CRO Playbook

A practitioner-level conversion rate optimization playbook covering the systematic process for identifying, testing, and implementing changes that increase conversion rates across websites, landing pages, and digital products.

---

1. The Full CRO Process

Step-by-Step Framework

CRO is not random A/B testing. It is a systematic, evidence-based process that moves from data collection to hypothesis to testing to implementation.

Step 1: Define the Conversion Goal

• What specific action do you want more users to take?
• Examples: form submission, account signup, purchase, demo request, download
• Choose ONE primary conversion goal per page or flow
• Secondary goals (micro-conversions) can be tracked but should not split your focus

Step 2: Pick the Device and Screen

• Analyze your traffic split by device (desktop, mobile, tablet)
• Start optimization on the device that represents the majority of your traffic
• Mobile and desktop conversion paths often differ significantly -- optimize separately
• Check if there is a device-specific drop-off that indicates a UX problem on one platform

Step 3: Google Analytics Flow Analysis

• Map the user journey through your conversion funnel using GA's funnel visualization
• Identify the step with the highest drop-off rate -- this is your biggest opportunity
• Compare funnel performance by traffic source (organic vs paid vs direct vs social)
• Look for unexpected exit pages that indicate confusion or friction

Step 4: Session Recording Analysis

• Use tools like Smartlook, Mouseflow, or Hotjar to watch actual user sessions
• Watch 50-100 sessions of users who abandoned the conversion flow
• Note patterns: where do they hesitate? What do they try to click that is not clickable? Where do they scroll and stop?
• Pay special attention to rage clicks (rapid repeated clicking on the same element) -- these indicate frustration with a non-responsive element or confusing UI

Step 5: Heatmap Analysis

• Scroll heatmaps: How far down the page do users actually scroll? If your CTA is below the fold and 60% of users never scroll to it, you have found a problem.
• Click heatmaps: Where do users click? Are they clicking non-interactive elements (indicating confusion)? Are they ignoring your primary CTA?
• Attention heatmaps: Where do users spend the most time looking/hovering?

Step 6: Form Your Hypothesis Based on the data collected, form a specific, testable hypothesis:

• "Changing [element] from [current state] to [proposed state] will increase [metric] by [expected amount] because [reasoning based on data]."
• Example: "Moving the CTA button above the fold will increase demo requests by 15% because scroll heatmaps show 45% of users never reach the current CTA position."

Step 7: A/B Test

• Create the variant (change only ONE element per test)
• Split traffic evenly (50/50) between control and variant
• Use a testing tool (Google Optimize successor, VWO, Optimizely, or a custom solution)
• Do not peek at results before reaching statistical significance
• Minimum test duration: 2 full business weeks (to capture day-of-week patterns)
• Minimum sample: Use a sample size calculator -- typically need 1,000+ conversions per variant for reliable results

Step 8: Measure for Statistical Significance

• Require 95% confidence level before declaring a winner
• If the test is inconclusive after adequate sample size, the difference is likely too small to matter -- move on to a higher-impact hypothesis
• Do not end tests early because of early positive results (regression to the mean is real)

Step 9: Implement Permanent Changes

• Winners get permanently implemented in the codebase (not left running as tests)
• Document the test results, hypothesis, and learning for future reference
• Move to the next highest-priority hypothesis

---

2. Exit Popup Strategies

When and How to Use Exit Popups

Exit-intent popups trigger when the user's cursor moves toward the browser's close button or address bar (desktop only -- mobile requires scroll-up detection or time-based triggers).

