PLAYBOOK — FREE FROM EGGKNITE

The Lifecycle & Retention Playbook

The operating manual for owned-channel revenue: deliverability that earns the inbox, capture that compounds, the five core flows with real benchmark ranges, segmentation that protects sender reputation, a compliant SMS layer and the retention math that proves it all.

From the Lifecycle & Demand Generation toolset.

$36
average return per $1 spent on email
Litmus — the highest-ROI channel in marketing
41%
of email revenue comes from automated flows
from roughly 5% of sends — Klaviyo Benchmark Report 2025
0.3%
the spam-complaint ceiling Gmail and Yahoo enforce
Google sender requirements, in force since 2024
25–95%
profit lift from a 5-point retention gain
Bain & Company

Lifecycle marketing is the closest thing growth has to compound interest. Paid media buys you the first purchase; email and SMS decide whether that purchase becomes a customer relationship. The economics are lopsided in your favor: Litmus pegs email’s average return at $36 for every dollar spent, and Klaviyo’s 2025 benchmark data shows automated flows driving about 41% of email revenue from roughly 5% of total sends. A flow you build once keeps selling every day after.

The catch is that the bar for entry has risen. Since February 2024, Gmail and Yahoo require authenticated mail, one-click unsubscribe and a spam-complaint rate below 0.3% from bulk senders — and Google has escalated from temporary failures to outright rejections for mail that misses the bar. On the SMS side, the FCC’s April 2025 revocation rules and state-level quiet hours have made compliance a design constraint rather than a legal footnote. Programs built on batch-and-blast habits are being priced out of the inbox.

This playbook is the system we run on lifecycle engagements: six phases, in build order. Foundation before capture, capture before flows, flows before campaigns — because each phase multiplies the one before it. Every benchmark cited here is sourced, and every KPI band is one we hold real programs against.

PHASE 01

Deliverability foundation

Earn inbox placement before you scale volume. Authentication, complaint hygiene and list discipline are the multiplier on everything that follows — a flow that lands in spam has an open rate of zero.

Authenticate every sending domain

Publish SPF, DKIM and DMARC on each domain that sends commercial mail — this has been mandatory for senders of 5,000+ daily messages to Gmail and Yahoo since February 2024, and it is table stakes at any volume. At least one of SPF or DKIM must align with your visible From: domain for DMARC to pass. Start DMARC at p=none to collect reports, then move to quarantine and reject once legitimate mail streams are verified.

Field note: Run your domain through our free Email Deliverability Checker before touching DNS — it grades SPF, DKIM, DMARC, MX and BIMI against the current Gmail and Yahoo requirements with exact fixes.

Ship one-click unsubscribe and honor it fast

Promotional mail must carry RFC 8058 List-Unsubscribe-Post headers so recipients can leave with a single tap in the inbox UI, and opt-outs must be processed within two days. Most major ESPs inject the headers automatically — verify rather than assume, especially on custom SMTP or transactional-turned-promotional streams. A frictionless exit is cheaper than the spam complaint it prevents.

Operate below the complaint ceiling

Google’s published guidance: keep user-reported spam under 0.1%, and treat 0.3% as the line where mitigation eligibility disappears. Wire up Google Postmaster Tools and Yahoo’s complaint feedback loop on day one and review them weekly. Complaint spikes almost always trace to a specific segment, offer or acquisition source — find it and cut it.

Field note: Complaints are counted against delivered mail that recipients actually report. Aggressive sends to stale segments are the classic spike: the 180-day-inactive cohort you “reactivate” before a sale is usually the culprit.

Warm up new domains and IPs deliberately

Cold infrastructure with sudden volume is the textbook spam pattern. Ramp over four to six weeks, starting with your most engaged recent openers and clickers and roughly doubling volume at each step while watching bounce and complaint telemetry. Separate marketing and transactional mail onto different subdomains so a promotional misstep never takes order confirmations down with it.

Write a sunset policy and automate it

Define the point at which a subscriber is costing you reputation — commonly no clicks or site activity in 120–180 days — and route them through a short re-permission series before suppressing. Apple Mail Privacy Protection auto-fires opens for a large share of iOS users, so anchor engagement definitions on clicks, site visits and purchases. A smaller list that wants your mail outperforms a big one that tolerates it.

