAND known. A market that does not yet know it has the problem needs education before any brand claim can land.
The buyer is building a shortlist, so the work is being comparable on their criteria rather than being persuasive. Content that refuses to acknowledge alternatives reads as evasion at exactly this stage.
What remains is usually risk rather than desire — the contract, the migration, the fear of looking foolish. Removing a barrier converts better here than adding another benefit.
It is the cheapest revenue in the funnel and the least resourced, because acquisition is visible and retention is the absence of an event. Nobody celebrates a customer who did not leave.
Advocacy is earned by the product rather than bought by the campaign, and it is the only stage whose output feeds the top of the funnel. That loop is what makes it worth more than its direct revenue.
It describes a sequence of mental states, not a sequence of channels, so all four can happen in one page or across six months. Treating each letter as a separate campaign stage is where it is misapplied.
A demo request shown to someone who has just learned the category exists converts nobody and is read as pushy. Matching the ask to the stage costs nothing and is the commonest unforced error in paid media.
A funnel treats each customer as an exit; a flywheel treats them as an input to the next acquisition. The difference is whether retention and advocacy are measured as costs or as channels.
A loop works when using the product produces something that attracts the next user — content, an invitation, a public artifact. Without that mechanical link, growth is a funnel being refilled by spend.
The value is in the steps nobody owns: the phone call, the forwarded email, the conversation with a colleague. A map built only from analytics reproduces the digital path and misses where the decision was made.
Most touchpoints were never designed as marketing — an invoice, a support reply, an error message — and they carry positioning anyway. Auditing them usually finds more contradiction than the campaign does.
A few interactions carry disproportionate weight: first use, first problem, first bill. Effort spread evenly across the journey under-serves exactly the points that decide the relationship.
It is an agreement between two teams rather than a property of a person, so a definition nobody negotiated produces the familiar argument about lead quality that no data can settle.
The distinction that matters is ACCEPTANCE: sales agreeing to work it. A marketing-side definition that sales never accepted measures the handoff from one side only.
Most buyers are not ready when first reached, so the job is to remain useful without asking, and the failure is a sequence that is really a series of asks with an interval between them.
The same message to a new customer and a three-year customer is wrong for at least one of them. Stage-aware messaging is mostly about suppressing what is no longer relevant rather than adding more.
A customer who never reached first value cannot be retained by any later campaign, so activation caps everything downstream. It is marketing work that usually sits in product's budget.
The stages describe the CATEGORY's trajectory, not one company's. Reading your own decline as a market in maturity — or the reverse — misdirects every investment decision that follows.
Introduction pays for education, growth for distribution and share, maturity for defense and efficiency, decline for harvest or exit. Running growth-stage tactics in a mature market is the expensive misread.
The groups differ in what evidence they require, not in enthusiasm. Early adopters accept a promise; the majority wants proof from people like themselves, which is why the same campaign stops working.
The gap exists because the majority's reference group is other mainstream buyers, and early adopters do not count as references for them. The bridge is a completely solved use case for one specific segment.
Five properties predict how fast something spreads, and complexity and observability are the two most often ignored — a product nobody can see being used spreads slowly even when it is excellent.
A lapsed customer already knows the product, so winback is cheaper than acquisition — except where they left for a reason that still holds, in which case the campaign re-advertises the failure.
The hard question is whether the program changes behavior or pays discounts to customers who would have stayed. Measuring it requires a control group, which is exactly what most programs never have.
It is a single stated-intent question, so it captures sentiment at one moment and predicts far less about actual behavior than its ubiquity implies. Useful as a trend, weak as a target.
D. Brand
Because it is a perception rather than an asset you own, it exists whether or not anyone managed it — a company with no brand strategy still has a brand, assembled by customers from whatever they encountered.
The logo is a recognition trigger; the brand is what gets triggered. Redesigning the mark changes the trigger and leaves the association untouched, which is why a rebrand alone rarely changes perception.
The gap between them is the diagnostic. Where image lags identity the communication is failing; where they differ on substance the product is failing, and more communication makes it worse.
It shows up as a price premium, faster trial, and forgiveness after a mistake. Those are measurable, which is what separates equity from a claim that the brand is valuable.
Recognition — knowing it when you see it — is easy to build and weak; recall — naming it unprompted in a buying situation — is what actually gets you into the consideration set.
They accumulate whether intended or not, from packaging, price, the people who use it and where it is sold. Unintended associations are the hardest part of a repositioning to undo.