Effective exit popup types:

The value offer:

• Offer a discount, free resource, or bonus that was not visible on the page
• Example: "Wait -- get 15% off your first order" with a countdown timer
• Works best for ecommerce and SaaS free trial pages

The objection handler:

• Address the most common objection to conversion
• Example: "Not sure if it's right for you? Chat with an expert for free."
• Works best for high-consideration purchases

The content lead magnet:

• Offer related downloadable content in exchange for an email
• Example: "Before you go -- download the complete guide to [topic]"
• Works best for content pages and blog posts

The survey:

• Ask why they are leaving (2-3 quick options)
• Example: "What stopped you from signing up today?" with radio buttons
• Works best for gathering data to inform future optimization

Exit popup rules:

• Trigger only once per session (multiple popups destroy trust)
• Include a clear, obvious close button
• Do not show to returning visitors who have already dismissed it
• Mobile alternative: use a sticky bottom bar instead of a popup (popups on mobile are penalized by Google and annoying to users)
• Test the popup against no popup -- some audiences respond negatively, and the popup can increase bounce rate

---

3. Scroll Popup Timing

The 30% Scroll Trigger

Scroll-triggered popups fire when a user has scrolled a specified percentage of the page. The optimal trigger point is around 30% scroll depth.

Why 30%:

• At 30%, the user has scrolled past the introduction and is actively reading the content
• They are an engaged reader, not a bouncer
• Triggering just before the halfway point catches users while interest is high but before they start losing attention
• Users who reach 30% scroll depth are 3-4x more likely to convert than those who do not scroll at all

Implementation:

• Use JavaScript scroll event listeners to detect scroll percentage
• Show a slide-in from the bottom-right or a non-intrusive bar rather than a full-screen overlay
• Content should be relevant to the page they are reading (not a generic newsletter signup)
• Example: On a blog post about email marketing, the 30% scroll popup offers a downloadable email marketing template

Performance benchmarks:

• Scroll popups at 30% typically convert at 3-5% of viewers
• This is 2-3x higher than time-based popups (which often fire for disengaged users)
• And 1.5-2x higher than entry popups (which fire before the user has any investment)

---

4. The Retargeting Cookie Delay Hack

Only Retarget Interested Users

By default, retargeting pixels fire immediately when a page loads. This means everyone who visits your site -- including accidental clicks, bot traffic, and immediate bouncers -- enters your retargeting audience.

The fix: Hide the retargeting pixel for 45 seconds using JavaScript.

Implementation:

// Delay retargeting pixel by 45 seconds
setTimeout(function() {
  // Insert your retargeting pixel code here
  // Facebook pixel, Google remarketing tag, LinkedIn Insight tag, etc.
}, 45000);

Why 45 seconds:

• Users who stay on your site for 45+ seconds have demonstrated genuine interest
• They have read content, explored navigation, or engaged with the page
• This filters out accidental visitors, bot traffic, and tire-kickers
• Your retargeting audience becomes smaller but dramatically more qualified

Impact:

• Retargeting audience size decreases by 40-60% (most visitors leave within 30 seconds)
• But retargeting CTR increases by 50-100% because you are only retargeting interested users
• Net effect: lower retargeting spend with higher conversion rates and better ROI
• CPL from retargeting campaigns drops significantly because you eliminate waste impressions

---

5. Localized Pricing Experiments

Geographic Price Sensitivity

Different markets have different willingness to pay. Testing localized pricing can unlock significant revenue gains.

Implementation approach:

• Use IP geolocation to detect the visitor's country
• Display pricing in local currency at market-appropriate price points
• Higher-GDP countries may support premium pricing
• Emerging markets often need lower price points for viable conversion rates
• Test systematically: start with 3-4 country tiers

Country tiering example:

Tier	Regions	Pricing Strategy
Tier 1	US, UK, Canada, Australia, Nordics	Full price
Tier 2	Western Europe, Japan, Singapore	85-95% of full price
Tier 3	Eastern Europe, Latin America, SE Asia	50-70% of full price
Tier 4	India, Africa, select developing markets	30-50% of full price

Purchasing Power Parity (PPP) pricing: Some companies openly offer PPP-adjusted pricing with a banner: "We've adjusted pricing for your region." This builds goodwill and significantly increases conversion rates in lower-income markets without cannibalizing Tier 1 revenue (most Tier 1 customers never see the lower prices).

