The Executive Summary
Agency founders at $60–$150K/month lose $10,000–$20,000 per churn incident because behavioral warning signs go undetected for 2–4 months.
Who this is for: Service agency founders at $60–$150K/month running 6+ active retainer clients who lose clients without warning
The signal blindness problem: Clients become quietly unhappy at month 1; cancellation emails arrive at month 4–5; the 2–4 month discovery gap costs $10,000–$20,000 per incident, up to $30,000/year across 2 incidents
What you’ll learn: The Delivery Dashboard — Client Health Score (5 signals, 1–3 scale, max 15 points), Red/Amber/Green Classification, Proactive Intervention Triggers, and Dashboard Review Cadence
What changes if you apply it: Client health becomes a number you govern weekly, not an impression you hold until a cancellation email arrives
Time to implement: First full scoring session runs in 55–80 minutes; weekly review cadence is 20 minutes; monthly reassessment cycle
Written by Nour Boustani for service agency founders at $60–$150K/month who want to catch at-risk clients early without waiting for them to raise concerns.
› Library Navigation: Quick Navigation · Service Agencies
Spot At-Risk Clients Before They Send the Cancellation Email
Agency founders at the Scaling band often learn a client is unhappy only when the cancellation email arrives. By then, the client may have spent 2–4 months accumulating frustrations that no one surfaced or resolved.
With 6–10 active retainer clients, a founder cannot keep a detailed picture of every relationship in their head. Frequent communication can make a client feel healthy. Silence can look like satisfaction, even when a client is close to leaving.
At an average retainer of $5,000/month, a client who could have been saved through early intervention represents an estimated $10,000–$20,000 in preventable churn. That estimate assumes a 2–4 month gap between the first signs of dissatisfaction and the cancellation. The opportunity to intervene existed earlier, but the founder had no system to spot it.
The Delivery Dashboard turns those signals into a weekly 20-minute review through four components:
Client Health Score
Red/Amber/Green Classification
Proactive Intervention Triggers
Dashboard Review Cadence
Together, they give the founder a way to identify relationship risk before a cancellation email becomes the first clear warning.
Where are you with this right now?
“I’ve lost clients who never said anything was wrong.” You’re inside the constraint. The five-signal health score below maps the specific indicators that precede cancellation decisions. Start at Component 1: Client Health Score.
“We have regular calls with clients so we’d know if something was off.” Regular calls produce the information clients choose to share. The five signals track what clients reveal through behavior regardless of what they say on calls: payment timing, revision count, engagement level, results trajectory, and expansion signals. Clients who are planning to leave often maintain cordial call behavior until they don’t. Start at the Stage Filter section.
“We only have 3 clients — this feels like overhead.” The stage filter for this system is 6+ active clients. Below that threshold, direct relationship management is sufficient. The prerequisite at 3 clients is building the governance that scales — I Spend All Day in My Inbox Answering ‘Where Is This’ Questions — Professional Project Governance installs the project transparency layer that makes client health scoring tractable at scale.
Try This Now
Think of your three longest-running current clients. When did each last give you unsolicited positive feedback about results? A polite response on a call does not count.
If you cannot remember for one or more of them, you are relying on an assumption, not a signal. The Client Health Score below turns that assumption into a number you can act on.
The 2–4 Month Discovery Gap and Its Cost
An agency that first discovers client unhappiness in a cancellation email has no early warning system.
What Quiet Clients Can Hide
A 6-person SEO agency generating $85K/month has eight active retainer clients. The founder speaks regularly with six, and those relationships feel healthy. The other two communicate only through project management comments and monthly report acknowledgments. The founder reads their silence as satisfaction.
One of the quiet clients cancels in month seven, saying they are “not seeing the results we hoped for.” A review of the monthly reports shows results below target in months three, four, and five. The client raised no concern on the monthly calls.
What Rising Revisions Can Signal
At a 4-person performance marketing agency generating $70K/month, a client paying $8,000/month submits nine revision requests in 30 days, up from an average of two. The founder notices but attributes the increase to the client being “thorough.”
