10 marketing concepts explained: behavioral triggers and the metrics that judge them

Ten marketing concepts, seven from behavioral science and three from unit economics. The behavioral ones explain how people notice offers, compare options, and feel about what they bought. The economic ones tell you whether any of it was worth funding.

Most marketing advice fails in one of two ways. It borrows a psychology term and turns it into a growth hack, or it reports a metric so flattering that nobody asks whether the business made money.

This guide covers ten concepts that sit on both sides of that gap. Each one gets a plain definition, a real application, the condition under which it stops working, and what to measure. Treat them as hypotheses you can test, not laws.

The three jobs these concepts do

Before applying any of them, know which problem you have.

• Attention and memory, covered by the Zeigarnik effect, the mere exposure effect, and the peak-end rule. These decide what customers notice and recall later.

• Choice and reassurance, covered by charm pricing, loss aversion, choice paralysis, and post-purchase dissonance. These decide how customers compare, commit, and feel afterwards.

• Growth economics, covered by contribution margin, blended CAC, and marginal ROAS. These decide whether revenue and acquisition are worth scaling.

1. The Zeigarnik effect

The Zeigarnik effect is the tendency for unfinished tasks to stay more mentally accessible than completed ones, which creates a pull to go back and finish them.

You meet it a dozen times a day. A progress bar sitting at 60 percent. An abandoned cart email. A LinkedIn profile telling you it is 83 percent complete. A series that ends mid-scene. Each one leaves a task open, and the open task keeps asking to be closed.

How to use it

The mechanism is motivated continuation, not suspense. That distinction decides whether it works.

• Show three of five onboarding steps already done instead of a blank checklist.

• Save partial progress in quizzes and product finders so people can resume where they stopped.

• End an email lesson at a real gap and resolve it in the next send.

• Keep half-built carts, quotes, and configurations alive between sessions.

• Run cart and browse abandonment flows, the most direct open loop in ecommerce.

Where it breaks

The popular version of this claim is stronger than the evidence. Pooled results across studies show no reliable memory advantage for interrupted tasks, though the tendency to resume them holds up better. How much the person cares about the task does most of the work. Interrupt a task nobody was invested in and you get nothing.

The practical failure is manufactured suspense. A subject line that withholds information the reader needed teaches them to ignore you. The loop has to open onto something they actually want.

What to measure

Return rate, step completion rate, time to completion, unsubscribe and complaint rate, and downstream conversion. Always test against a version of the same message that resolves fully.

2. The mere exposure effect

The mere exposure effect is the tendency to prefer things simply because they are familiar. Repeated contact with a brand, face, sound, or shape tends to make people rate it more positively, as long as the repetition is not annoying.

Zajonc demonstrated it in the 1960s, and it quietly underwrites most of brand building. Distinctive assets work because they are repeated, not because any single execution was brilliant.

How to use it

• Keep one color system, one logo treatment, one typographic voice. Familiarity only compounds when the thing being repeated stays the same.

• Buy the same placements repeatedly instead of spreading reach thin.

• Retarget at a cadence you would tolerate yourself.

• Repeat the same product shape, use occasion, and problem statement across channels.

• Work with creators to get repetition of one message through different faces.

Where it breaks

Frequency is a lever, not a strategy. Past a certain point extra impressions produce irritation instead of preference, and that ceiling arrives faster with simple, repetitive creative. Exposure also has to be non-aversive. Repeating an ad people dislike builds familiarity with the dislike.

What to measure

Branded search volume, aided and unaided recall, consideration lift, incremental reach, frequency distribution, negative feedback rate, and creative fatigue curves. Impressions on their own tell you nothing about whether preference moved.

3. The peak-end rule

The peak-end rule says people judge a past experience mostly by its most intense moment and how it ended, rather than by its average quality or how long it lasted.

In experiments on unpleasant episodes, peak and final discomfort predicted overall ratings well while duration added little. This is why a two-hour flight with a miserable baggage claim gets remembered as a bad flight.

How to use it

Map the customer experience, find the emotional high point and the exit, then spend your effort there.

• Checkout and order confirmation, which is the last screen before a wait.

• The first time the product works, usually the emotional peak of a new relationship.

• Delivery and unboxing, an ending you fully control.

• Problem resolution, where recovery done well becomes the moment people quote in reviews.

• Renewal and cancellation, both endings, both remembered.

Where it breaks

This is a design lens, not a full model of a long relationship. It does not license neglect in the middle, and it fits awkwardly onto B2B accounts with many touchpoints and many endings. A staged peak sitting on top of a broken core experience also reads as compensation.

