RB
ROIBASE.

LIFECYCLE AUTOMATION

CDP & Retention Engineering

Identity graph, RFM + LTV + churn scoring, 12-18 lifecycle flows, channel orchestration and control-group experimentation — a full operation that builds retention as an engineering discipline, not a 'campaign'.

Talk strategy
Contact

Acquisition is expensive, retention isn't negotiable; but every company that runs retention as 'monthly newsletters' is silently re-paying its CAC.

Retention's paradox: most brands treat it as a marketing activity, when successful retention is actually a product engineering discipline — identity graph, RFM/LTV/churn scoring, lifecycle automation, experimental design and channel orchestration. Roibase's retention engineering is built on six principles; every principle produces 'fewer but more triggering actions' instead of 'more emails'.

Roibase perspective

METHODOLOGY

DIAGNOSE to ARCHITECT to AUTOMATE to EXPERIMENT to OPTIMIZE to GOVERN — retention engineering

A six-stage process; every stage produces a written decision, a test and a transferable output.

01

01

DIAGNOSE

Retention curve, cohort decay, segment health, channel performance, deliverability + consent audit; the quantitative visibility of lost revenue.

02

02

ARCHITECT

Identity graph, segment strategy, KPI tree, lifecycle flow map and experiment design; stakeholder approval and 'north-star metric' are set.

03

03

AUTOMATE

CDP + ESP + SMS + push + in-app integration; 12-18 flows go live; every flow starts with a control group.

04

04

EXPERIMENT

Weekly experiments with statistical power calculations; incremental lift + interaction effects are measured; winners go 'always-on', losers are killed.

05

05

OPTIMIZE

Segment score updates, channel mix optimization, creative/offer refresh, frequency capping improvements; on a quarterly cadence.

06

06

GOVERN

Consent granularity, right-to-erase pipeline, vendor DPAs, deliverability monitoring and SLA + on-call rotation handed over in writing.

— COMPARISON

ESP-only vs all-in-one CDP vs Roibase retention engineering

How three approaches define retention and what that produces operationally.

DimensionESP-only (Klaviyo/Mailchimp)All-in-one CDP vendorRoibase retention engineering
Identity resolutionEmail-onlyVendor black-box modelDeterministic + probabilistic, open model
Predictive scoringNone or simple LTVVendor default scoresCustom churn + LTV + NBA model
Lifecycle flow count4-6 core flows8-12 templates12-18 custom + control-grouped
Channel orchestrationEmail onlyEmail + SMS + pushEmail + SMS + push + in-app + paid + web
Experiment disciplineAd-hoc A/BVendor A/B toolControl group + power analysis + incremental lift
Governance + consentBasic opt-outVendor DPA + consent signalRunbook + granular consent + RTE pipeline
Warehouse + reverse ETLNoneLimited connectorsWarehouse-native, vendor-agnostic
Total annual cost40-90k€ (ESP + SMS)180-400k€ (CDP license heavy)110-240k€ (operation + warehouse)

PROOF

Outcomes, measured

+27%
Repeat rate improvement

Average in the first 6 months after winback + activation flows.

4x
Email revenue

After Klaviyo migration + segment rewrite + deliverability repair.

-18%
Churn rate

Over 12 months with predictive churn score + proactive outreach + offer testing.

+38%
D60 LTV

Segment average after lifecycle flow count grew from 5 to 14.

99.2%
Inbox delivery

After SPF/DKIM/DMARC + BIMI + ISP reputation monitoring.

14
Lifecycle flows

Typical number of live control-grouped flows in an implementation.

WHAT WE DO

Engagement scope

Every offering is an outcome-based work package. Roibase blends strategy and execution inside a single team — no hand-offs.

01 / 10

Identity resolution

Deterministic (email/phone/user_id) + probabilistic (fingerprint, household) matching; a single customer graph + cross-device journey.

02 / 10

RFM + LTV scoring

Every customer gets predictive LTV + recency/frequency/monetary tier + churn probability; segments are built on these scores.

03 / 10

Churn probability model

Gradient boosting + behavioral features produce 30/60/90-day churn probability; high-risk segments trigger proactive outreach.

04 / 10

Lifecycle journey mapping

Welcome, activation, repeat, retention, winback, VIP, sleep, reactivation — 12-18 flows; each with a control-group A/B and statistical power calculation.