The client cancels in month four, citing a “creative mismatch.” The increase in revisions was a signal to investigate, but no one flagged it.
What Slower Payments Can Signal
At a 7-person creative agency generating $110K/month, one client pays three days late in month two, five days late in month three, and eight days late in month four. The founder follows up each time but does not connect the pattern.
The client cancels in month five. Slower payment did not establish why the relationship was weakening, but the pattern warranted a conversation before the cancellation.
The Discovery Gap
Month 1: The client becomes unhappy. Quiet signals appear, but the founder sees no warning.
Months 2–3: Revision requests rise or payments slow. The founder does not connect the changes.
Months 3–4: The client makes an internal decision to leave. The relationship can still look normal.
Months 4–5: The cancellation email arrives. The founder discovers the problem after a 2–4 month lag, with an estimated $10K–$20K in preventable churn.
The problem is signal blindness. Clients may not say, “I am becoming unhappy with your agency.” Instead, their behavior changes:
They pay slightly later.
They request more revisions.
They stop engaging with reports.
They stop mentioning expansion.
They answer questions on calls but stop asking their own.
These shifts are measurable. Without a system to track them, they remain invisible until the cancellation email arrives. Clients who appear to cancel without warning may have given warnings the agency had no way to read.
The Limits of Strong Client Relationships
“Build strong client relationships and they’ll tell you when something is wrong” is useful advice, but it is not an early warning system. It relies on a dissatisfied client choosing to start a difficult conversation.
A client may avoid that conversation because they are unsure what to say, reluctant to create conflict, or not ready to renegotiate the engagement. They stay quiet while they consider their options, then leave with a brief explanation such as “budget” or “direction change.”
The fix is not to stop investing in relationships. It is to pair those relationships with a way to measure changes in client behavior.
Stage Filter: Scaling Band ($60–$150K/month), 6+ Active Clients
At the Scaling band, the constraint is visibility across 6+ client relationships, not necessarily relationship quality.
When a cancellation comes as a surprise, a founder may improve the service without adding a way to detect dissatisfaction earlier. Delivery improves, but the discovery gap remains.
Use the Client Health Score for active retainer clients with at least 60 days of delivery history. A client in their first 30 days does not yet have enough behavioral history for a meaningful score.
The Real Cost of Late Detection
For an agency with eight active retainer clients averaging $5,000/month, the estimated preventable churn exposure is $10,000–$20,000 per incident if deterioration goes undetected for 2–4 months.
A client who remains silently unhappy for 90 days represents $15,000 in billing at a $5,000 monthly retainer. That figure describes revenue billed during the detection gap, not revenue already lost. The risk is that the relationship ends without a timely chance to intervene.
Use these assumptions to calculate your own exposure:
Completed Example
- Average monthly retainer: $5,000
- Discovery lag: 3 months
- Revenue billed during the gap per incident: $5,000 × 3 = $15,000
- Active clients without health scores: 8
- Assumed incidents per year: 2
- Illustrative annual exposure: $15,000 × 2 = $30,000Your Calculation
- Average monthly retainer: $[amount]
- Estimated discovery lag: [number] months
- Revenue billed during the gap per incident: $[amount] × [number] = $[amount]
- Active clients without health scores: [number]
- Assumed incidents per year: [number]
- Illustrative annual exposure: $[amount] × [number] = $[amount]What Changes as Detection Is Delayed
Within 30 days: The first Client Health Score pass takes 60–90 minutes across active clients and produces a ranked list of relationships to review. An amber or red client can receive an intervention before the next billing cycle.
After 30–60 days: Behavioral changes continue without a review. A check-in that could have addressed an early concern may now require a more direct conversation.
After 60+ days: A client may already have decided to leave. The score can still flag the relationship, but the response may need to shift from a proactive check-in to a recovery conversation.
Check Whether You’re Ready to Install the Dashboard
Proceed only if all three criteria are met:
You have 6 or more active retainer clients.
You have at least 60 days of delivery history per client to establish a behavioral baseline.
You can access payment, communication, and project data in one session.
Pass: All three criteria are met. Proceed to scoring.