---

6. The Unsubscribe Page Retargeting Technique

Turning Unsubscribes into Pipeline

When someone clicks "unsubscribe" in your email, they land on an unsubscribe confirmation page. Most companies treat this as a dead end. It is actually an opportunity.

The technique:

1. Create a custom redirect page that serves as the unsubscribe confirmation
2. On this page, alongside the "You've been unsubscribed" confirmation, present a compelling offer or landing page
3. This could be: a different product, a downgraded offering (e.g., "Instead of weekly emails, get our monthly digest"), a one-time discount, or a piece of high-value content
4. Place a retargeting pixel on this page
5. Build a custom audience of unsubscribe page visitors
6. Run retargeting ads to this audience with a re-engagement offer

Why this works:

• Unsubscribers are not necessarily disinterested -- they may have email fatigue, changed roles, or prefer other channels
• The retargeting ads reach them on social media and display, offering a different engagement path
• Even a small conversion rate from this audience is pure incremental revenue from people you would have otherwise lost completely
• This turns a loss (unsubscribe) into a pipeline opportunity

Implementation details:

• The redirect must still actually process the unsubscribe (do not trap people)
• The landing page offer should feel like a value exchange, not a guilt trip
• Test different offers: content download, webinar invite, discount, alternative subscription frequency

---

7. Web Personalization with CRM Data

Dynamic Page Content Based on Visitor Identity

When you can identify a website visitor (through login, email link click, or cookie match), personalize their experience using data from your CRM.

Personalization layers:

By industry:

• Show industry-specific hero images, case studies, and testimonials
• Adjust language to use industry-specific terminology
• Surface industry-relevant features or use cases

By company size:

• Show pricing appropriate to their segment (startup vs enterprise)
• Feature case studies from similarly-sized companies
• Adjust CTA language (small: "Start free trial" vs enterprise: "Contact sales")

By funnel stage:

• New visitor: Educational content, thought leadership, top-of-funnel offers
• Returning visitor: Case studies, product tours, comparison pages
• Known lead: Personalized demo, ROI calculator pre-filled with their data
• Existing customer: Upsell offers, feature announcements, support resources

By engagement history:

• Show content related to topics they have previously engaged with
• Hide content they have already consumed
• Surface next-logical-step content based on their consumption pattern

---

8. Pricing Page Experiments

Testing Displayed vs. Actual Price

An advanced pricing experiment: test the difference in conversion rate between the displayed price and the actual billing price.

Concept:

• Display a price that anchors perception (e.g., "$99/month")
• After the user initiates checkout, the actual price may differ (annual billing, introductory rate, etc.)
• The experiment measures whether different displayed prices affect signup rates even when the actual billing amount is the same

Practical application:

• Test showing monthly price vs. annual price divided by 12
• Test showing price per user vs. total team price
• Test showing price before vs. after discount
• Test "Starting at $X" vs. exact pricing

Example: Showing "$8.25/month (billed annually at $99)" vs. "$99/year" -- same price, different framing. The monthly breakdown often converts better because the per-period cost feels smaller.

---

9. CTA Optimization

Color and Design Psychology

Bright colors for CTAs, dark colors for opt-outs:

• Primary CTA buttons should use a bright, contrasting color that stands out from the page palette
• Common high-performing CTA colors: green, orange, blue (test for your specific audience)
• Opt-out or secondary options should use muted, less prominent styling (gray text, no button, smaller font)
• The visual hierarchy should make the desired action the obvious choice

CTA Copy That Converts

Speak like your customers, not like a marketer:

Generic (Lower CTR)	Customer Language (Higher CTR)
Subscribe to our newsletter	Get the email course
Sign up for updates	Send me the tips
Request a demo	See it in action
Contact sales	Talk to an expert
Learn more	Show me how it works
Download whitepaper	Get the free guide

The distinction: "Subscribe for the email course" sounds like a commitment. "Get the email course" sounds like a benefit. Use language that emphasizes what the user receives, not what they are signing up for.