05 / 10

Channel orchestration

Email + SMS + push + in-app + web + paid retargeting under a single frequency cap + channel priority rules; the user is reached without fatigue.

06 / 10

Experimentation platform

Control group + power analysis + Bayesian or sequential testing for every automation; nothing scales without measured incremental lift.

07 / 10

Next-best-action engine

The next best action per customer (product recommendation, channel, offer); pushed to activation channels via reverse ETL.

08 / 10

Deliverability + consent ops

SPF/DKIM/DMARC setup, BIMI logo, ISP reputation monitoring, consent granularity and right-to-erase pipeline.

09 / 10

CDP platform

Segment, RudderStack, mParticle, Bloomreach, Klaviyo Data Platform or warehouse-native CDP (Census/Hightouch) — vendor-agnostic setup based on need.

10 / 10

Retention KPI tree

Every metric's source, owner, threshold and triggered action is written down; one tree from D30 repeat rate to LTV/CAC.

— BENEFIT

The concrete, quantitative return of retention engineering

The financial, operational and risk impact of moving from ESP to an engineering discipline.

+34% LTV/CAC

CAC pressure drops

Higher LTV + repeat rate makes the same CAC deliver much better payback; paid budget can flow freely into profitable segments.

4x email revenue

Email becomes a revenue engine

After deliverability repair + segment rewrite + lifecycle flows, email grows to 25-40% of total revenue.

-18% churn

Churn becomes predictable

Predictive scores surface risk 30-60 days early; proactive winback + VIP retention produce real incremental lift.

-47% unsub

Channel fatigue ends

With frequency capping + channel priority order, users aren't burned out; unsubscribe and spam complaint rates fall.

100% audit-ready

Consent + compliance assurance

Granular consent + right-to-erase + vendor DPAs make a KVKK/GDPR audit easy to defend.

+41% capacity

Your team's capacity opens up

Automation + NBA engine + self-serve analytics move the marketing team from producing campaigns to producing strategy + experiments.

DELIVERABLES

Concrete, written deliverables for every retention project

Architecture, code, configuration, playbook and training — all handed over to your team.

  • Retention diagnostic report

    Cohort decay, segment health, channel performance, deliverability + consent audit; 40-60 pages.

  • Identity graph + data model

    Deterministic + probabilistic matching rules, household detection, unified customer view schema.

  • RFM/LTV/churn model pack

    Trained models + feature registry + retraining pipeline + model card documentation.

  • Lifecycle playbook

    Visual map of 12-18 flows, each with target metric, control design, creative brief and measurement plan.

  • CDP + ESP + SMS setup

    Live Segment/Klaviyo/Braze/mParticle setup, event taxonomy, data contracts + monitoring.

  • Orchestration engine

    Channel priority rules, frequency capping, suppression rules, holdout groups.

  • Deliverability setup

    SPF/DKIM/DMARC + BIMI logo + ISP warm-up plan + reputation monitoring dashboard.

  • Experimentation framework

    Power analysis template, Bayesian/sequential testing tools, experiment log + decision matrix.

  • KPI tree + dashboard

    D30/D60/D90 repeat rate, LTV, churn, engagement score; metrics tied to business decisions.

  • Consent + RTE pipeline

    Granular consent tables, right-to-erase automation, vendor DPA library, monthly audit report.

  • NBA engine + reverse ETL

    Next-best-action generation + push to activation channels via Census/Hightouch + monitoring.

  • Runbook + 3-week training

    Operational runbook, on-call rotation, SLA contract + 3 weeks of hands-on training for your team.

— SCOPE

What we do, what we don't — clear boundaries

The retention engineering scope is in writing; it prevents scope surprises and hidden invoices.

