YESWARD

THE PILLAR

Considered purchases: why buyers hesitate, and what to do about it.

A considered purchase is one people research before they commit. The decision takes days, not minutes; it involves comparison, configuration and often another person; and hesitation is the normal state of the journey rather than a defect in it.

DEFINITION

A considered purchase sits roughly in the €300–€3,000 band, involves comparison and configuration before checkout, and is bought once every few years rather than weekly. The buyer is neither browsing nor certain. They are doing careful work, and they can be lost at any step of it without ever telling you why.

This page is the argument Yesward is built on: that these journeys are the ones where restraint has commercial value, because the buyer's attention is already committed to the decision and interrupting it costs more than it does on an impulse purchase.

WHY THE DISTINCTION MATTERS

Tactics that work on a €40 basket are actively harmful on a €1,800 decision.

Impulse / low considerationConsidered purchase
Decision windowMinutes, one sessionDays to weeks, several sessions and devices
What abandonment meansMild interest that fadedA specific unanswered question, usually a knowable one
Effect of urgency tacticsOften works — the stakes are low enough to gambleReads as pressure on a decision the buyer is being careful about
Effect of a discountTips a marginal basket over the lineOften pays people who would have bought anyway, and rarely answers the real question
Value of one recoveryCents of marginEnough that one recovered contract can pay for months of platform
Right defaultAsk for the saleGet out of the way, and answer only what is actually blocking them

This is why we do not think of Yesward as an ecommerce conversion tool with a considered-purchase mode. The economics of restraint only work when a single recovered decision is worth real money — which is exactly this band.

THE JOURNEY

Four steps. Hesitation clusters at the seams.

comparison → configuration → price → checkout / application

STEP 01 · COMPARISON

Which of these, and why

Three tabs open, a spec table half understood, and no basis on which to separate two options that both look fine.

Dominant friction: choice

SIGNALS: A↔B LOOPS · COMPARE PAGE RETURNS

STEP 02 · CONFIGURATION

Will it actually work for me

Dimensions, dates, address, eligibility, compatibility. The step where a buyer discovers a constraint nobody warned them about.

Dominant friction: coverage

SIGNALS: REPEAT CHECKS · CHANGED INPUTS

STEP 03 · PRICE

What will this really cost

Not “is it too expensive” but “what is the total, over what period, with what added”. Usually a clarity problem wearing a price costume.

Dominant friction: price, then commitment

SIGNALS: FEE FAQ · CALCULATOR RE-RUNS

STEP 04 · CHECKOUT / APPLICATION

Am I sure about committing

Contract length, cancellation, what happens if it goes wrong, whether this company can be trusted with a direct debit.

Dominant friction: commitment, trust

SIGNALS: TERMS PAGE · TERM TOGGLES · STALL

Analytics can tell you which step people leave from. It cannot tell you which of the four questions above they were stuck on — and the answer determines whether the right response is a sentence, a question, or nothing at all.

THE VOCABULARY

Five frictions. Each one has a different cheapest answer.

Full definitions in the methodology →

Choice friction

Two or more options look equally valid and nothing on the page separates them.

Cheapest answer: the one distinction that actually decides it — usually not price.

TYPICAL LEVEL · L1–L2

Price friction

The total cost is unclear, or the shape of it over time is unclear.

Cheapest answer: information. A discount is a last resort you switch on deliberately.

TYPICAL LEVEL · L1–L3

Commitment friction

The term, the lock-in, the exit. Very common in subscription and service purchases.

Cheapest answer: one question about the buyer's horizon, then the matching term.

TYPICAL LEVEL · L3 · THE WORKED EXAMPLE

Trust friction

Doubt about the company, the promise, or what happens when something goes wrong.

Cheapest answer: the specific promise in writing. Invented social proof makes it worse.

TYPICAL LEVEL · L1

Coverage friction

Eligibility, availability, dates, address, compatibility — a hard constraint, not a doubt.

Cheapest answer: often none. If the answer is no, the honest move is to say so early.

TYPICAL LEVEL · L0 OR L5

Note what is missing from this list: “not interested”, “just browsing”, “price too high”. Those are conclusions, not frictions — nothing on your website can resolve them, so the correct response is silence, and a receipt recording that we decided so.

WHERE THIS SHOWS UP

Four verticals where hesitation is the whole game.

Chosen for structural fit rather than sector glamour: comparison behaviour, configuration steps, real acquisition spend, and margin data that exists server-side.

01 · HOME & FURNITURE

Will it fit, and will it last

High order value, visual products, specification and material questions, delivery windows, financing. Comparison behaviour is intense and mostly invisible in analytics.

DOMINANT: CHOICE · COVERAGE

Case-study shape: reduce product comparison abandonment without discounting.

