What is customer lifecycle communication?
Customer lifecycle communication is a way of organizing messages around the state of a relationship, rather than the company's desire to fill a sending calendar. Someone who has just started needs to understand the next step. Someone using the service needs information that helps preserve its value. A customer whose activity has declined may need an obstacle removed rather than another offer. When a relationship has ended, we should not assume the person wants to return or speak as though they already have. The practical question is: what do we know about the current state, which message fits it, and do we have permission and the ability to deliver whatever that message promises?
The idea is not that everyone will pass through the same states in the same order, or that automation can read their intentions. This is an operating map that helps teams choose more relevant communication and avoid contradictory messages. An inactive customer may resume using a service. A seasonal customer may remain silent until the natural moment of need. A new buyer may purchase once and need nothing else. We will therefore treat states as operational definitions open to review, distinguishing low activity from refusal of communication, and a genuine return from the company's wish to recover revenue. All the following examples and calculations are hypothetical teaching illustrations, not real customer results.
Read the states and messages without imposing an inevitable path
The map contains four states. A new customer is associated with a welcome program, introductory offers, a survey and newsletters. A known active customer is associated with statements, cross-selling or upselling, birthday or anniversary messages and loyalty offers. A known inactive customer is associated with early dormancy notifications, a reactivation program, and a survey or offers. For a lost customer, a welcome-back message appears when the customer returns. These pairings reflect the states and options in the course material. They are not instructions to send every message to every person or use every option in every business.

Original teaching diagram by Karim Elganainy, based on the mapping on pages 159 and 81 of the course materials, where the figure is attributed to Striata. The map is qualitative, not a timetable or a measurement of interaction, and welcome back is conditional on the customer's return. Public reference for the figure: lecture deck, page 26.
The original diagram's axes are brand interaction and time, but they do not establish that dormancy begins after a fixed number of days or that every relationship rises and falls in the same pattern. Use the map to ask questions, not to extract predictions from the curve. Some communication options may be entirely unsuitable for your business, such as statements the service does not produce or personal-occasion messages the customer would not welcome. Respecting those differences is part of applying the framework correctly. Useful automation chooses, excludes and waits. It does not succeed simply because it can run a long sequence without human intervention.
Define the state through a clear event and an appropriate group
Before writing a message, define the unit of the relationship. Are you communicating with a person, a company account, a branch or a particular subscription? Someone may be active in one service and inactive in another. One employee may use a product while a colleague pays the bill. Mixing these units produces messages that appear personal but rest on the wrong state. Document the definition of each state, the event that enters it, the event that exits it, the data source and the last update time. Add an insufficient-information state when the record cannot support a reliable judgment, rather than forcing every name into a category.
Separate the signal from its interpretation. An absence of logins may indicate non-use, but it may also reflect a tracking problem or work performed outside the system. Opening a message does not establish that the customer has received value, and closing it does not prove rejection of the brand. Choose events relevant to the service, then review their suitability with the team that understands actual usage. An event might be a completed order, a meaningful work session or a renewal becoming due. Record the definition before measuring, and do not change it later merely to improve the activity rate or reduce the reported number of inactive accounts.
New customers: reduce uncertainty at the start before promoting
A good welcome program explains what has begun, what comes next and whom to contact when help is needed. It does not require a long message listing every product feature. Choose the customer's first important task and offer a simple way to complete it, including conditions that might prevent completion. Someone may have subscribed only to learn more, while another has paid and is awaiting onboarding; do not use the same message for both. The promise should remain consistent with the page or conversation preceding signup, so the relationship does not begin with a discrepancy between what the person understood and what the welcome message offers.
An introductory offer, survey and newsletter are purposeful options, not mandatory additions. A short survey can help identify the intended use if the answer will genuinely change the assistance provided next. Collecting preferences nobody uses merely adds effort without clear benefit. Do not push another purchase before customers understand what they have received, or use an offer to conceal an important condition. When reviewing the message, ask whether someone can understand what to do, why it matters and how to stop optional messages. If the answer is unclear, improving the beginning takes priority over increasing the number of touches.
Active customers: protect value before asking for expansion
An active customer does not constantly need proof that the company remembers their name. They need relevant information that helps them manage and benefit from the relationship. Statements or periodic summaries may be useful when they form part of the service, provided they are accurate and understandable. Cross-selling or upselling can be appropriate when there is a genuine need and the ability to fulfill it, not simply because someone purchased before. Explain the difference between expanding the current solution and adding a complementary one, making costs and limits clear. Activity is a relationship signal, not unlimited permission for every offer, channel or frequency.