We do

  • Retention diagnostic + cohort + deliverability audit
  • Identity graph + data model design
  • RFM + LTV + churn predictive model training
  • 12-18 lifecycle flow implementation
  • Channel orchestration + frequency capping
  • Control-group experimentation framework
  • CDP + ESP + SMS + push + in-app setup
  • Deliverability repair (SPF/DKIM/DMARC/BIMI)
  • Consent granularity + right-to-erase pipeline
  • NBA engine + reverse ETL activation
  • KPI tree + dashboard + alerting
  • Runbook + 3-week hands-on training

We don't

  • Legal counsel (DPA/consent review coordinated via partner lawyer)
  • Volumetric creative production (separate scope with content/design teams)
  • ESP/SMS/CDP license resale (vendor-agnostic recommendations, no commission)
  • CRM sales functions (B2B sales engagement is a different scope)
  • Loyalty program mechanics (a separate decision with business model + finance)
  • Ad account management (separate scope with PPC teams)
  • Manual monthly campaign production (scope is 'engineering', not 'campaign team')
  • Customer support + helpdesk operations (coordinated with your CX teams)

HOW WE WORK

First live flow in 4 weeks, full system in 16 weeks — written cadence

01

Weeks 1-2: diagnostic

Cohort decay, segment health, deliverability + consent audit; stakeholder interviews + north-star metric.

02

Weeks 3-4: architect + quick-win

Identity graph + KPI tree + first lifecycle flows (welcome + browse-abandon) go live; first demo.

03

Weeks 5-6: CDP + channel setup

Segment/mParticle/Klaviyo integration, SMS + push + in-app channel connection, event taxonomy.

04

Weeks 7-8: predictive scoring

RFM + LTV + churn models are trained, feature registry is set up, scores start feeding segment creation.

05

Weeks 9-10: lifecycle expansion

Activation, repeat, retention, VIP, winback flows go live with control groups; the experimentation framework opens.

06

Weeks 11-12: orchestration + deliverability

Frequency capping + channel priority rules + SPF/DKIM/DMARC + BIMI; suppression + holdout configuration.

07

Weeks 13-14: NBA + reverse ETL

Next-best-action engine goes live; Meta CA, Google CM, Klaviyo segments pushed via Census/Hightouch.

08

Weeks 15-16: govern + handoff

Consent + RTE pipeline + monthly audit report + runbook; 3 weeks of hands-on training, SLA + on-call rotation handed over.

— TOOLKIT

The tools we use — vendor-agnostic but decisive choices

We pick what fits each customer; we protect independence by taking no commissions.

CDP & IDENTITY

SegmentRudderStackmParticleBloomreachKlaviyo Data PlatformHightouchCensus

EMAIL & MESSAGING

KlaviyoBrazeIterableCustomer.ioHubSpotAttentivePostscriptOneSignal

EXPERIMENTATION & MODELING

StatsigEppoGrowthBookOptimizelyPython (scikit-learn, lightgbm)dbtBigQuery ML

DELIVERABILITY & GOVERNANCE

Postmark ReputationGlockAppsGoogle PostmasterOneTrustCookiebotDidomi

QUESTIONS

Frequently asked

An integrated stack like Shopify + Klaviyo is enough under 100k monthly customers. For 500k+ customers, multi-channel, multi-brand or B2B accounts, a CDP is recommended as a separate layer. But a 'warehouse-native CDP' (BigQuery/Snowflake + Census/Hightouch) is often more economical + flexible than a traditional CDP. The choice gets made through a vendor-agnostic audit.

— GLOSSARY

CDP & Retention Engineering terminology

Twelve critical terms that give your team and stakeholders a shared language.