Field note: Suppress rather than delete. Suppressed profiles keep their history for analytics and can re-enter via a new purchase or site event — deletion throws away the data and the win-back trigger.
Where this phase fails
  • Leaving DMARC at p=none permanently — you get the reporting benefit while spoofers keep the domain’s reputation hostage.
  • Measuring engagement on opens in a post-MPP world, which quietly rots sunset logic and segment quality.
  • Blasting the full list from a new ESP or domain on day one and burning the reputation the warm-up was meant to build.
  • Mixing purchased or scraped addresses into the list — spam traps and complaint clusters follow within weeks.
Spam complaint rate <0.1%DMARC policy quarantine or rejectHard bounce rate <0.5%Domains authenticated 100%Sunset automation live
PHASE 02

Capture & zero-party data

Turn anonymous traffic into addressable profiles at a rate that outruns list churn — and collect the one or two data points that make every later message smarter.

Run a popup with a real offer

Klaviyo’s guidance is a submit rate of about 3% or higher for popups and flyouts; typical unoptimized ecommerce popups sit at 2–3%, while well-built forms with strong offers routinely double that. Trigger on intent — time on page, scroll depth or exit — rather than instantly on load, and keep the first screen to a single decision.

Field note: The offer matters more than the design. “10% off” against “free shipping” against a dollar amount against early access is the highest-leverage A/B test on the form — run it before you touch colors.

Go multi-step: email, then SMS, then one question

Ask for email on step one, present SMS opt-in on step two with its own TCPA-compliant consent language, and close with a single preference question. Each step is optional and the email is already banked, so added steps cost little and every completion enriches the profile. Two-step formats consistently outperform single walls of fields.

Collect zero-party data you will actually use

One or two questions — category interest, shopping-for-whom, skin type, dog or cat — captured at signup power welcome-flow branching and campaign segmentation from message one. Only ask what a flow or segment will consume within 30 days; every unused field is friction you paid for. Store answers as profile properties so they survive across tools.

Field note: Progressive profiling beats long forms: ask the second question inside the welcome flow or on the preference page, once the relationship exists.

Cover every capture surface

The popup is one door among many: footer embeds, checkout opt-in, post-purchase upsell pages, account creation, back-in-stock alerts, QR codes on packaging and in-store signage all feed the same machine. Tag every source so you can compare downstream revenue per subscriber by origin. Checkout-sourced subscribers usually monetize best; make that box impossible to miss.

Set a growth target against measured churn

Every list loses people monthly to unsubscribes, bounces and sunset suppression — measure that outflow and set the capture target above it, because a flat list is a shrinking revenue base once fatigue sets in. Report net list growth, never gross signups. When growth stalls, revisit offer strength and traffic mix before adding another popup.

Where this phase fails
  • Bribing for emails with an offer so deep that the list fills with coupon hunters who never pay full price.
  • Launching SMS capture with email-style consent copy — SMS requires prior express written consent with carrier and frequency disclosures.
  • Firing the popup instantly on page load, which tanks submit rate and can degrade Core Web Vitals and Google page experience.
  • Collecting preference data that no flow, segment or campaign ever reads.
Popup submit rate ≥3%Net list growth positive every monthSMS opt-in at capture step two liveSignup sources tagged 100%Zero-party question ≥1 per signup
PHASE 03

The core flows

Build the five automations that do the heavy lifting. Klaviyo’s 2025 data attributes about 41% of email revenue to flows from roughly 5% of sends — and Omnisend found cart, welcome and browse abandonment alone account for 87% of automated orders.

Welcome flow: your highest-leverage sequence

First-message open rates commonly clear 50% — the warmest audience you will ever mail. Klaviyo benchmarks put average welcome-flow revenue near $2.65 per recipient with the top decile above $21, a spread that says execution is the variable. Run three to five emails: deliver the offer immediately, then brand story, best sellers and social proof, with an offer-expiry reminder closing the arc.

Field note: Branch on the zero-party answer from capture. A “shopping for my dog” subscriber seeing dog products in email two is segmentation earning its keep on day one.