Repeat purchase can be habit, inertia or a lack of alternatives. Genuine loyalty shows up as resistance when a competitor discounts, which is the only test that distinguishes it.
One name carries equity between products and carries risk the same way; separate names contain both and pay for each build. The decision is really how correlated the reputations should be.
An endorsed sub-brand borrows credibility while keeping room to differ. The lending direction is the one to check: what the sub-brand does will also attach to the parent.
It works when the existing associations are relevant to the new category, which is a question about what the brand MEANS rather than what it currently sells.
A poor fit does two kinds of damage: the extension underperforms, and the parent's meaning blurs. The second cost is larger, slower and invisible in the extension's own P&L.
Each party lends credibility and inherits the other's failures, so the due diligence is on the partner's practices rather than their audience. The association outlasts the campaign.
It is carried by ordinary surfaces — error messages, invoices, support replies — far more than by advertising, because those are where the volume is and where nobody checks.
Their purpose is to hold fixed the few things recognition depends on, so a document that constrains everything equally gets ignored. Name what must never change, and leave the rest to judgment.
Colors, shapes and sounds work only after long consistent use, which makes them expensive to build and cheap to destroy. A refresh that touches them discards the investment that made them work.
Being thought of and being findable are separate failures with separate fixes. A brand people recall but cannot buy, and one stocked everywhere that nobody thinks of, both convert badly.
Buying is triggered by situations rather than by brands, so the practical question is which occasions bring you to mind. Owning more entry points grows a brand more reliably than deepening loyalty.
A values claim invites scrutiny of the practice behind it, and the gap between the two is what gets covered. If the claim would embarrass you under examination, the fix is the practice.
It is warranted when the meaning has genuinely changed — a merger, a new category, an association that must be shed. Done for freshness it spends recognition to solve a problem customers did not have.
Speed and candor dominate: most lasting damage comes from the delay and the hedging rather than from the original event. The instinct to wait for complete information is usually the costly one.
Protection follows distinctiveness and actual use, and it decays without enforcement. A descriptive name is cheap to choose and expensive to defend, which is a marketing decision as much as a legal one.
It only tells you anything as a time series against a consistent question, so the method matters more than the number. Changing the wording resets the baseline and hides the trend you were tracking.
Activation converts existing demand and reports quickly; brand building creates demand and reports slowly. Optimizing on what reports quickly reliably starves the half that makes the other half cheaper.
Presence relative to competitors, not absolute spend, is what predicts share movement — so a budget that grew while the category's grew faster is a cut in the terms that matter.
Hold the recognition assets and vary the execution. Teams tire of their own campaign long before the audience has noticed it, and that impatience is what destroys the consistency compounding depends on.
E. Research and insight
You control the question, the sample and the method, and you pay for all three. Its advantage is fit to your decision; its risk is that a badly designed study is more persuasive internally than no study.
Cheap, fast, and collected to answer somebody else's question — so the definitions, the timeframe and the population rarely match yours exactly. Check those three before citing the number.
They explain mechanisms and generate hypotheses, and they cannot tell you how many. Reporting a count from eight interviews is the error, and it happens because the quotes feel conclusive.
They measure magnitude and can establish difference, and they cannot tell you why. A survey that finds a preference without explaining it usually produces a decision nobody can defend later.
The order is usually qualitative first to find what to ask, then quantitative to size it. Running the survey first fixes the answer options before you know what the options are.
Order effects, scale labels and the options offered all change the answer, so the instrument is part of the finding. Two honest surveys of the same population can disagree entirely on wording alone.
The tell is that the question contains the conclusion — asking how much someone values a benefit presumes they value it. Neutral phrasing produces less flattering numbers, which is the point.
Who COULD be selected decides what the result describes, and it is usually narrower than the claim made from it. Surveying existing customers about why people buy misses everyone who did not.
Precision improves with the square root of sample size, so quadrupling the sample halves the error — and beyond a point more respondents buy very little. Representativeness matters more than size past that point.
The people who answer differ from those who do not in ways usually connected to the topic, so a low response rate is a warning about DIRECTION rather than about precision. More responses collected the same way do not fix it.
People report the behavior they would like to have, so stated recycling, exercise and reading all exceed the observed versions. Ask about the last specific occasion rather than about habits.
What people say they would pay and what they pay diverge systematically. Where a behavioral measure exists — a price test, a pre-order, an abandoned cart — it outranks any stated intention.