01
Identity Resolution
Linking activity across devices/channels to a single user identity; deterministic (email/user_id) + probabilistic (fingerprint/household).
CDPHousehold
02
RFM
Recency, Frequency, Monetary — the classic 3 dimensions of customer value; the foundation of scoring in modern retention.
LTVChurn Probability
03
LTV
Lifetime Value; a customer's total lifetime net contribution. Predictive LTV estimates it within 30/60/90 days.
RFMCAC
04
Churn Probability
The probability that a user will leave the active base in the next N days — computed by an ML model (logistic regression, gradient boosting, survival analysis) on RFM, product-usage, payment and engagement signals. The basis for prioritising win-back and VIP communication.
RFMNBA
05
Cohort Decay
The decline over time in activity or revenue contribution of a cohort of users acquired in a given period. In healthy products the curve flattens at some point (plateau); a cohort that decays without bottoming out is the most reliable signal of missing product-market fit.
Retention CurveLTV
06
Lifecycle Journey
The staged customer journey like welcome → activation → repeat → retention → winback; different messaging and offers per stage.
NBAOrchestration
07
Next-Best-Action (NBA)
The engine that recommends the best real-time action (product, offer, channel, message) for each customer; a recommender + rule engine combination.
Reverse ETLOrchestration
08
Orchestration
The coordination of cross-channel message flow under frequency capping + priority rules + suppression.
Frequency CappingNBA
09
Frequency Capping
The upper limit of messages sent to a user within a time window; prevents channel fatigue.
OrchestrationDeliverability
10
Deliverability
The rate at which messages actually reach the inbox; managed via SPF/DKIM/DMARC/BIMI + reputation monitoring.
Frequency CappingGovernance
11
Reverse ETL
Pushing segments/scores from the warehouse to activation channels like Klaviyo, Meta, Google, Braze.
CDPNBA
12
Incremental Lift
The extra revenue/behavior an automation produces compared to a control group; the real success measure of retention experiments.
ExperimentationChurn Probability
13
RFM Segmentation
Scoring customers on Recency (last purchase date), Frequency (cadence) and Monetary (total spend) and bucketing them into behaviour-based cohorts. The canonical engine for win-back, VIP, dormant lifecycle signals; foundational CRM/CDP module.
14
Win-back Campaign
A communication flow targeting customers who have not purchased for a defined period (typically the RFM "dormant" segment). Reactivates with a personalised discount, a "missed you" message or a new-collection hook; success measured by reactivation rate and incremental LTV.
15
Upsell
Selling an existing customer a more expensive, more capable version of the product/plan they bought. SaaS basic → pro → enterprise, e-commerce "128GB instead of 32GB"; the fastest way to grow LTV, leveraging existing trust.
16
Cross-sell
Recommending a complementary/related product to an existing customer. A case for the phone they bought, related titles for the book, an add-on for the SaaS plan; the "Frequently Bought Together" widget in e-commerce runs on this — grows AOV.
17
NPS (Net Promoter Score)
A customer-satisfaction metric from the answer to "On a 0-10 scale, how likely are you to recommend us to a friend?" 9-10 promoters, 7-8 passives, 0-6 detractors; NPS = %promoters - %detractors. The real value is industry benchmark + trend over time.
18
CSAT (Customer Satisfaction)
A point-in-time satisfaction score asked after a specific interaction ("how satisfied were you with the resolution of this support ticket?") on a 1-5 or 1-7 scale. Unlike NPS it is episodic and transactional; the iteration signal for product teams.
19
Drip Campaign
A sequence of emails delivered automatically on a predefined time cadence. Typically 5-9 emails covering lead nurturing, onboarding or education flows. Each message follows the "one topic, one CTA" rule; layered with behavioural triggers it becomes a true lifecycle programme.
20
Welcome Series
A 3-5 email welcome sequence sent to newly registered users. The first email often clears 50% open rate — the most valuable window to set tone, USPs and product benefits. A welcome series in the first 14 days lifts LTV by 25-40%.
21
Abandoned Cart Email
An email triggered after a user adds items to the cart but leaves without checking out. Typical sequence: 1-hour reminder, 24-hour incentive, 72-hour last chance. A well-tuned flow recovers 10-15% of otherwise lost revenue.
22
Triggered Email (Behavioural)
An email fired in real time by a user behaviour event — login, purchase, profile update, dwell time. Engagement is 4-7× that of schedule-based campaigns; requires a solid CDP / event-tracking foundation.
23
Transactional Email