Social Pressure Around CTAs

Add social proof directly adjacent to CTA buttons:

• "Join 10,000+ marketers who read this weekly"
• "N of your connections also use this product" (requires social login data)
• "237 people signed up today"
• "Rated 4.8/5 by 1,200 users"

Social pressure near the CTA reduces decision anxiety at the exact moment of conversion. It answers the implicit question: "Am I making a mistake?" with evidence that others have made the same choice.

Using social pressure from the user's network: If you have social login data, showing that specific friends or connections use your product is extremely powerful. "5 of your friends are also using this product" creates peer-validated social proof that generic numbers cannot match.

---

10. Gate Popups on Link Clicks

Inline Content Gating

Instead of gating entire pages behind forms, gate specific high-value elements within the content.

Implementation:

• Display the first section of a guide, report, or template freely
• When the user clicks a link to download, access additional sections, or view detailed data, trigger a popup requesting an email
• The user has already consumed enough content to understand the value, making them more willing to exchange their email

Why this outperforms full-page gates:

• The user self-selects by clicking -- they want the specific resource
• They have already experienced value from the ungated portion
• The ask feels proportional to the value offered
• SEO benefit: the ungated content is indexable, driving organic traffic to the page

Testing approach:

• Test gate position: after 30%, 50%, or 70% of content
• Test gate format: inline form vs. popup vs. slide-in
• Test what you ask for: email only vs. email + name vs. email + company
• More fields = lower conversion rate but higher lead quality

---

11. Signup Flow Optimization

Capture Signups Before Full Details

For businesses with multi-step conversion flows (booking, application, checkout), capture the email address as early as possible in the flow.

The strategy:

1. Step 1: Collect email address and basic info (name, email)
2. Step 2: Collect detailed information (booking details, preferences, payment)
3. Many users abandon at Step 2 -- but you now have their email from Step 1

The follow-up:

• Send an automated email to users who completed Step 1 but not Step 2
• "You started booking but didn't finish -- here's your link to continue"
• Typically 90% of users who complete Step 1 do not complete Step 2
• Even if only 1% of those 90% complete the booking from the reminder email, that represents roughly a 10% lift in total completed bookings -- significant revenue at zero additional acquisition cost

Implementation:

• Save the email on Step 1 form submission (before redirecting to Step 2)
• Set up an automated email trigger that fires 1 hour, 24 hours, and 72 hours after abandonment
• Include a direct link that returns them to Step 2 with their information pre-filled
• Personalize the email with whatever information they provided in Step 1

---

12. Visual Demonstrations of Savings

Why Visuals Outperform Text

When communicating savings, discounts, or value comparisons, visual demonstrations dramatically outperform text-based claims.

Example test result: Replacing a text-based savings claim ("Save $200 by inviting friends") with a visual demonstration (animated calculator showing the savings accumulate as friends are added) increased invite rates by several multiples.

Why visual savings work:

• Text claims are abstract -- "Save 30%" is a cognitive exercise
• Visual demonstrations make the savings concrete and emotionally tangible
• Animations create a dopamine response as the user watches savings grow
• Interactive calculators let users personalize the savings to their situation

Visual savings formats:

Format	Use Case	Expected Impact
Before/after price comparison	Discounts, upgrades	2-3x lift
Animated savings calculator	Referral programs, volume pricing	3-7x lift
Side-by-side plan comparison	Tier upgrades	1.5-2x lift
Running total savings counter	Subscription vs. one-time	2-4x lift
Personalized ROI calculator	B2B, high-consideration purchases	2-3x lift

Implementation tips:

• Keep the visualization simple -- it should be understood in under 5 seconds
• Use green for savings, red for costs (universal color associations)
• Show the outcome, not just the percentage (people understand "$150 saved" better than "15% off")
• If possible, make it interactive (let the user adjust variables and see results change)

---

13. Device Restriction as a Conversion Lever

The Synced Device Upgrade

For products with multi-device usage (streaming, software, file storage), restricting the number of synced devices on the base plan creates a natural upgrade trigger.