One confident participant moves the room, and the output is a group's negotiated position rather than several individual ones. Individual interviews cost more and avoid the mechanism entirely.
Ask about what happened, not what they would do: the last time they faced the problem, what they tried, what it cost. Predictions about future behavior are the least reliable thing an interview can collect.
Reconstructing an actual purchase — what triggered it, what was considered, what nearly stopped it — produces the decision path rather than a rationalization assembled for the interviewer.
The alternative set includes doing nothing, doing it manually, and a different budget line entirely. A map of named competitors misses the option that usually wins.
Its discipline is keeping internal and external separate — a strength is something you control, an opportunity is not. Collapsing the two produces a list of hopes in four boxes.
It is a checklist against blind spots rather than a strategy, and its value is proportional to how specific each entry is. Listing that regulation exists changes no decision.
It explains why an INDUSTRY is profitable rather than why a company is, which makes it an entry-and-exit tool. Using it to explain your own performance attributes firm-level results to structure.
The three figures answer different questions and are routinely quoted interchangeably. The obtainable one is the only one a plan can be built on, and it is the one people skip.
Top-down takes a share of a published total and flatters; bottom-up multiplies real units by real prices and usually shocks. Doing both and explaining the gap is worth more than either number.
Statistical agencies publish method and sampling; industry reports often do not. That difference is what decides how much weight a figure can carry, independent of how precise it looks.
Recency, method, and who paid for it. The third is the one most often skipped and most often decisive — a study commissioned by a party with an interest in the result needs its method read, not its summary.
Testing the idea before building it is cheap, and the common failure is testing a description rather than the thing: people respond to a well-written concept that no product could deliver.
An observation is true; an insight changes what you would do. If the finding leaves every option equally attractive, it was a fact rather than an insight, however surprising it felt.
F. Channels and campaigns
The three differ in control and credibility, inversely: you control paid entirely and it is trusted least, you control earned least and it is trusted most. A plan built on one alone inherits that tradeoff undiluted.
It reaches people who have already stated the need, which is why it converts well and why it cannot create demand. Spend here scales with existing intent rather than generating more of it.
It reaches people who were not looking, so the creative has to do the work of establishing relevance. Applying a search-style bottom-funnel ask to an interrupted audience is the standard misfire.
Enormous reach at low cost per impression, with viewability, fraud and placement quality all eating the difference. The measured metric to check is not impressions but whether a human plausibly saw it.
One is an asset that compounds slowly and stops costing; the other is a tap that delivers immediately and stops when the spend does. Choosing between them is largely a question about time horizon.
It earns attention by being useful before asking for anything, which makes it slow and cumulative. The failure mode is content that is really an ad with an article's formatting.
The only major channel you own outright — no algorithm between you and the recipient — and the one with the clearest consent obligations. That list is an asset precisely because it cannot be bought.
They form a chain, so a weak final number usually has an earlier cause: delivery gates opens, opens gate clicks. Diagnosing a click problem without checking delivery first is the common waste.
Authentication records and sending reputation decide whether mail arrives at all, and a campaign that never lands reports as poor engagement. It is an infrastructure problem wearing a marketing symptom.
High attention, low tolerance, and strict consent rules. The cost of over-sending is not a lower open rate but a permanent opt-out, which removes the channel for that person entirely.
Paying for performance moves risk to the partner and moves your brand into the hands of people you do not manage — whose disclosures and claims you are nonetheless associated with.
You are renting trust that someone else built, and the relationship must be disclosed because that trust depends on the audience believing the opinion is unbought.
The requirement is that the audience can actually notice it: placement, prominence and proximity to the claim. A disclosure below a fold or buried in a hashtag block fails the test it was written for.
Matching the surrounding form is the point and the hazard — it must remain identifiable as advertising, and the better it blends the more explicit the label needs to be.
Earned coverage carries credibility precisely because you did not control it, which is the same reason the message arrives altered. Treating a journalist as a distribution channel reliably fails.
You buy access to a defined audience in a receptive setting, at a cost per contact far above digital. It earns its place where the contact is qualitatively different, not where it is merely bigger.
Broad reach with very limited targeting and almost no direct response, so it does brand work. Judging it on click-equivalent metrics measures the thing it was never doing.
Physical, harder to ignore and expensive per contact, with response measurable by code or unique URL. Its economics work for high-value or geographically concentrated offers.
It re-reaches people already interested, which is why it reports excellent conversion and why that number is misleading: many would have returned anyway, so it claims credit rather than creating it.