A critical-information email sent in response to a user action — order confirmation, receipt, password reset, OTP. Marketing-law constraints (CAN-SPAM, opt-in rules) don't apply; deliverability practice recommends a separate IP pool from marketing.
24
ESP (Email Service Provider)
A SaaS platform that handles bulk-email sending. Mailchimp, Klaviyo, Braze, Iterable and Customer.io are examples; provides bounce/complaint handling, IP reputation, segmentation, A/B testing and deliverability monitoring. The technical foundation of any company's LTV/retention engine.
25
Email Deliverability
The ability of sent emails to land in the inbox rather than the spam folder. Driven by SPF/DKIM/DMARC + IP reputation + content quality + engagement signals + list hygiene. Bulk senders target 95%+ inbox placement; under 85% is a serious red flag.
26
Sender Reputation / Sender Score
A 0-100 trust score that ISPs assign to a sender. Computed from spam-complaint rate, bounce rate, engagement and blacklist status; SenderScore.org, Google Postmaster Tools and Microsoft SNDS are the public tools. 80+ is ideal, under 50 means imminent blacklist risk.
27
Soft Bounce vs Hard Bounce
Hard bounce: a permanent delivery failure (invalid address, missing domain) — must be removed from the list immediately. Soft bounce: a temporary failure (mailbox full, server down) — treated as permanent after 3-5 retries. Bounce rate above 2% drags down sender reputation.
28
IP Warm-up
The process of slowly ramping up sending volume from a new or dormant IP until ISPs deem it trustworthy. Typical plan: day 1 = 50, day 2 = 100, day 3 = 200… day 30 = full volume. Skip it and you get ISP throttling and spam-folder placement.
29
Apple MPP (Mail Privacy Protection)
A feature shipped in iOS 15 (2021) that pre-fetches the open pixel through a proxy whenever an Apple Mail user opens an email. It inflates open rates and hides location. ESPs now segment Apple Mail opens to preserve true engagement metrics.
30
Subject Line A/B Testing
The practice of sending two subject lines for the same campaign to a small holdout, then sending the winner to the rest. A built-in feature in most ESPs; 5-10% sample sizes with a 4-hour wait are typical. The highest-ROI email test, lifting open rates 15-30%.
31
Preheader / Preview Text
The short snippet that appears under the subject line in the inbox preview. Reinforces the subject or offers a different hook; a well-written preheader can lift open rate 20-40%. Aim for 80-150 characters; the first 35 matter most on mobile.
32
CTOR (Click-To-Open Rate)
The share of openers who click at least once (clicks / opens). Independent of the subject line, it measures content quality; the industry average is 10-15%. Combined with open rate it answers the question "is it the subject or the content that's broken?".
33
List Hygiene
The practice of regularly removing undeliverable, unengaged or complaint-prone addresses from an email list. Sunsetting 90-day inactives via a re-engagement flow and auto-deleting hard bounces is the standard. A dirty 100K list performs far worse than a clean 50K one.
34
Double Opt-in
A consent method that requires the user, after the signup form, to click a confirmation email landed in their own inbox before joining the list. Lifts list quality dramatically and lowers complaint rate; legally mandatory in some markets (Germany, Austria).
35
Personalization Token
A placeholder in an email body that dynamically injects user-specific data — name, last order, recommended product. Syntax like {{first_name}} or %FIRST_NAME%; ESP merge tags are standard. Used well, tokens lift open and click rates by 15-25%; the "Hi {{first_name}}" cliché is no longer enough.
36
Send-time Optimization (STO)
A feature that uses machine learning to send each user an email at their personally most-active time. ESPs derive each subscriber's ideal hour from past-engagement patterns; open rates rise 5-15% versus single-time blasts.
37
AMP for Email
A technology Google introduced in 2019 that lets you embed interactive elements (forms, carousels, real-time content) inside an email. Users can answer surveys, pick products or book a meeting without leaving the inbox. Supported by Gmail, Yahoo and Mail.ru; sent as an extra text/x-amp-html MIME part.
38
SMS / WhatsApp Marketing
A retention channel that complements or replaces email via SMS and the WhatsApp Business API. Open rate exceeds 95%, but cost and consent are strict. WhatsApp requires template messages plus customer-service windows; SMS imposes a 160-character limit and a mandatory STOP keyword.
39
Web Push Notification
A browser-delivered notification sent via a Service Worker that appears on desktop or mobile even when the site is closed. With Apple Safari iOS 16.4+ support it became a true channel; opt-in rates are 5-25% and click rates 5-15%. Its CTR competes with native push.
40
CSAT (Customer Satisfaction Score)
A single-question customer-satisfaction metric — "How satisfied were you with this experience?" Asked on a 1-5 scale; 4s and 5s count as positive. Measures a specific touchpoint (shipping, call, product); the short-term satisfaction counterpart to NPS's long-term loyalty focus.