The technique:

1. Set the free/base plan to support 1-2 devices
2. When a user attempts to connect a third device, show a clear upgrade prompt
3. Offer a discounted upgrade (not full price) to reduce friction at the moment of need
4. The upgrade is positioned as solving an immediate problem they are actively experiencing

Why this works:

• The user is experiencing the limitation in real-time (not hypothetically)
• The upgrade solves a problem they have right now
• The discounted price feels like a concession, increasing conversion likelihood
• Users who connect multiple devices are typically your highest-engagement segment, making them the most likely to convert and the highest-LTV customers

---

14. A/B Testing Best Practices

Traffic Distribution

• Always split traffic evenly (50/50 for two variants, 33/33/33 for three)
• Do not run tests with uneven splits (90/10) unless you are running a bandit algorithm and understand the tradeoffs
• Use server-side splitting when possible to avoid flicker (the flash of the original page before the variant loads)
• Ensure randomization is session-persistent (users should see the same variant throughout their session and on return visits)

Common A/B Testing Mistakes

Ending tests too early:

• Minimum 2 full business weeks regardless of traffic volume
• Must reach 95% statistical significance
• Early results are unreliable due to novelty effects and sample bias

Testing too many things at once:

• Change one variable per test
• If you change the headline, button color, and layout simultaneously, you cannot attribute the result to any specific change
• Multivariate testing exists but requires significantly more traffic

Not accounting for seasonality:

• Run tests for full week cycles (Monday-Sunday) to capture all day-of-week patterns
• Avoid starting tests during holidays, events, or promotional periods
• If a test spans a seasonal change, extend it to capture both periods

Ignoring segment-level results:

• A test might show no overall winner but have a clear winner for mobile users
• Always check results by: device type, traffic source, new vs returning visitor, and geography
• Segment-level insights can inform targeted implementations even when overall results are inconclusive

What to Test (Priority Order)

1. Headlines and value propositions (highest impact, easiest to test)
2. CTA text and placement (high impact)
3. Form length and fields (high impact for lead gen)
4. Social proof placement and content (medium-high impact)
5. Page layout and visual hierarchy (medium impact)
6. Images and video (medium impact, slower to test)
7. Color and design elements (lower impact than most people expect)
8. Micro-copy and help text (lower impact but easy wins)

---

15. Measurement and Reporting

CRO Dashboard Metrics

Track these metrics for every page in your conversion funnel:

Metric	Frequency	Alert Threshold
Conversion rate	Daily	Drop greater than 20% week-over-week
Bounce rate	Daily	Increase greater than 15%
Average session duration	Weekly	Drop greater than 25%
Funnel step completion	Daily	Any step dropping greater than 10%
Form abandonment rate	Weekly	Increase greater than 10%
Revenue per visitor	Weekly	Drop greater than 15%
CTA click rate	Daily	Drop greater than 20%

Calculating Test Impact

Revenue impact formula:

• Additional conversions per month = (Winning variant conversion rate - Control conversion rate) x Monthly visitors
• Monthly revenue impact = Additional conversions x Average order value (or lead value)
• Annual impact = Monthly revenue impact x 12

Example:

• Control conversion rate: 3.0%
• Variant conversion rate: 3.6% (20% relative improvement)
• Monthly visitors: 50,000
• Additional conversions: (0.036 - 0.030) x 50,000 = 300
• Average order value: $100
• Monthly revenue impact: 300 x $100 = $30,000
• Annual impact: $360,000

This kind of quantification helps prioritize future tests and secure resources for CRO programs.