Abandoned cart & checkout: the revenue workhorse

Baymard’s meta-analysis puts average cart abandonment at 70.2%, which makes this flow the single largest recoverable pool in ecommerce. Klaviyo’s analysis of 143,000+ cart flows benchmarks a 50.5% open rate, 6.25% click rate, 3.33% placed-order rate and $3.65 revenue per recipient on average — with the top 10% reaching 13.3% clicks, 7.7% conversion and $28.89 RPR. Send three touches across 24–48 hours: a reminder within the hour, objection handling at ~24 hours, and scarcity or an incentive last.

Field note: Baymard finds unexpected costs are the top abandonment cause, cited by 48% of US shoppers — answer shipping and returns questions inside email two before reaching for a discount in email three.

Browse abandonment: catch intent one step earlier

Viewers who never carted are a colder but far larger audience, and the flow monetizes accordingly — Klaviyo pegs average revenue around $0.90–$2 per recipient depending on order value. Keep it to one or two lightweight emails showing the viewed product with adjacent recommendations, and cap frequency so repeat browsers are never stalked. This flow depends on identified traffic, so form fills and click-throughs directly expand its reach.

Post-purchase: the flow that builds the second order

Post-purchase messages earn the highest open rates of any flow — buyers actively want them — which makes the sequence your cheapest channel for reviews, UGC and education even though immediate revenue per send is modest. Sequence it: order reassurance, delivery follow-up, usage tips, review request timed after the product has been experienced, then a cross-sell matched to what they bought. Time the cross-sell to your median repurchase interval instead of a fixed 30 days.

Field note: Separate first-time and repeat buyers. First-timers get brand education and the review ask; repeat buyers get loyalty, referral and replenishment logic.

Winback: recover customers before they are gone for good

Trigger when a customer exceeds their expected repurchase window — typically 60–120 days past median — with two or three escalating touches: a nudge, a stronger reason, then a last-call offer. Klaviyo data puts average winback revenue near $0.84 per recipient for mid-AOV brands; per-send numbers look small while the margin is nearly pure, since the audience costs nothing new to reach. Feed non-responders straight into the sunset ladder from Phase 1.

Extend with transactional-adjacent triggers

Back-in-stock, price-drop, low-inventory and replenishment reminders are small flows with outsized conversion because the customer declared the intent themselves. Build them after the big five are live and tested. Each one is an afternoon of work that pays rent indefinitely.

Where this phase fails
  • Single-email flows — Klaviyo’s cart data shows three-message sequences dramatically out-earn one-and-done sends.
  • Leading every flow with a discount, which trains your best customers to abandon carts on purpose.
  • Ignoring flow-vs-flow suppression: a subscriber in the cart flow should exit browse abandonment immediately.
  • Setting flows live and never re-testing — the gap between average and top-decile RPR is roughly 8x, and it is made of iteration.
Flow share of email revenue 30–45%Cart flow RPR $3.65+ (top 10%: $28.89)Cart placed-order rate 3.3%+Welcome RPR $2.65+Core flows live 5 of 5
PHASE 04

Campaigns & segmentation

Flows print money on autopilot; campaigns set the tempo. Segmentation is what lets you send more to people who respond and less to people who complain — the mechanism that grows revenue and protects deliverability at the same time.

Build an RFM model

Score every customer on recency, frequency and monetary value — quintiles per axis — and collapse the grid into named tiers: champions, loyal, promising, at-risk, hibernating. Most ESPs can compute this natively or via computed properties; refresh scores on a schedule so the tiers stay honest. RFM turns “who should get this campaign?” from a debate into a lookup.

Field note: Start with five tiers at most. A 125-cell grid nobody actions is a dashboard; five tiers wired into cadence rules is a system.

Layer engagement tiers over the customer tiers

Bucket profiles by last email click or site visit: 30, 60, 90 and 180 days. Your most engaged tier can absorb three to five sends a week; the 90-plus tier should see one or two carefully chosen sends a month at most. This single discipline is the difference between a growing program and a complaint-rate incident.

Set cadence by tier and test upward carefully

Default the calendar to two or three campaigns per week to engaged segments, then test additional frequency against unsubscribe and complaint deltas rather than against revenue alone — revenue always rises with volume right up until reputation breaks. When a send must go wide, exclude the unengaged tiers by default. Omnisend’s 2025 dataset puts average campaign opens at 30.2%; falling well below your own baseline is the early-warning light.

Field note: Book a recurring monthly “list health” review: complaint rate, unsub rate, net growth and revenue per recipient by tier, on one page.