The channel has to match both the segment's habits and the purchase cycle's length — a considered six-month purchase needs channels that can sustain contact, not one that delivers a single impression.
One objective per campaign, because the creative, the targeting and the measurement all change with it. A campaign asked to build awareness and drive conversions optimizes for neither.
The insight is the part that cannot be delegated and is most often missing — without it the brief lists requirements and the work becomes a committee compromise.
The same budget buys many people once or few people often, and which is right depends on whether the job is recognition or persuasion. The tradeoff is fixed; the choice is not.
Continuous presence defends memory; bursts create salience around an event. Flighting decided by budget availability rather than by the buying cycle is the usual cause of wasted weight.
Test creative and audience at small spend before scaling, and hold the winner long enough to be sure. Scaling on an early result is how noise gets a budget.
G. Metrics and measurement
It counts delivery, not attention — an ad served below the fold, in a background tab or never scrolled to still counts. Viewability is a separate measure, and the gap between them is large.
Impressions divided by reach gives frequency, so quoting impressions alone hides whether you reached many people once or a few people many times. Those are different campaigns with the same number.
It moves inversely with reach: broadening the audience adds impressions faster than clicks, so the rate falls while the campaign improves. Judging breadth by rate penalizes exactly the right decision.
Comparable only when the denominator is defined identically over time. A change in what counts as a visitor — bots filtered, a new landing page — moves the rate with nothing having improved.
A price, not a result. It falls when you buy cheaper attention, which is often less qualified attention, so an improving cost per click beside a worsening cost per acquisition is the normal warning.
The standard unit for comparing reach-buying across media, and it says nothing about whether the impressions were seen or relevant. Cheap inventory is cheap for reasons.
Whether it is good depends entirely on what an acquisition is worth, so the number is meaningless without lifetime value beside it. Optimizing it alone pushes toward cheap customers who leave.
Fully loaded means salaries, tools and creative as well as media, divided by customers WON rather than leads. Excluding the cost of failures is the standard way this figure is made flattering.
It attributes revenue to spend using whatever attribution model is configured, so it inherits that model's bias — and it usually ignores margin, counting revenue the business never keeps.
It is MARGIN over the relationship, not revenue, and it depends on a retention assumption that is a forecast. A generous churn estimate makes acquisition economics look sound that are not.
It is far more sensitive to churn than to acquisition cost, because retention compounds. A small retention improvement moves it more than a large efficiency gain in media.
The ratio says whether acquisition pays; payback says when — which is the cash question. A healthy ratio with a long payback can still exhaust the money before the return arrives.
Each one is an assumption about how credit should be divided, not a measurement. Switching models changes every channel's apparent performance without anything in the world changing.
It distributes credit across the path, which is more plausible and still not causal — it tells you what preceded conversions, never what caused them. Only a holdout does that.
The lookback period decides which touches exist at all, so a short window structurally favors late channels like retargeting and search, and a long one favors awareness. It is a choice, and it is rarely stated.
The only method that answers whether the spend CAUSED the outcome, because it compares against people who were not exposed. Every attribution model is a proxy for this and none replaces it.
Withholding exposure from a random slice costs the revenue that slice would have produced, which is exactly what makes the measurement real. Teams resist it for the same reason it works.
It estimates contribution from aggregate spend and outcome data without tracking individuals, which makes it robust to privacy changes and coarse — good for budget allocation, poor for tactical decisions.
The test is whether the number could rise while the business worsened. Followers, impressions and page views all can, which is why they are reported when the ones that cannot are unflattering.
One measure that moves only when customers get real value — so it resists gaming by construction. Choosing one that a team can raise without delivering anything defeats the purpose entirely.
Revenue reports what happened; pipeline, qualified traffic and brand search predict it. Managing only on the lagging one means learning about a bad quarter too late to change it.
Small differences on small samples are usually noise, and calling a test early is the most common way noise gets a budget. Set the sample and the duration before starting, and hold to both.
An average can hide two segments moving in opposite directions, so an aggregate that looks flat may contain a large win and a large loss. Splitting by segment is where the decision usually is.
Every metric on a dashboard should have a decision attached to it. Numbers with no decision consume attention and provide cover, which is worse than not reporting them.
Consent requirements and platform tracking limits mean a growing share of activity is unobservable, so measured conversions undercount and the gap is not uniform across channels. Modeling and holdouts are the responses.
H. Law, ethics, and boundaries
The standard is the overall impression on a reasonable consumer, not the literal accuracy of each sentence. An ad composed entirely of true statements can still be misleading, and that is actionable.