---

16. Conversion Optimization Deep Cuts

These are the techniques that do not fit neatly into standard CRO frameworks but consistently produce measurable lifts when implemented thoughtfully. They sit at the intersection of value communication, device-specific behavior, lead capture psychology, and offline-to-online attribution.

Proposition Mapping Framework

Most conversion rate problems are not design problems — they are communication problems. You know your product is valuable, but the page fails to articulate that value in a way that resonates with the visitor's specific situation. The proposition mapping framework forces you to break your value proposition into six concrete categories and load each with proof points.

The six value categories:

Category	Question It Answers	Example
Time	What time does this save me?	"Cut onboarding from 3 weeks to 3 days"
Cost	What does it cost or save me?	"Reduce your cost per lead by 40%"
Quality	How good is the outcome?	"98.7% uptime over the last 24 months"
Pain	What frustration does it eliminate?	"No more manual CSV exports and copy-paste reporting"
Access	What can I do that I could not before?	"Get enterprise-level analytics without the enterprise price tag"
Status	What does using this say about me?	"Join 500+ growth leaders who switched this quarter"

How to use the matrix:

1. List every proof point, data point, testimonial, and feature benefit your product offers
2. Assign each to one or more of the six categories
3. Identify which categories are overloaded (you probably have too many Quality points) and which are empty (most companies neglect Pain and Status)
4. Fill the gaps — interview customers, review support tickets, and mine sales call recordings for the specific language people use around time savings, frustration relief, and status benefits

The payoff for conversion optimization: This matrix makes it trivial to generate landing page headline variants, ad copy angles, and email subject lines. Instead of brainstorming from scratch, you pick a category, grab a proof point, and write. Need a headline for a price-sensitive audience? Pull from Cost. Running a retargeting campaign to people who visited but did not convert? Lead with Pain — remind them of the frustration they are living with. Testing new ad copy? Mix categories to find which resonance angle produces the highest CTR for each audience segment.

Testing CTAs by Device

Mobile and desktop users are not the same audience, and they do not respond to the same calls to action. A CTA that performs well on desktop may actively hurt your mobile conversion rate.

Why device-specific CTA testing matters:

• Desktop users are typically in research mode with more time and screen space. They tolerate longer CTAs with more detail: "Start Your 14-Day Free Trial — No Credit Card Required."
• Mobile users are action-oriented and impatient. They want short, tappable CTAs with minimal text: "Try Free" or "Get Started."
• The thumb zone on mobile devices means button placement matters more than copy. A perfectly worded CTA that requires awkward stretching to tap loses to a mediocre CTA in the natural thumb position.

How to test:

1. Set up device-specific variants in your A/B testing tool
2. Run separate tests for desktop and mobile — do not combine results
3. Test both copy and placement simultaneously on mobile (tap target size and position are as important as words)
4. Allow 3-4 weeks per test to account for different usage patterns by device and day of week

Email Capture Tactics

Getting a visitor's email address is the conversion that funds every subsequent conversion. Each capture mechanism works through different psychology and fits different moments in the visitor journey.

Exit-intent popups:

• Trigger when the visitor's cursor movement signals they are about to leave (desktop) or after a scroll-up pattern (mobile approximation)
• The value exchange must be immediate and specific — not "join our newsletter" but "get the 12-point checklist" or "see the data behind our claims"
• Show only once per session. Multiple popups destroy trust and trigger ad blocker escalation.
• Benchmark: 2-5% conversion rate on exit-intent popups

Persistent top bars:

• A narrow, non-intrusive bar fixed to the top of every page with a single compelling stat or social proof statement and a small email input
• Works because it is always visible without being disruptive — it sits in peripheral vision and catches attention at natural scroll pauses
• Best content for the bar: a quantified benefit ("Join 12,000 marketers getting weekly growth tactics") or a specific free resource
• Benchmark: 0.5-2% conversion rate, but it compounds because every page view is an impression