What to measure

Post-interaction satisfaction, review sentiment and star distribution, repeat purchase rate, renewal rate, and referral rate. Compare cohorts before and after you change the ending.

4. Charm pricing

Charm pricing sets a price just below a round number, such as $29.99 instead of $30.00, so the price reads as lower than it is.

Two things drive it. The left-digit effect means people anchor on the leftmost digit and read $9.99 as closer to nine than ten. Separately, the ending itself carries a price image. A nine-ending signals discount and value whether or not the arithmetic misleads anyone.

How to use it

• Use nine-endings where a good deal is the message you want.

• Put a nine-ending on the entry tier of a pricing page to sharpen the contrast above it.

• Keep nine-endings for promotional and clearance pricing, where the ending reinforces the discount.

• Use round numbers for premium positioning. $200 reads more confident than $199.99.

This is the odd-even pricing decision. Odd endings sell savings, even endings sell quality, and luxury, professional services, and trust-sensitive categories usually do better with clean numbers.

Where it breaks

The effect is contextual. Price level, category, and the shopper's goal all change it, so no ending is universally better. The real risk is strategic rather than statistical. A nine-ending that lifts checkout conversion by a point while telling buyers you are the cheap option can be a bad trade for a premium brand.

What to measure

Conversion rate at the price point, average order value, revenue per visitor, mix shift across tiers, and refund rate. Change one ending at a time and hold the price level constant, so you are measuring the ending and not the discount.

5. Loss aversion

Loss aversion is the finding from prospect theory that people judge outcomes against a reference point and feel losses more strongly than equivalent gains.

The practical consequence is framing. "You are losing $360 a year on this plan" lands harder than "you could save $360 a year," even though both describe the same arithmetic.

How to use it

The honest version makes a real cost of inaction concrete.

• Quantify the status quo: hours wasted, fees paid, risk carried, revenue leaked.

• Use free trials and freemium tiers, where people work to keep access they already have.

• State genuine deadlines and real stock counts, which are fine when the constraint is true.

• Frame switching as recovering something they are currently losing.

• Price the risk in B2B. A security page that quantifies an unpatched vulnerability beats one promising peace of mind.

Where it breaks

This is where behavioral marketing most often turns into dark patterns. Invented scarcity, countdown timers that reset on refresh, and fear framing that hides alternatives can all move a weekly number while raising refunds, complaints, and churn. Loss framing also backfires with low-trust audiences and in categories where anxiety is already high, such as health and finance.

What to measure

Conversion rate, but always with guardrails: refund and cancellation rate, complaint volume, review sentiment, repeat purchase rate, and brand perception tracking. A loss-framed test that wins on conversion and loses on retention has not won.

6. Choice paralysis, or choice overload

Choice paralysis, also called choice overload or the paradox of choice, is the drop in action, confidence, or satisfaction that can happen when a choice set is too large or too hard to compare.

The best-known demonstration is the jam study, where a small tasting assortment produced more purchases than an extensive one. It became famous because it inverts the intuition that more selection is always better.

How to use it

The goal is less mental work for the buyer, not necessarily fewer products.

• Recommend one option. A best-for-most-people badge does most of the work on a pricing page.

• Say who each option is for. Buyer fit beats feature counts.

• Build filters around use case and outcome, not spec dumps.

• Show a handful of variants on a listing page and let people drill in.

• Compare three or four differences that matter instead of twenty rows of checkmarks.

• Use guided selling. A three-question finder often beats a full catalog.

Where it breaks

Choice overload is conditional, not automatic. Pooled evidence ties the effect to four conditions: how complex the choice set is, how difficult the task is, how uncertain the buyer's preferences are, and what the buyer is trying to achieve. When people know what they want, large assortments help them, and cutting products in a category where buyers arrive with firm preferences will cost you sales.

So diagnose before you prune. Reduce uncertainty first, then reduce options.

What to measure

Selection rate, funnel completion, time to choose, filter and search usage, pre-sales support contacts, return rate, and post-purchase satisfaction. Time to choose is the most direct read on how hard the decision feels.

7. Post-purchase dissonance

Post-purchase dissonance, commonly called buyer's remorse, is the doubt customers feel after committing, especially when the alternatives were attractive or their expectations are still unresolved.

It is worse online, because the buyer never touched the product before paying. Everything they believe about it came from your copy and your images, so the gap between promise and reality shows up at delivery.

How to use it

Treat the period after checkout as a designed stage, not a receipt.

• Restate the use case they bought for, so the confirmation reads as more than a receipt.

• Set honest expectations on delivery windows, setup effort, and total cost including fees.