The source includes birthday or anniversary messages and loyalty offers, but choosing them requires relevance, permission and reliable information. Do not infer a personal occasion from a source not intended for that purpose, or collect a private date merely to imitate a marketing template. The anniversary of starting a service may be more relevant than a personal occasion; sending nothing may be the best choice. Loyalty is not measured by the number of promotional codes sent. It can instead be examined through the continuity, value and conditions of the relationship. Do not turn a small reward into a claim that the customer is satisfied or willing to recommend the business.
Inactive customers: examine the obstacle before sending an incentive
This state starts with a reasonable definition of declining activity in the context of the service, not a universal deadline copied from a different business. A seasonal retailer and software used weekly do not share the same rhythm. Before triggering a notification, check data completeness, unresolved service problems, and whether the customer requested a stop or changed communication channels. Declining activity may reflect a need that has ended, temporary pressure, a fault or an unsuitable offer. A good message does not claim to know the reason. It offers an appropriate step or a limited question that helps establish what can actually be repaired.
The map's early dormancy notification, reactivation program, survey and offer are possible tools, not compulsory escalation. Do not make every silence a reason for a larger discount: that may conceal an operational problem or train the audience to wait for incentives. Set a limit on attempts, a review period and a stopping condition. When there is an objection or a request for help, addressing the problem should come before promotion. Successful reactivation is not merely a link click. Define the event that constitutes a useful return, then monitor its continuation within a declared window rather than announcing success at the first interaction.
Lost customers: do not say welcome back before they return
Lost is an operational classification requiring a definition, not a judgment about the person or a promise that they will return. A contract may have ended, an account may have closed, or the relationship may have stopped under the business's conditions. The rule must distinguish what is known from what is assumed, and separate the end of a relationship from withdrawal of communication permission. A system may retain historical records for appropriate purposes, but the existence of a record does not establish eligibility for marketing messages. Do not treat every old name as an unrestricted recovery opportunity, or use the lost classification to override a customer's choice or try alternative channels without permission.
The framework's welcome-back message is conditional on an actual return. Define the event that establishes it in practice, such as restarting a service or completing a new purchase, not merely opening an advertisement. Then consider whether the person has returned to the same conditions or a different product, and what requires clarification again. Do not assume old preferences remain valid, or send a complete onboarding sequence when it is unnecessary. The purpose is to acknowledge the return and help with the next step, not to use warm language to conceal the fact that the company has not verified the state or the relevance of the current message.
Design priority and stopping rules before expanding automation
Someone may satisfy several conditions simultaneously: they purchased, opened a support request and became eligible for an offer. If systems operate separately, they may receive congratulations, a reminder and an expansion offer while awaiting a solution. Establish priorities reflecting purpose and context, distinguishing necessary service messages from optional messages through the review appropriate to the business. We are not assuming a universal legal classification for any message category here. The organization needs to know why it sends, who owns the decision, how it prevents conflict, and what happens when systems disagree about a customer's state or data validity.
For each automated path, document entry and exit conditions, frequency limits, exclusions and the person responsible for exceptions. Drawing arrows in a marketing tool is not enough. Test delayed data, duplicate events, delivery failures and preferences changing mid-sequence. Stopping must be possible and clear; an old synchronization should not re-enter someone into a sequence they voluntarily left. Keep a reviewable record explaining why a message was sent without exposing information the reviewer does not need. These rules make automation a means of executing a defined decision, rather than a machine that increases contact whenever another activity appears.
First hypothetical example: software for small businesses
Suppose a software service follows 120 new accounts during a 30-day test. Every number here is hypothetical, not a market benchmark. The account, rather than the individual user, is the measurement unit. Under the test conditions, 100 accounts were eligible for a welcome message through the selected channel, while 20 were ineligible under their recorded preferences. The message reached 90 eligible accounts. Eligibility is approximately 100 / 120 = 83.33%, and delivery among eligible accounts is 90 / 100 = 90%. We do not describe ineligible accounts as sending failures or change their preferences to improve coverage.
The team defined activation as completion of the first approved project, not opening the message. Of the accounts receiving the message, 54 completed that event, giving 54 / 90 = 60%. Across the original group, 66 accounts activated, giving 66 / 120 = 55%. This means 12 accounts activated outside the delivery group; not every activation can be attributed to the welcome message. These are precisely described, overlapping groups, not an experiment establishing causation. An appropriate comparison design is needed before claiming the message made the difference. We also need to investigate non-activation rather than assume that more messages alone will solve it.