02 · TRAVEL

Dates, rooms, refunds

€500–€3,000 bookings, multiple travellers, cancellation conditions, room and date selection, and enormous acquisition spend behind every session.

DOMINANT: COMMITMENT · CHOICE

Case-study shape: reduce booking hesitation at the date and refund step.

03 · INSURANCE & CONSUMER FINANCE

Am I eligible, and what is excluded

Friction around cost, eligibility, commitment and complexity — the strongest product fit of the four. The trade-off is a slower sales cycle into regulated organisations.

DOMINANT: TRUST · COVERAGE · PRICE

Case-study shape: reduce quote abandonment mid-application.

04 · ENERGY, TELECOM & MOBILITY

Which tariff, for how long

Tariffs, contract terms, hardware, installation, financing and comparison sites. Our first beachhead, and the journey the demo account is modelled on.

DOMINANT: COMMITMENT · PRICE

Case-study shape: reduce tariff comparison abandonment. See the test site ↗

THE RESPONSE

Match the smallest response to the friction you can actually evidence.

Every row below is a decision the engine makes thousands of times a day. The right-hand column is the part most tools skip: what we deliberately do not do, and why.

The Intervention Ladder in full →
Evidence saysSmallest useful responseWhat we do not do
Weak or ambiguous signalNothing. L0, with a receipt.Guess, and call the guess personalisation
Choice frictionThe one distinction that separates the options, in page flowRecommend the highest-margin option and call it help
Price frictionTotal cost made explicit, over the period the buyer is thinking inOpen with a discount. Off by default, policy-gated
Commitment frictionOne question about their horizon, then the term that matchesOpen a chat and hope a conversation happens
Trust frictionThe specific promise, in writing, where the doubt occurredBadges, counters, and “2 people are viewing this”
Coverage friction, answer is noSay so early and stop spending their timeCapture the lead anyway
Buyer already progressingNothing. Especially nothing.Congratulate ourselves on 96% personalisation coverage

MEASUREMENT

Long journeys make attribution easy to fool.

A considered purchase spans sessions and devices, which means a large share of buyers convert some time after any intervention you made. Last-touch logic will happily credit your nudge with sales it had nothing to do with.

The only defence is a control group that persists as long as the programme does — and a reported unit that cannot be inflated by giving margin away.

The measurement maths → · Three ways NBA results get overstated →

FOUR METRICS WORTH THE ARGUMENT

Incremental gross profit

Treated minus permanent holdout, net of discount, model spend and fees. The headline number, with its interval.

Qualified outcome rate

Share of eligible sessions reaching a genuinely valuable outcome — not clicks, not conversations started.

Cost of speaking

Dismissals, rage clicks, bounce delta and suppressed sessions, reported next to the lift rather than in an appendix.

Restraint rate

Share of sessions deliberately left alone. A number most vendors would rather you did not ask for.

RESEARCH · IN PROGRESS

The Considered Purchase Friction Report

Every deployment observes friction. Aggregated anonymously across accounts, that becomes something no competitor can write from a keyword tool: which behavioural patterns indicate which friction, which responses resolve it, how often silence wins, and what each treatment was worth.

Not published yet, and we will not pretend otherwise. The first edition needs enough sessions across enough accounts to say anything defensible.

PLANNED CUTS

by friction typeby journey stagemobile vs desktopnew vs returningby industrywhich responses resolve ithow often silence wins

Founding pilots contribute anonymised aggregates and get the first edition before it is public.

How a founding pilot works →

FREQUENTLY ASKED

Considered purchases, briefly.

What is a considered purchase?

A purchase people research before committing — typically €300–€3,000, involving comparison and configuration, bought rarely enough that the buyer has no habit to fall back on.

Why do buyers hesitate on them?

Because a specific question is unanswered: which option, will it work here, what is the total cost, what am I committing to, can I trust this company. Hesitation is rational, not a lack of interest.

Is a discount the fastest fix?

It is the fastest way to raise conversion rate and often the fastest way to lower profit, because most of it goes to people who would have bought anyway. It rarely answers the actual question.

How is this different from cart abandonment?

Cart abandonment is one symptom at one step. On considered purchases most loss happens earlier — during comparison and configuration — where there is no cart to recover and no email to send.

Does this work for B2B?

Where the purchase completes online at this value, yes. Enterprise deals with a sales cycle and a committee are a different problem, and we would rather say so than stretch the claim.

Where should I start?

One journey, one outcome, two weeks in shadow mode. You get the friction distribution for your own traffic before anything is shown to a visitor. The pilot →

Get the friction distribution for your own journey.

Two weeks in shadow mode tells you which of the five frictions your buyers actually hit, and how often the answer is nothing.

Review one leaking journey