• Engineer a first success. Proof the thing works ends doubt faster than reassurance does.

• Send reviews and photos from buyers of the same product after the sale.

• Make the path to a human obvious. It reduces regret more than a discount does.

• Keep returns and warranty terms next to the buy button rather than in a footer.

Where it breaks

You cannot communicate your way out of a product that underdelivers. Post-purchase messaging that oversells while the experience disappoints speeds up the return. Sending too much email after checkout creates its own irritation. Resolve dissonance with product truth, not volume.

What to measure

Activation rate, time to first value, refund and return rate, exchange rate, repeat purchase rate, support ticket volume and topics, review sentiment, and referral rate.

8. Contribution margin

Contribution margin is revenue minus variable costs. Per unit, it is the money each sale leaves behind to cover fixed costs and then profit.

Contribution margin = revenue minus variable costs

Contribution margin ratio = contribution margin divided by revenue

Take a $100 order carrying $40 of variable cost across product cost, shipping, payment fees, and pick and pack. Contribution margin is $60 and the ratio is 60 percent. That $60 is the entire budget available for acquisition, overhead, and profit on that order.

Contribution margin compared with gross margin

Gross margin subtracts cost of goods sold. Contribution margin subtracts every variable cost, so it also takes out shipping, payment processing, and fulfillment. Gross margin is the better read on product and pricing health. Contribution margin is the number that should drive marketing decisions.

Many operators split it further. A first level takes out product cost, a second takes out fulfillment, and a third takes out marketing. The level after fulfillment and before marketing is the one that sets your acquisition ceiling, because it is what an order genuinely brings home before you spend to win it.

Why it matters for marketing

Contribution margin is how you set a defensible maximum CAC. A campaign can produce impressive revenue and still be unfundable if product cost, shipping, fees, commissions, and variable service cost eat most of it. Your target CAC should sit clearly below that ceiling so fixed costs and profit have room.

What to measure

Contribution margin by product, by channel, by campaign, and by customer cohort. Check it after discounts and after returns, because both quietly compress it.

9. Blended CAC

Blended customer acquisition cost is total acquisition spend divided by all new customers in the same period, whichever channel brought them.

Blended CAC = total acquisition spend divided by new customers acquired

Blended CAC compared with paid CAC

Blended CAC counts all acquisition spend against all new customers, and it answers whether the business is acquiring efficiently overall. Paid CAC counts only paid media, agency, creative, affiliate, and paid influencer spend against the customers attributed to paid, and it answers whether paid media is doing its job.

Paid CAC is almost always the higher of the two, because blended CAC gets to divide by customers that organic, brand, referral, and content brought in at no incremental media cost. The gap between the two numbers shows how much your organic engine is subsidizing your paid math. When paid CAC rises but blended CAC holds steady, organic is quietly covering for you, and that is worth knowing before you cut the content budget.

Define it before you report it

Blended CAC has no single standard definition, which is why two people can calculate it for the same company and disagree by 40 percent. Write down and circulate five decisions.

• Marketing cost only, or fully loaded with sales salaries and tooling.

• Which line items count as acquisition rather than retention.

• Who qualifies as a new customer, and whether reactivated customers count.

• The time window, and whether spend and customers use the same one.

• How returns, cancellations, and refunds are handled.

Where it breaks

Blended CAC is an operating benchmark, not a diagnostic. It reports the business-level result and says nothing about which channel caused it. Read it alongside paid CAC, marketing efficiency ratio, CAC payback period, and the ratio of lifetime value to CAC. The commonly cited three-to-one target on that ratio hides enormous variation by category and margin, so treat it as a starting point rather than a rule.

10. Marginal ROAS

Marginal ROAS is the return on the next increment of spend, rather than the average return across all spend.

Marginal ROAS = change in conversion value divided by change in ad spend

Average ROAS compared with marginal ROAS

Average ROAS divides total conversion value by total spend, so it answers what everything you spent returned. Marginal ROAS answers what the next dollar will return. Average ROAS stays flattering as budget grows. Marginal ROAS declines. Use the first for reporting and the second for deciding whether to scale.

Why the difference decides budgets

Your first dollars buy the cheapest, highest-intent demand: people already searching for you, retargeting pools, warm audiences. Average ROAS carries the memory of those cheap conversions forever. Marginal ROAS does not. As budget grows you reach people further from purchase, and each additional dollar returns less.

This is why a campaign can report 4x average ROAS while the last increment of spend runs at 1.2x and loses money against contribution margin. The scaling question is never whether your ROAS is good. It is whether the next increment clears the hurdle your contribution margin sets.