Interpreting the first example: separate returning from continuing
In a separate group of 40 accounts classified as inactive under the test definition, 30 were eligible for communication. Of these, 24 received a help message addressing a usage obstacle known within the scenario. Among recipients, 9 returned to the qualifying usage event, giving a return rate of 9 / 24 = 37.5%. After a separate follow-up window, 6 of those returning still met the continuation criterion. This is approximately 6 / 9 = 66.67%, or 6 / 24 = 25% of the original recipients. These percentages answer different questions; choosing the largest and calling it the success rate of the whole sequence would be misleading.
Suppose the specified cost of preparing and running this test was SAR 1800. Dividing by observed returning accounts gives 1800 / 9 = SAR 200 per observed return; dividing by continuing accounts gives 1800 / 6 = SAR 300 per observed continuing account. Neither is a causal acquisition cost, profit or proven saving. These accounts might have returned without the message, and the value and costs of serving them have not been calculated. The proposed decision is to examine what prevented the other three from continuing and review the obstacle addressed by the message, rather than automatically increase the budget on the strength of one apparently high return rate.
Second hypothetical example: a seasonal household-supplies retailer
Suppose a retailer sells household supplies associated with widely separated seasons. A week without a purchase cannot establish inactivity. The team therefore derives its rules from a purchasing rhythm appropriate to the business and checks them before implementation. In a hypothetical test group of 200 customers classified as inactive under the selected rule, 150 were eligible for one offer and 135 received it. Eligibility is 150 / 200 = 75%, and delivery is 135 / 150 = 90%. These calculations do not establish that the classification rule is valid in itself. They describe what happened after applying a definition that needs independent review.
During the test window, 27 recipients purchased, giving an observed purchase rate of 27 / 135 = 20%. There were 3 returned orders in this group, leaving 24 retained purchases. The retained-purchase proportion is approximately 24 / 27 = 88.89% among buyers and 24 / 135 = 17.78% among recipients. Here we assume one order per buyer to keep the units consistent. That assumption must be declared because multiple orders would make the number of buyers different from the number of orders. Do not call the first purchase alone a recovery of long-term loyalty, or omit returns from the story after reporting the initial result.
Interpreting the second example: review value and welcome-back conditions
If each retained order has an average value of SAR 250, the total is 24 × 250 = SAR 6000. Assume a contribution margin of 30% after the cost of goods and before campaign costs: contribution is 6000 × 30% = SAR 1800. If the specified campaign cost is SAR 1200, the remainder is 1800 - 1200 = SAR 600 under those assumed line items only. This is neither the company's net profit nor a proven causal return. Other costs may exist, and we do not know how much purchasing would have occurred without the offer. The purpose is to explain why sales value alone is insufficient for judging the commercial case for reactivation.
In another, separate record, the retailer observed 10 customers previously classified as lost who subsequently completed a new purchase. Of these, 8 were eligible for the optional welcome channel and 7 received the message, giving delivery of 7 / 8 = 87.5%. Do not mix these customers with the previous sample or send welcome back to everyone classified as lost. The real event preceded and triggered the message, so their return cannot be attributed to communication delivered only after the purchase. Instead, examine whether the message clarified the next step, respected current preferences and prevented repetition of an introductory offer no longer appropriate to the customer.
A dashboard that does not reward message volume alone
Build the dashboard around clearly defined groups: eligible states, those included under the communication purpose and channel permissions, delivered messages, useful events, and continuation within a specified window. State the denominator and period beside every percentage. Do not compare a path contacting everyone eligible with another selecting only the most ready, then call the difference superior content. Separate relationship state from message state: failed delivery does not make an active customer lost, and successful delivery does not reactivate anyone. When data are incomplete, show the gap rather than silently assigning it to success or failure.
Monitor indications of harm or poor fit alongside response, such as unintended repetition, objections, stop requests and service problems connected to the message's promise. Do not set targets encouraging the team to ignore these signals in pursuit of conversion. Qualitative case reviews may be needed to understand causes when numbers are small. A message that produces no purchase can still be useful if it answers a service question, but that result must be named honestly. Choose the measure from the purpose. Do not redefine the purpose after seeing results, or present inbox activity as a substitute for the relationship value you intend to support.
A limited implementation plan before scaling the paths
Start with one service, one state and one message whose effects can be reviewed. Bring together the content, data and service owners, agreeing on the definition, event, permission and authoritative information source. Test invented records before introducing real ones: a new customer, an existing customer, someone who stopped communication, delayed data and an open support request. Examine which message would be sent and what would prevent it. The ideal case working is not enough. The real value of operating rules appears when they prevent inappropriate sending or reveal a conflict requiring human judgment before it reaches the customer.