41
CES (Customer Effort Score)
A metric that asks "How hard was it to solve your issue?" — measuring customer effort. Scored 1-7; the 2010 HBR study proved that low effort correlates with high loyalty more strongly than satisfaction. Where CSAT asks "are you happy?", CES asks "did we make it easy?".
42
tNPS / rNPS (Transactional / Relational NPS)
Two modes of NPS: relational NPS asks about overall brand loyalty quarterly or annually; transactional NPS fires right after a specific touchpoint (order, call, invoice). A low tNPS pinpoints which process is broken; rNPS is the macro-health indicator.
43
FCR (First Contact Resolution)
The rate at which a customer's issue is resolved on their first support contact. Targets are typically 70-80%; every repeat contact doubles cost (AHT × 2) and drags down CSAT. Driven up by AI agent assist, KB optimisation and tier skipping; the strongest indicator of support efficiency.
44
AHT (Average Handle Time)
The average duration of a support interaction — talk time + hold time + after-call work. The core efficiency metric in contact centres; low AHT is good only when it doesn't hurt FCR or CSAT. AI summarisation and real-time agent-assist tools are the modern levers to bring it down.
45
ASA (Average Speed of Answer)
How long, on average, a customer waits in queue on a support channel. The classic call-centre SLA is "80% answered within 20 seconds"; abandonment grows exponentially once it crosses 60 seconds. Workforce management plus chatbot deflection are the two levers to keep it healthy.
46
Deflection Rate
The share of customer requests resolved by self-service (KB article, chatbot, IVR) without ever reaching a live agent. Modern support targets 30-60%; every 10 points of deflection saves seven figures in FTE cost. AI agents like Intercom Fin and Zendesk AI now push verticals past 50%.
47
Knowledge Base (KB)
A searchable library of self-service articles and videos written for both customers and agents. Search analytics, content-gap analysis and disciplined version updates keep it alive; it underpins FCR and deflection metrics. Zendesk Guide, Help Scout Docs, Notion and Document360 are common platforms.
48
Omnichannel CX
A support experience in which context never breaks as the customer moves between web, app, store, phone, email, WhatsApp and social. Requires a single customer ID, a unified conversation history and shared agent screens. The evolution of multichannel and the modern support standard.
49
Voice of Customer (VoC)
A programme that systematically collects and analyses customer feedback. Sources: surveys (NPS, CSAT, CES), support tickets, social-media comments, app-store reviews and agent notes. Modern VoC stack: Qualtrics, Medallia, ChurnZero plus LLM-based topic clustering.
50
Customer Journey Map
A document that visualises the customer's journey from awareness → consideration → decision → onboarding → retention → advocacy. Columns capture touchpoints, emotions, pain points and opportunities; the output surfaces CX improvement priorities. The front-stage half of a service blueprint.
51
Service Blueprint
An extended diagram that exposes everything behind the customer journey map: customer actions (front-stage), agent actions (back-stage) and the supporting systems and processes. Surfaces the true root cause of CX bugs; the central tool of service design.
52
Escalation Matrix
A table that defines which ticket escalates when, to whom and at what priority. Tier 1 → Tier 2 → engineering, with severity (P1/P2/P3) and elapsed-time thresholds. Modern helpdesks automate it; a well-designed matrix cuts mean-time-to-resolve by 30-50%.
53
Tier 1 / Tier 2 / Tier 3 Support
The tiered structure of a support organisation: Tier 1 is front-line, scripted and handles 70-80% of tickets; Tier 2 is specialist for complex troubleshooting; Tier 3 is engineer-level for code fixes. Solid tier-skipping rules, AI agent assist and runbooks balance FCR against cost.
54
ITIL (IT Infrastructure Library)
A best-practice framework for IT service management (from Axelos), built on five pillars: Service Strategy, Design, Transition, Operation and Continual Improvement. Standardises Incident, Problem, Change and Release Management; the philosophical foundation of ServiceNow and Jira Service Management.
55
ITSM (IT Service Management)
The discipline of designing, delivering and managing IT capability as "services". Operationalised through the ITIL framework, it bundles ticket, change, asset and knowledge management. ServiceNow, BMC Helix, Jira Service Management and Freshservice are leading platforms.
56
Incident vs Request vs Problem (ITIL)
ITIL's three ticket types: Incident — "it's broken right now, restore it" (urgent); Service Request — "give me a new laptop / VPN access" (planned); Problem — "these incidents keep recurring, find the root cause" (long-running). Conflating the three is the most common helpdesk-SLA bug.