Set the hurdle from margin

If contribution margin is 60 percent, spend breaks even at roughly 1.7x. Anything below that on the next increment is buying revenue at a loss. The hurdle belongs to the increment, not to the account average.

The measurement warning

Platform-reported conversion value is attribution, not causation. Both average and marginal ROAS built from in-platform numbers inherit every attribution artifact, and marginal ROAS is not directly observable from platform data at all. Where the spend decision is material, validate it.

• Geo holdout and geo lift tests.

• Randomized conversion lift studies.

• Scheduled spend-down or blackout tests.

• Media mix models calibrated against experiment results.

What to measure

Marginal ROAS by increment and by channel, break-even ROAS derived from contribution margin, incremental CAC, marketing efficiency ratio, and the ratio of platform-reported conversions to experimentally measured ones.

How to put these together

Ten concepts is too many to apply at once, so sequence them.

Diagnose the bottleneck first. Is the problem attention, decision friction, buyer anxiety, or unit economics? Applying a behavioral concept to a measurement problem is the most common waste on this list. If marginal ROAS is below break-even, no amount of charm pricing fixes it.

Write the mechanism before the test. State why the change should work, which customer it helps, and what could go wrong. A hypothesis you cannot state is one you cannot learn from.

Change one meaningful thing. Pick a primary metric, pick guardrail metrics, and decide the pass rule before you look at the data.

Convert the result into money. Translate the win into contribution margin, CAC, payback period, and retention, and into incremental return where you can measure it. A conversion lift that lowers contribution margin is not a win.

Keep the persuasion honest. Clear comparisons, real deadlines, and better onboarding compound over time. Hidden alternatives and invented urgency borrow from next quarter.

Six questions for your next campaign review

Ask these before any budget increase.

• What reference point is the customer using, and are we clarifying a real loss or inventing one?

• Which options genuinely differ for this buyer, and can we make the choice easier without hiding useful alternatives?

• What is the customer's first proof of value, and does the post-purchase flow get them there?

• Does the price ending reinforce the brand signal we want, or contradict it?

• What contribution margin and marginal ROAS must the next budget increment clear?

• Which result is merely attributed, and what evidence suggests it is incremental?

FAQ

Does charm pricing still work?

Often, but not everywhere. The left-digit effect and the value signal a nine-ending carries are both real, and nine-endings are still common across retail. The size of the effect depends on price level, category, and what the shopper is trying to do, and round prices tend to suit premium and trust-sensitive positioning better. Test it instead of adopting it as a default.

Is the paradox of choice real?

The effect is real but conditional. Large assortments reduce action mainly when the set is complex, the comparison is hard, or the buyer is unsure what they want. When preferences are clear, more options help. Reduce difficulty before you reduce inventory.

What is the difference between contribution margin and gross margin?

Gross margin subtracts cost of goods sold. Contribution margin subtracts every variable cost, including shipping, payment processing, and fulfillment. Contribution margin is the more useful number for marketing, because it defines what you can actually afford to spend acquiring a customer.

What is a good contribution margin?

It varies enough by category that cross-industry figures are close to useless. What matters is whether the margin clears your target CAC with room left for fixed costs and profit.

Should I use blended CAC or paid CAC?

Both. Blended CAC tracks whether the business is acquiring efficiently. Paid CAC tracks whether paid media is working. Reporting only blended CAC lets an underperforming paid channel hide behind organic growth.

What is the difference between ROAS and marginal ROAS?

ROAS is total conversion value divided by total spend, an average across everything you spent. Marginal ROAS is the return on the next increment of spend. Average ROAS is for reporting. Marginal ROAS is for deciding whether to scale.

How do I measure incremental ROAS?

With an experiment, not a dashboard. Geo holdouts, randomized conversion lift tests, and scheduled spend-down tests all estimate what would have happened without the ads. Media mix models are useful once they are calibrated against those experiments.

Are behavioral marketing tactics manipulative?

It depends on whether the thing you are pointing at is true. Making a real cost of inaction concrete, surfacing a genuine deadline, or recommending the option that fits most buyers all help people decide. Inventing scarcity, hiding alternatives, and inducing anxiety produce short-term lifts and long-term churn. The test is simple: would the customer still approve of the tactic if they could see exactly how it worked?

Start with one

Behavioral concepts tell you what to change. Unit economics tell you whether the change deserves more money. Run only the first half and you get clever campaigns that lose money. Run only the second half and you optimize spreadsheets for offers nobody wants.

Pick your current bottleneck from the three jobs above, choose the one concept that addresses it, and run a single clean test this month. Judge the result in contribution margin.