Fail: Any criterion is unmet. Address it before scoring:
Fewer than 6 clients: Manage relationships directly and return to the dashboard at 6+ clients.
Less than 60 days of history: Wait until the client reaches 90 days before installing the score.
Data scattered across inaccessible tools: Consolidate the sources before scoring.
Scoring with less than 60 days of history can create a false baseline and inflate Green designations.
One thing from this section:
The signals that precede client cancellation are behavioral, not verbal — they are measurable if a measurement system exists and invisible without one.
The cost is documented. The framework that installs the measurement layer has four components, and the sequence is fixed: you cannot govern health tiers before you have a score, and you cannot trigger interventions before you have tiers.
How to Build a Client Health Dashboard That Flags Churn Risk Early
A client management system that waits for clients to raise concerns is reactive. The Delivery Dashboard gives the agency a way to review behavioral signals and act before a cancellation arrives.
Its four components work together:
Client Health Score produces a score from observable behavior.
Red/Amber/Green Classification assigns a tier to the score.
Proactive Intervention Triggers specify when to act.
Dashboard Review Cadence keeps the review running without relying on founder initiative.
Component 1: Score Client Health Using Behavioral Signals
Assess every active client monthly against five signals. Score each signal from 1 to 3, for a maximum total of 15. Use evidence from delivery, written communication, engagement, and payment rather than relying only on what a client says on calls.
Results vs. targets: Score 3 when agreed deliverables meet or exceed the benchmarks set at onboarding; 2 when slightly below target; 1 when materially below target for two or more consecutive months.
Communication sentiment: Score 3 when written communication is engaged and positive; 2 when neutral or transactional; 1 when terse, delayed, or minimal.
Engagement level: Score 3 when the client attends calls and engages with reports and deliverables; 2 when attendance or response rates drop; 1 when the client consistently misses calls or leaves reports unacknowledged.
Payment timeliness: Score 3 when payments are consistently on time; 2 for one late payment in the last 90 days; 1 for increasing lateness or a payment currently overdue.
Expansion/reduction signals: Score 3 when the client mentions expansion, additional projects, or referrals; 2 when there is no expansion discussion; 1 when they raise scope reduction, a pause, or restructuring.
Add the five scores and assign a tier:
Green (10–15): Maintain the standard delivery cadence.
Amber (6–9): Investigate the change and initiate proactive communication.
Red (below 6): Start a founder-led intervention within 48 hours.
First Scoring Pass: Example
Before: Eight active clients with no health monitoring.
Diagnostic finding: Two amber clients and one red client had gone undetected.
Action: Complete interventions for the flagged clients.
After: Two relationships stabilized.
Timeline: 30 days from the first scoring pass.
The output should be a ranked list of every active client, with a total score and tier. This makes it easier to decide which relationships need attention instead of relying on the founder’s impression.
The score does not directly measure satisfaction on calls. A client may sound positive while other signals deteriorate. Review the five signals together, and use the score to prompt a conversation, not to replace one.
Quick Signal
Pick the client you would call “the quiet one” and score all five signals now. If three or more signals score 1, investigate the relationship even if the last call felt fine. Under the total-score thresholds above, those three scores alone do not necessarily make the client amber; use the calculated tier alongside the signal pattern.
Component 2: Turn Health Scores Into Required Actions
The Red/Amber/Green Classification turns each client’s health score into a response with a deadline. A tier is useful only if it triggers action.
Green (10–15)
Maintain the standard delivery cadence.
Review the score monthly. No additional intervention is required.
Amber (6–9)
Within 5 business days, schedule an unscheduled, agenda-free check-in. Ask, “How are you finding the engagement so far?” Make space for concerns rather than turning it into a status call.
If results vs. targets is the low-scoring signal, prepare a results recovery narrative before the next scheduled call. Do not wait for the client to raise the shortfall.
Move the client to a weekly health review until two consecutive monthly scores return to Green.
Red (below 6)
Within 48 hours, the founder leads a direct review, not an account manager. Open with: “I want to make sure I understand exactly where we stand and what would make this engagement work better for you.”