Plan a content mix beyond the discount

A calendar that alternates product launches, education, social proof, community and seasonal moments keeps engagement alive between promotions and gives the algorithmic inboxes positive signals to learn from. Aim for a majority of sends that would be worth opening with no coupon attached. Promotion-only calendars decay measurably within two quarters.

Make testing a standing habit

Every campaign carries one test: subject line, preview text, send time, offer framing or module order. Judge subject lines on clicks and revenue per recipient — MPP-inflated opens flatter the wrong variants. Log results centrally; a year of small compounding wins is where top-decile programs come from.

Where this phase fails
  • Sending every campaign to the full list — the fastest route from healthy metrics to a 0.3% complaint spike.
  • Chasing open-rate wins that Apple MPP fabricated while click and conversion rates stand still.
  • Running RFM once as a project instead of refreshing it as infrastructure.
  • Letting the promo calendar overrule engagement tiers during Q4, precisely when inbox providers tighten filtering.
Campaign open rate 30%+Campaign click rate 1.5%+Unsubscribe rate <0.3% per sendEngaged-tier share of sends ≥80%Tests logged per month 4+
PHASE 05

The SMS layer

SMS is the highest-attention channel you can own — and the most regulated. Built as a precision layer on top of email, it adds urgency where timing matters; built as a second blast channel, it adds legal exposure at $500–$1,500 per message.

Get consent architecture right first

US marketing texts require prior express written consent: clear disclosure of program, frequency, message-and-data rates and STOP/HELP language at the point of signup, with proof logged per subscriber. TCPA statutory damages run $500–$1,500 per message with no cap, and class actions climbed sharply through 2025 — consent records are the asset that makes the channel safe to scale. Keep SMS consent separate from email consent in both the form and the database.

Field note: Screenshot and version every consent-collecting form. When a demand letter arrives, the timestamped consent record and the form language it came from are your entire defense.

Comply with the 2025 revocation rules

Since April 11, 2025, the FCC requires honoring opt-outs made by any reasonable method — STOP and its synonyms, email, voicemail, web form or plain language — within ten business days, with real-time processing as the operating standard. You may send exactly one non-promotional clarification message, within five minutes, if the scope of the revocation is genuinely unclear. Audit that your platform catches misspelled and free-text opt-outs, because “please stop texting me” now counts.

Respect quiet hours everywhere you send

Federal quiet hours are 8am–9pm in the recipient’s local time, and several states go further — Florida, Oklahoma and Washington cut off at 8pm, and Texas restricts marketing texts to 9am–9pm Monday through Saturday and noon–9pm on Sundays under its 2025 rules. Configure quiet hours by recipient location, never by your office timezone. Queue rather than drop messages that land in a restricted window.

Pair SMS with flows before campaigns

Klaviyo’s benchmarks show flow-based SMS clicking near 10% on average with top performers past 16% — roughly double campaign SMS — because the message arrives at a moment the customer created. The proven pairings: a cart-abandonment text 30–60 minutes after the event ahead of the email sequence, back-in-stock alerts, shipping updates and welcome-offer delivery. Let email carry the narrative and let SMS carry the moment.

Field note: In cross-channel flows, check for email engagement before the text fires. Reserving SMS for messages email missed keeps per-message cost and opt-out rate down simultaneously.

Ration campaign SMS and track revenue per message

Four to eight campaign texts a month is the workable band for most brands; each send costs real money and a slice of goodwill, so every message needs urgency the inbox could never match — drops, restocks, final hours. Watch opt-out rate per send and revenue per message as the channel’s twin health metrics. SMS lists shrink faster than email lists when abused and forgive far less.

Where this phase fails
  • Migrating email habits to SMS — frequency tolerance is a fraction of email’s and opt-outs are permanent.
  • Relying on keyword-only opt-out handling after April 2025, when any reasonable revocation method must be honored.
  • Sending campaign texts on a national schedule that violates state quiet-hour windows in Florida, Texas or Washington.
  • Skipping the ROI math: at $0.01–0.03 per segment, an untargeted SMS blast can cost more than it returns.
Flow SMS click rate 8–10%+Opt-out rate per campaign <1%Opt-out processing real-timeQuiet-hours compliance 100%Campaign volume 4–8/month
PHASE 06

Retention economics

Prove the program in the language finance respects: repeat rate, cohort LTV and revenue per recipient. Bain’s research puts the prize in plain terms — a 5-point retention improvement lifts profits 25–95% — and this phase is how you claim credit for it.