Full-page welcome mats on second visit:

• Do NOT show on the first visit — let the visitor experience your site and build interest first
• On return visits, display a full-screen overlay with a high-value offer that acknowledges they are back: "Welcome back — this time, grab our complete playbook"
• The return visit signals genuine interest, which means the visitor is more receptive to a bolder ask
• Benchmark: 5-15% conversion rate on second-visit welcome mats (much higher than first-visit mats)

Testing across tactics: Run all three simultaneously but measure each independently. Different tactics capture different visitor segments — exit-intent catches people on their way out, top bars catch browsers, and welcome mats catch engaged returners. The combined system captures more emails than any single mechanism.

Tracking Offline Conversions Back to Web Sources

If your final conversion happens offline — a phone call, a text message, an in-person meeting, a mailed check — you lose the connection between marketing spend and revenue unless you build a deliberate bridge between online and offline.

The tracking architecture:

1. Assign unique identifiers per marketing channel and campaign. This can be a unique phone number (dynamic number insertion), a unique URL parameter, a unique promo code, or a unique landing page per campaign.
2. Capture the visitor's web identifier at the point of initial contact. When a visitor fills out a form, starts a chat, or creates an account, store their analytics visitor ID (the client ID from your analytics platform) alongside their contact information in your CRM.
3. Track subsequent offline events. When that contact later calls, visits a location, receives a recommendation, or makes a purchase, log the event in your CRM tied to their contact record.
4. Send offline events back to your analytics platform. Use your analytics platform's measurement protocol or data import feature to push offline conversion events back into the web analytics system. Include the original visitor ID so the offline conversion is attributed to the correct web session, traffic source, and campaign.

What this unlocks:

• You can see which Google Ads keywords produce phone calls that result in sales — not just form fills
• You can attribute in-store purchases to the Facebook campaign that drove the initial website visit
• You can calculate true end-to-end ROI for every marketing channel, including channels where the last touchpoint is a human conversation rather than a checkout page
• Your automated bidding strategies in ad platforms become dramatically more effective because they can optimize for actual revenue events, not just proxy conversions

Implementation priority: Start with your highest-volume offline conversion path. If most of your revenue comes from phone calls, implement call tracking with dynamic number insertion first. If most revenue comes from in-person consultations that start as web form fills, build the CRM-to-analytics bridge first. Solve the biggest attribution gap before expanding to secondary channels.

---

NPS-Based Automation Branching

Most companies collect NPS scores and look at them in a dashboard. Very few wire NPS directly into their automation to trigger different actions based on the score. This is a missed opportunity.

The pattern: after a key customer milestone (first major interaction, onboarding completion, first renewal period, or a significant event), trigger an NPS survey. Set a delay of 2-3 weeks to allow responses to accumulate. Then branch the automation based on the score.

Low NPS path: trigger a webhook to your CRM that creates a task for a customer success rep or account manager to personally follow up. This is not a marketing email. It is a human reaching out to understand the issue. The webhook should include the NPS score, any written feedback, and the customer's account details so the rep has full context before making the call.

High NPS path combined with a high-value signal (like account size, engagement level, or usage frequency): trigger an upsell or expansion offer. These are your happiest, most engaged customers. They are the most receptive to a premium tier, an add-on product, or an expanded commitment. The offer should feel like a reward, not a sales pitch.

Both paths merge back into a shared check-in email with a buffer delay (7-14 days). This ensures every customer gets a touchpoint regardless of which branch they went through, and it gives the rep follow-up time to complete before the next automated message.

This turns a passive survey into an active revenue and retention mechanism. Low NPS becomes an early warning system that routes problems to humans before they become churn. High NPS becomes a qualified upsell trigger that reaches customers at the moment they are most satisfied. The automation handles the routing. The humans handle the conversations that matter.