Within 5 days, make a specific written resolution offer that addresses the concern. Options include revised deliverables, an adjusted timeline, an additional resource, or scope renegotiation.
Within 14 days, if the offer has not stabilized the relationship, prepare a full account restructure or a clean exit plan.
An amber label without a check-in does not close the discovery gap. A red label without founder action leaves the relationship exposed. The classification works only when the required actions happen.
Component 3: Act on Early Client Risk Signals
Proactive Intervention Triggers prompt action before a client raises a concern. Check both the current score and how it has changed.
Single-signal deterioration: If any of the five signals drops to 1, check that issue directly, regardless of the total score. For example, a payment timeliness score of 1 with an overall score of 11 calls for a conversation about payment, not the full Amber protocol.
Two-point drop in 30 days: If the total score falls by 2 or more points between monthly assessments, run an immediate mini-review. A move from 12 to 10 remains Green, but the downward trend needs attention.
Three consecutive months at Amber: If a client stays Amber for three consecutive months without returning to Green, escalate to the Red protocol regardless of the current score. The Amber actions have not resolved the issue.
A tier is not a permanent status. A client who remains Amber for six months has had an unresolved risk signal for half a year.
Quick Signal
Review the last three months of invoices, call notes, and project comments for one client. Were they more engaged in month one than they are now? If so, name the specific signal that changed. Start the intervention there.
Component 4: Review the Dashboard Every Week
Score every client monthly. Review the dashboard weekly for 20 minutes. Monthly scoring brings together evidence across all five signals; the weekly review catches changes and tracks interventions between scores.
During each review, check:
Amber and Red clients: What happened this week on each open intervention?
New signals: Did a payment arrive late, an email thread turn terse, or a client miss a call? Log the change before it becomes a pattern.
Upcoming scores: Is any client approaching their monthly assessment, and does their current trajectory suggest the tier may change?
Put the 20-minute review on the calendar at the same day and time each week. Do not leave it to “when I have time.”
Weekly Dashboard Review Template
Client name: [name]
Current tier: [Green/Amber/Red]
Intervention open: [Yes/No]
- Payment status this week: [status]
- Communication tone this week: [observation]
- Single-signal flares: [signal or none]
- Intervention update, if Amber or Red: [action and status]Time to complete: 20 minutes across the dashboard.
By week four, the founder should have a ranked client health list reviewed weekly, with visible risks and open actions. The dashboard cannot guarantee that a cancellation will not be a surprise, but it gives the founder a chance to notice deterioration and respond.
What the Delivery Dashboard Changes
The four components build a habit of checking behavioral evidence rather than relying on how a relationship felt during the last call. Before a renewal conversation, the founder can review the client’s score and trajectory instead of working from a general impression.
The approach is similar to tracking accounts receivable: a payment trend prompts attention before it becomes a larger problem. Here, the founder watches for changes across the client relationship and acts on the trigger.
What AI-Assisted Monitoring Looks Like
For eight active clients, a manual monthly scoring pass takes an estimated 60–90 minutes to review project history, payment records, and communication threads. An AI-assisted pass is estimated at 20–30 minutes when the relevant data is available. AI can help surface potential signals; the founder still needs to check the evidence, assign the scores, and decide what action to take.
AI Prompt for Monthly Client Health Scoring
Use this prompt with Claude or ChatGPT. Include the agreed targets, payment history, and enough client activity to assess all five signals. Remove confidential information you are not permitted to share.
Assess client health for [client name] using only the information below.
Client data
- Agreed targets: [targets]
- Results for the last two months: [results]
- Emails, project comments, and call notes from the last 30 days: [communications]
- Call attendance and responses to reports or deliverables: [engagement]
- Payment due dates and actual payment dates for the last 90 days: [payment history]
- Expansion, reduction, or pause discussions: [details or none]
- Previous monthly health score: [score or unavailable]
- Consecutive months at Amber: [number or unavailable]
Score these five signals from 1 to 3:
- Results vs. targets: 3 = on track or exceeding; 2 = slightly below target; 1 = materially below target for two or more consecutive months.
- Communication sentiment: 3 = engaged and positive; 2 = neutral or transactional; 1 = terse, delayed, or minimal.