The evidence must exist BEFORE the claim is made, not be assembled when challenged. That ordering is the whole rule, and it is what separates a supportable claim from an accurate guess.
You are responsible for what the ad conveys, including implications you did not write. Imagery, juxtaposition and omission all make claims, and they are judged the same way explicit ones are.
Subjective, unmeasurable praise that no one takes as a factual claim sits outside the rules. The boundary is measurability — the world's best coffee is puffery, the world's strongest coffee is a claim.
The evidence bar rises with the potential for harm, so claims touching health, safety or finances are held to a higher standard than ordinary product claims. Testimonials are not evidence here.
Broad terms like green or eco-friendly are unqualified and therefore hard to support. A specific, qualified, substantiated claim about one attribute is both safer and more persuasive.
Comparisons are permitted and must be accurate, current and like-for-like. Stale comparisons are the quiet failure — true when written, false once a competitor changed pricing, and still running.
A testimonial implies typical results unless it is clearly qualified, and burying the qualification does not cure it. Choosing your best outcome and presenting it plainly is the pattern the rules target.
Any relationship that would affect how an audience weighs an endorsement — payment, free product, employment, family — must be disclosed. The test is the audience's weighting, not the size of the benefit.
Placement, prominence and proximity to the claim, in the same medium as the claim. A disclosure only the diligent would find is designed not to be found, and is judged on that basis.
A footnote cannot contradict the headline. Where the main impression is misleading, adding a qualification elsewhere does not repair it — the fix is the headline.
Interface choices that obtain consent the user did not intend — preselected boxes, hidden opt-outs, confusing negatives. They convert well and are increasingly the subject of enforcement, which is the arithmetic to notice.
A reference price has to be one the product was genuinely offered at, and a sale has to be a real reduction. Perpetual sales against an invented list price are the archetypal violation.
The word carries a high bar: any condition — a purchase, a subscription, shipping — must be disclosed clearly and next to the claim, not discovered at checkout.
Where silence renews a charge, the obligations concentrate on disclosure before signup and a cancellation route as easy as the signup was. Requirements vary by jurisdiction; the design principle does not.
Children cannot reliably distinguish advertising from content, so restrictions are tighter on both the persuasion and the data collection. This is a category where the general rules are the floor, not the standard.
Identify the sender, say it is an advertisement where required, include a working opt-out, and honor it promptly across every list. Compliance failures usually come from lists nobody connected.
There must be a lawful basis for collecting and using personal data, and the requirements differ by jurisdiction and by data type. Consent obtained by confusion is not consent in any of them.
Collecting only what the use requires reduces both risk and obligation, and it is the cheapest privacy control available. Data you never collected cannot leak, be subpoenaed or be misused.
Precision makes it possible to target a vulnerability — debt, addiction, grief, illness — and the capability does not make it acceptable. The question is whether the targeting exploits the state or serves it.
Persuasion gives reasons a person can evaluate; manipulation works by bypassing that evaluation. Fabricated scarcity, engineered confusion and pressure are on the wrong side of a line the audience can eventually see.
The pattern is a claim outrunning the practice, and the exposure is proportional to how prominent the claim was. Stating less than you do is safe; stating more is the risk.
Captions, contrast, alt text and keyboard navigation widen the audience and are increasingly a legal expectation. Marketing material is often the least accessible part of a product's surface.
Outcome promises are unsupportable, because the result depends on the market, the product and the customer. They are also a liability — a guarantee is a claim, and a claim requires substantiation.
Every shortcut draws on the same account: the audience's willingness to believe the next thing you say. That balance is invisible until it is spent, which is why the cost always arrives later than the gain.
I. Web and engagement analytics
The definition is a threshold, not a judgment — a visit qualifies by lasting long enough, firing a key event, or viewing enough pages. Because the bar is configurable, the metric is comparable only against itself.
A share of sessions, so it moves when either the numerator or the traffic mix changes. A campaign bringing cheaper, broader traffic lowers it while increasing engaged sessions in absolute terms.
It is the complement of engagement rate under the current definition, which means it is no longer simply single-page visits. A high figure on a page designed to answer one question may be the page succeeding.
Keentune is not affiliated with or endorsed by the organizations whose documentation informs these maps.
All exam, test, and product names and trademarks are the property of their respective owners and are used here for identification and reference only. Keentune is independent study practice — not affiliated with, authorized, or endorsed by any of these organizations.