Baseline your repeat purchase rate

Compiled Shopify-store benchmarks put the average ecommerce repeat purchase rate near 27–28%, with wide spread by category — consumables and subscriptions run far higher, considered purchases lower. Compute yours as customers with 2+ orders over trailing twelve months divided by all customers in the window, and track it monthly. This single number is the cleanest referendum on whether lifecycle is working.

Field note: Segment repeat rate by first product purchased. Most brands find one or two “gateway” products whose buyers return at twice the average — that insight should reshape both paid acquisition and the post-purchase cross-sell.

Build cohort LTV curves

Group customers by acquisition month and plot cumulative revenue per customer at 30, 60, 90, 180 and 365 days. Cohort curves reveal what blended averages hide: whether recent cohorts are strengthening, when the repurchase cliff hits, and how much a customer is truly worth against CAC. Every flow-timing decision in Phase 3 — cross-sell delay, winback trigger — should be read off these curves.

Make revenue per recipient the working metric

RPR normalizes performance across list sizes, flows and campaigns — it is the number Klaviyo builds its benchmarks on and the fairest way to compare a 3,000-person segment against a 300,000-person blast. Review RPR by flow and by campaign segment quarterly against the Phase 3 benchmark bands. When RPR falls while volume rises, you are strip-mining the list; the fix is segmentation rather than another send.

Track owned-channel share of revenue

Klaviyo’s 2025 benchmark data shows established ecommerce brands typically attributing 25–40% of total store revenue to email and SMS. Below that band, the usual culprits are missing flows, weak capture or deliverability problems — walk phases 1 through 3 in order. Well above it, the brand may be leaning on the list to compensate for stalled acquisition, which caps growth from the other direction.

Field note: Use a consistent attribution window — Klaviyo’s default is 5 days for email — and sanity-check platform-attributed revenue against a holdout or against order coupon codes once a quarter.

Report retention like a P&L line

One monthly dashboard: repeat rate, net list growth, flow and campaign RPR, owned-channel revenue share, complaint rate and cohort LTV at 90 days. Annotate it with what shipped — new flows, offer tests, cadence changes — so causation stays legible. Existing customers spend meaningfully more per order than first-timers as the relationship matures; the dashboard is how that compounding becomes budget next planning cycle.

Where this phase fails
  • Reporting opens and clicks to leadership when the durable claim is repeat rate, LTV and owned-revenue share.
  • Letting the ESP’s generous default attribution inflate the story until finance audits it for you.
  • Averaging LTV across all cohorts, which hides the decay or improvement that recent changes caused.
  • Treating retention as the lifecycle team’s job alone while product, CX and shipping experience set the ceiling.
Owned-channel revenue share 25–40%Repeat purchase rate 27%+90-day cohort LTV tracked monthlyRPR review cadence quarterlyRetention dashboard live
MODELS

Frameworks to steal

The 0.1 / 0.3 operating bands

Run complaint rate like an SRE runs error budgets. Under 0.1%: normal operations, room to test frequency and reactivation. Between 0.1% and 0.3%: incident mode — freeze list expansion, cut the unengaged tiers from sends, find the offending segment or source. At 0.3%: Gmail’s mitigation eligibility is gone and recovery takes weeks of exemplary behavior. The band you are in dictates this week’s send plan.

The 41/5 build order

Klaviyo’s benchmark finding — roughly 41% of email revenue from about 5% of sends — is an instruction, read correctly: automation earns a wildly disproportionate share, so build in coverage order. Welcome, cart, browse, post-purchase and winback come before the campaign calendar gets ambitious, and each flow gets three-plus messages before any flow gets a redesign. Only after flows cover the big five moments does incremental effort shift to campaigns.

The RFM 5×5×5 grid

Score recency, frequency and monetary value into quintiles, then collapse the 125 cells into five actionable tiers: champions (5-5-5 neighborhood) get early access and referral asks; loyal get cross-sell and reviews; promising get nurture toward order two; at-risk get the winback flow before the discount; hibernating get one re-permission attempt, then sunset. The grid is rebuilt monthly; the tier definitions stay stable so trends mean something.