57
Zendesk
The most widely deployed cloud helpdesk and CX platform (founded 2007, Copenhagen). Email, chat, voice, social and WhatsApp omnichannel; Zendesk AI delivers auto-suggestion and answer-bot deflection. 100K+ customers — the "Salesforce of CX" for modern support operations.
58
Intercom
Founded in 2011, a support and marketing platform focused on web/mobile in-app messaging. Live chat plus automated workflows plus Intercom Fin (AI agent) drive high deflection. Among SaaS companies "in-app chat" is usually Intercom; integrations with Klaus and Productboard are strong.
59
Freshdesk
Freshworks's affordable helpdesk product. Multi-channel ticketing, a KB, basic automations and AI summarisation; the SMB and mid-market alternative to Zendesk. Combined with Freshchat and Freshcaller it forms an omnichannel experience under one roof.
60
AI Agent Assist
A system that surfaces AI suggestions to a live agent during a conversation — similar resolutions from past tickets, KB articles, customer context, sentiment alerts and suggested responses. Modern Zendesk AI, Intercom Fin Copilot and Salesforce Service Cloud Einstein are the canonical examples.
61
Customer Health Score
A composite score that predicts a customer's churn risk or growth potential. Inputs: usage frequency, last login, NPS, CSAT, support-ticket trend and payment timing. The main dial Customer Success teams use for proactive intervention; ChurnZero, Gainsight and Vitally automate it.
62
Self-Service Containment
The share of sessions that start in self-service and complete without ever bouncing to a live channel. The IVR/chatbot-focused measurement of deflection — "of those who entered the AI bot, what percent transferred to an agent?". The most direct KPI for tracking the ROI of CX investments.
63
CSAT Driver Analysis
A statistical analysis that pinpoints which sub-factors (response speed, agent warmth, resolution clarity) drive a CSAT score. Run via logistic regression, SHAP or topic modelling; tells you where low CSAT is coming from and which investment buys the most score uplift.
64
Topic Modeling (CX)
An NLP technique that automatically clusters recurring themes across thousands of support tickets or reviews. LDA is the classic; BERTopic and GPT-clustering are the modern picks. The data-driven answer to "what are our customers actually complaining about?" — a goldmine for the product roadmap and KB content gaps.
65
ATS (Applicant Tracking System)
A hiring tool that captures applications, parses CVs and lets recruiters track candidates through a pipeline. Greenhouse, Lever, Workable, Ashby, BambooHR and Workday Recruiting are common; LinkedIn Recruiter and Indeed integrations are standard. The CRM of the modern hiring funnel.
66
HRIS (Human Resources Information System)
A system that holds employee master data — personal info, contract, compensation, leave, performance — as a single source of truth. Workday, BambooHR, Rippling, Personio, Hibob, Deel and ADP lead. The CRM of modern HR and the centre of payroll, benefits and performance integration.
67
OKR (Objectives & Key Results)
A goal-management framework that Andy Grove developed at Intel and John Doerr scaled at Google. Objectives are qualitative and inspiring ("be the global leader"), while Key Results are measurable and ambitious (3-5 numbers). Pushes teams to focus on impact rather than shipping motion.
68
Performance Review (360 / Peer / Manager)
The process of evaluating an employee's performance from multiple angles. Manager review is the classic; peer review covers teammates; 360 combines manager, peers, reports and self; upward review goes from report to manager. Lattice, 15Five, Culture Amp and Leapsome run 360s; calibration sessions soften bias.
69
Calibration Session (HR)
A meeting where managers compare and adjust performance ratings together. Normalises bias like "everyone on my team got a 5/5 even though sales had a tough quarter". Uses tools like a talent map (9-box grid) or forced distribution; mandatory at every promotion cycle in modern tech companies.
70
Stack Ranking
A practice that Microsoft made famous in the 1990s — forcing employees into a bell curve (top 20%, middle 70%, bottom 10%). Microsoft scrapped it in 2013 amid critiques of "siloed competition" and killing collaboration. Modern tech mostly avoids it but it survives in things like Amazon's URA.
71
Async-First Workplace
A working model where documents, async video and thread-based communication are the default — meetings and live chat are not. Pioneered by GitLab, Automattic, Doist and Buffer; advantages include deep work, a shared written record and timezone-friendly global teams. The cure for the "too many meetings, too little work" era.
72
Hybrid Work