After review, run a limited test within existing permissions and set an evaluation date and stopping conditions. Retain the message version, inclusion rule and event sources so the result can be explained. If service conditions change, do not leave the sequence running against an old version. As you scale, add another state only when you can explain responsibility, the reason for communication and the exit mechanism. Do not test content while changing price, channel and timing together, then attribute the result to one wording choice. Greater risk or more sensitive data requires appropriate specialists and tighter boundaries, not greater confidence in automation merely because it runs quickly.
Practical exercise: choose the right message and denominator
You have a hypothetical group of 80 inactive accounts. Of these, 60 are eligible for the reactivation channel and 48 receive the message. A total of 12 return to the defined usage event, and 9 continue through the follow-up window. The test costs SAR 2400. At the same time, a separate record contains 5 lost accounts that have actually returned, but 2 are ineligible for optional welcome communication through the available channel. Calculate delivery, return, continuation, cost per returning account and cost per continuing account. Then identify who enters the welcome-back message without mixing the groups or assuming the return creates new permission.
The answer: delivery is 48 / 60 = 80%; return among recipients is 12 / 48 = 25%; continuation among returning accounts is 9 / 12 = 75%, and among recipients it is 9 / 48 = 18.75%. Cost per return is 2400 / 12 = SAR 200, and cost per continuing account is approximately 2400 / 9 = SAR 266.67. Optional welcome communication includes 3 accounts that actually returned, not every lost account or the two ineligible accounts. These hypothetical descriptive results do not establish the message's causal effect. Before scaling, ask why three returning accounts did not continue, whether the usage definition remains appropriate and what evidence would change your decision if the test showed that inadequate communication was not the problem.
Application limits and mistakes that should stop the process
This framework does not replace understanding the product and relationship. Purchases may naturally be infrequent, a service may be a project that ends successfully, or the organization may remain a customer while its employees change. Do not label the legitimate end of a need as a loss that must be repaired at any cost. The diagram also supplies no universal dormancy thresholds and does not determine what local rules or contractual terms permit. Review actual implementation with specialists when necessary. Use only data that serve the purpose: personalization that surprises customers with information they did not expect to be used may weaken trust even if it temporarily increases clicks.
Stop when message opens are confused with activity, welcome back is sent before a return, stop requests are ignored or expansion offers are triggered during unresolved service problems. Also stop when denominators change to improve reporting or all revenue is attributed to the last message. Correction begins with a clear definition and an owner able to change the rule, not an additional automated apology for every error. Review cases in which automation prevented sending as well as those in which it sent, because appropriate restraint is part of system quality. Message volume should not become the only evidence that the team is working or the relationship is improving.
Conclusion: send because the state warrants it, not because the calendar allows it
My practical choice is to begin with the transition creating the most confusion for the customer: an unclear start, declining use with a known obstacle, or a return requiring a newly explained next step. Write down the state, evidence, permission and responsibility. Then design one message and test the rules that prevent it before expanding. Preserve the distinctions between new, active, inactive and lost, but do not let the labels replace conversation or observation. The best improvement may be changing a service procedure or removing an unnecessary message, rather than adding another channel or sequence to a system already struggling with contradictory definitions.
The value of lifecycle communication is that it connects a message to an understandable reason in the relationship and makes its timing and promise reviewable. It does not guarantee that a customer will stay, return or become loyal, but it helps avoid inappropriate communication at an inappropriate time. When interpreting results, ask what you actually observed, who the measurement included and what cannot be attributed to communication. When a customer returns, welcome them because they returned, not because you wish they had. With that simple rule, personalization becomes assistance respecting a person's context and choices, rather than a personal-looking surface concealing automation that does not understand the relationship's limits.
Sources and the limits of the map we use
This lesson is based on physical page 159 of the introductory digital-marketing material and physical page 81 of the customer-experience material in the supplied course files. Both show a diagram attributed on the slide to Striata, pairing customer state with suitable communication types. The customer-experience explanation clearly separates new and active customers, while some introductory commentary groups them together; this lesson uses the explicit separation. The same figure can be reviewed in the published lecture deck, page 26. The numbers 159 and 81 remain references to the supplied files, not the public deck. The attribution does not establish when the diagram was invented or that it is a universal standard, and it gives no basis for adding periods or numerical thresholds absent from those pages.
The event definitions, priority rules, test plan and measures discussed here are original practical analysis, not additional instructions attributed to the diagram. For comparison with the work preceding an agreement, consult the Arabic lesson on the business-to-business buying journey. That lesson focuses on understanding an organization's decision, while this one focuses on communication appropriate to the relationship's state. We do not confuse someone's state in a messaging system with agreement to purchase, or treat movement between lifecycle states as a complete measure of profitability, satisfaction or the quality of service actually received.

Comments
No published comments yet.