The master checklist

  • SPF, DKIM and DMARC pass and align on every sending domain
  • DMARC upgraded from p=none to quarantine or reject
  • One-click unsubscribe (RFC 8058) live on all promotional mail, honored within two days
  • Google Postmaster Tools and Yahoo feedback loop connected; complaints reviewed weekly
  • Spam complaint rate holding under 0.1%
  • Marketing and transactional mail separated onto distinct subdomains
  • New domains and IPs warmed over 4–6 weeks, engaged segments first
  • Sunset policy live: re-permission series, then automatic suppression
  • Popup running at a 3%+ submit rate with a tested offer
  • Multi-step form captures email, SMS consent and one zero-party question
  • Every capture source tagged and feeding the welcome flow immediately
  • Welcome flow: 3–5 emails, offer delivered up front, branched on preference data
  • Abandoned cart flow: 3 touches across 24–48 hours, objections answered before discounts
  • Browse abandonment flow live with frequency caps and cart-flow suppression
  • Post-purchase flow splits first-time and repeat buyers; review ask timed to delivery
  • Winback triggers past the median repurchase window and feeds the sunset ladder
  • RFM tiers built, refreshed monthly and wired into campaign cadence
  • Engagement tiers gate frequency; unengaged excluded from wide sends by default
  • Every campaign ships with one logged test
  • SMS consent is prior-express-written, logged per subscriber with form versioning
  • Opt-outs honored by any reasonable method, processed in real time
  • Quiet hours enforced by recipient-local time, including stricter state windows
  • SMS paired into cart, back-in-stock and shipping flows before campaign use
  • Monthly retention dashboard: repeat rate, cohort LTV, RPR, owned-revenue share, complaint rate

Frequently asked questions

Do the Gmail and Yahoo rules apply if I send fewer than 5,000 emails a day?
The strict requirements formally target senders of 5,000+ daily messages to those providers, but the filtering systems reward the same signals at every volume — and most growing lists cross the threshold during their first big promotion, when a compliance scramble is most expensive. Authenticate, add one-click unsubscribe and watch complaints from day one; the setup takes hours and pays back indefinitely.
How many campaigns per week should I send?
Cadence is a property of the segment rather than the brand. Engaged tiers at most ecommerce brands comfortably absorb two to five sends a week; 90-day inactives should see one or two a month at most. Increase frequency one step at a time on the engaged tier and watch unsubscribe and complaint deltas for two weeks before locking it in — revenue alone will always vote for more volume, right up until deliverability breaks.
What share of total revenue should email and SMS drive?
Klaviyo’s 2025 benchmark data shows established ecommerce brands typically attributing 25–40% of store revenue to owned channels. Meaningfully below that usually means missing flows, weak capture or deliverability trouble. Far above it can signal over-reliance on the existing list while acquisition stalls — healthy programs grow both sides.
Are open rates still worth tracking after Apple Mail Privacy Protection?
As a directional signal within one audience over time, yes; as a success metric or a segmentation input, no longer. MPP auto-fires opens for a large share of Apple Mail users, inflating rates and corrupting open-based logic. Anchor engagement definitions, sunset policies and test verdicts on clicks, site activity and revenue per recipient — the numbers MPP is unable to fake.
Should the welcome offer be a discount?
A discount is the highest-converting default, and it carries a real training cost if it becomes the brand’s only lever. Test alternatives with real equity — free shipping thresholds, gift-with-purchase, early access, content or a loyalty-point head start — and reserve the deepest offers for the last email of the cart flow rather than the first touch of the relationship. Margin-friendly offers that hold conversion within a point of the discount are common findings.
How quickly do flows pay back?
Faster than almost anything else in growth. Cart and welcome flows start producing within days of launch because the triggering behavior already exists at volume; the build cost is typically days of work per flow. The compounding part takes longer — the roughly 8x gap between average and top-decile revenue per recipient in Klaviyo’s data is closed through months of iterating offers, timing and creative on a live baseline.
Is SMS worth adding if the email program is already strong?
Yes, as a precision layer — a strong email program is exactly the prerequisite. Flow-based texts click near 10% on Klaviyo’s benchmarks because they arrive at moments the customer created: an abandoned cart, a restock, a shipped order. Start with consent-clean capture and two or three flow pairings, prove revenue per message, and let campaign SMS stay scarce enough to keep its urgency.