A model where employees spend part of the week at the office and part remote. 3-2 (three days in, two remote) and Tuesday/Thursday office days are classic patterns. Post-pandemic, more than 60% of knowledge workers prefer it; office-utilisation, hot-desking and schedule-sync tools have emerged.
73
Remote-First
A "remote by default, office optional" working model. Communication, hiring, onboarding and performance are all designed remote-first. GitLab, Zapier, Doist, Buffer and Basecamp are pioneers; the office becomes a perk, a social hub or a team-offsite venue. Different from hybrid: the office is genuinely optional.
74
EVP (Employee Value Proposition)
A structured answer to "why work here?" — pillared on compensation, benefits, career, culture and impact. A strong EVP cuts recruiting cost up to 50% and lifts retention 20%. The HR application of marketing's customer-value proposition.
75
Compensation Band
A company-policy structure that defines the salary range for a role and level — for example "Senior Software Engineer L5: $180K-$240K plus RSUs plus bonus". Set against 25th, 50th and 75th percentile market data sourced from Radford, Mercer, Pave and Levels.fyi.
76
RSU (Restricted Stock Unit)
Company shares granted to an employee that vest over time. The standard is a 4-year schedule with a 1-year cliff; in tech, RSUs make up 30-50% of an SWE comp package. Liquid in public companies and valued via the annual 409A in private ones — both the dilemma and the wealth maker of modern tech compensation.
77
ESOP (Employee Stock Option Plan)
A plan that gives employees the option (call option) to buy company shares. The employee profits when the strike price sits below the fair-market value; after vesting, the exercise and tax planning are critical. Common at early-stage startups in place of RSUs; ISOs and NSOs are the two flavours with very different tax treatments.
78
Vesting Schedule (4-year + 1-year cliff)
The standard equity-vesting schedule of the tech industry. The one-year cliff means nothing vests in the first year, then 25% vests at the one-year mark; the remaining 36 months vest 1/48 monthly. Hedges against early departures and is a building block of modern tech retention.
79
Hiring Loop / Interview Loop
A standardised hiring process of 4-7 interviews that evaluates a candidate on role fit, culture fit and technical skill. Typical sequence: recruiter screen → hiring manager → technical → systems design → behavioural → exec round → offer. Google standardised "4 interviews" in 2017.
80
Onboarding Buddy
A more senior teammate assigned to a new hire who plays the role of question-answerer, culture explainer and internal-network opener for the first 90 days. GitHub's Buddy System, Microsoft Onboarding Buddy and Buffer Bestie are example programmes; lifts 30-day and 90-day retention by 25%+.
81
Continuous Listening (Employee Experience)
A modern HR approach that gathers employee feedback continuously instead of waiting for an annual survey. Combines pulse surveys, onboarding feedback, exit interviews, lifecycle moments, Slack-bot pulses and open-text sentiment. Microsoft Viva, Glint (LinkedIn) and Qualtrics EmployeeXM lead — replacing the "annual moment of truth" with a real-time pulse.
82
Internal Mobility Marketplace
An internal platform where employees discover and apply to open positions, short-term projects and mentorship inside the company. Gloat, Eightfold, Fuel50 and SAP SuccessFactors lead. Brings talent retention, internal hiring and skill development under one roof; Unilever, Mastercard and AT&T are large adopters.
83
Skill Taxonomy
A structured catalogue of the skills a company values — the answer to "what capabilities matter to us?". Includes hard skills (Python, financial modelling) and soft skills (leadership, communication), rated from beginner to expert. The foundation of career frameworks, internal mobility and learning roadmaps.
84
Workforce Planning
A discipline that forecasts the company's future people needs and matches them to the growth plan. Covers headcount budget, hire-vs-build-vs-buy decisions, succession planning and attrition forecasting. Tooled with Visier, Workday and Anaplan; a shared CFO + CHRO responsibility reported quarterly to the board.
85
L&D (Learning & Development)
The HR function that systematically grows employee skills. Covers LMS platforms (Cornerstone, Docebo, 360Learning), microlearning (Udemy Business, LinkedIn Learning, Coursera Plus), mentoring programmes and internal academies. Modern tech companies spend $1,500-3,000 per employee on L&D.

— DECISION TREE

Is a retention engineering operation right for you?

Answer 4 questions Yes/No; get a clear recommendation.

01 / 04

Is your active customer base 50k+?

The lower threshold for retention investment to be economically meaningful.

— LET'S BEGIN

Do you notice when customers are leaving you?

In a 3-hour retention diagnostic we surface churn patterns, the lost segment and deliverability issues.

Request a retention diagnosticSee a sample lifecycle map