- Engagement level: 3 = fully engaged; 2 = attendance or responses have dropped; 1 = consistently disengaged.
- Payment timeliness: 3 = consistently on time; 2 = one late payment in the last 90 days; 1 = increasing lateness or a payment currently overdue.
- Expansion/reduction signals: 3 = positive expansion signals; 2 = no expansion discussion; 1 = scope reduction or pause language.
Output
- List each signal, its score, and one sentence explaining the evidence.
- If evidence is missing, write “insufficient information” for that signal. Do not guess.
- If all five signals are scored, add the total out of 15 and assign Green (10–15), Amber (6–9), or Red (below 6).
- Flag any signal scored 1 for a targeted check, even if the total is Green.
- Flag a drop of 2 or more points from the previous monthly score, if provided.
- If Amber has persisted for three consecutive months, recommend escalation to the Red protocol.
- End with the next action and its deadline. Do not assign a Green tier to an incomplete score.A structured review can help a founder notice a gradual shift from enthusiastic to transactional communication instead of anchoring on the most positive recent exchange. AI can organize the evidence and propose scores, but the founder must verify them. It cannot assess a 90-day payment pattern or two consecutive months of results from only 30 days of data.
With complete inputs, the AI-assisted workflow is estimated to produce a client health ranking within 30 minutes. The value is not the speed alone. It is having a repeatable way to examine quiet clients as closely as vocal ones.
I’ve seen the red client revelation land hard on founders who have been in business for five or six years. Not because the $10,000–$20,000 per incident figure surprises them in the abstract. It’s because the first scoring pass shows a client they would have called ‘fine’ sitting at a 5. They’ve billed that client $15,000 in the three months since the first signals appeared. The score didn’t create the problem. It made it visible.
The five-signal score does not improve a relationship by itself. It makes the available evidence easier to act on. If a client scores 1 on two or more signals, investigate those signals promptly; do not dismiss them as a bad month or assume they prove the client has decided to leave.
Check That Every At-Risk Client Has an Intervention
Before closing the scoring session, confirm all three criteria:
Every active retainer client with 60+ days of history has a health score.
Every Amber client has a check-in scheduled within 5 business days.
Every Red client has a founder-led review booked within 48 hours.
Pass: All three criteria are met.
Fail: Any criterion is unmet. Do not close the session until you address it.
A client has a score but no scheduled intervention: Assign and schedule the next action.
A Red client has no review booked within 48 hours: Stop and book the founder-led review.
An Amber client’s check-in is deferred beyond 5 business days: Reclassify the client as Red and apply the Red protocol.
A health score without a scheduled next action is a label, not a governance system.
Premium Toolkit available for members
The Delivery Dashboard System includes:
Client Health Scorecard — identify behavioral churn signals early and prioritize at-risk clients before they cancel
Health Tier Governance Protocol — trigger clear interventions for Green, Amber, and Red clients before risk compounds
Health-Score-to-Churn Predictor Calculator — build agency-specific churn patterns that improve retention decisions over time
Plug-and-play AI diagnosis sessions — drop into Claude, Gemini or ChatGPT, answer a few questions, save hours of guessing, get your exact next move
Audio key points — concentrated frameworks you can absorb in minutes, implement while you move
Unlock 750+ ready-to-use constraint toolkits — built to solve every business problem operators actually face.
Prevent $10,000-$20,000 churn incidents by identifying behavioral risk signals two to four months before cancellation.
Cancel anytime. Every download you’ve accessed stays with you.
This system is built for Scaling-band agency founders with six or more active retainer clients.
If project data is unstructured, install I Spend All Day in My Inbox Answering ‘Where Is This’ Questions — Professional Project Governance first.
The first health score pass is one session. The toolkit produces the first at-risk client identification before the next billing cycle closes.
One thing from this section:
A client health score makes the state of every client relationship legible in a number — removing the subjective impression that allows deteriorating relationships to go undetected until cancellation.
The framework is built. The next step is the implementation sequence that produces a working health score for every active client in a single session.
Install Your Client Health Dashboard in One Session
The first session should produce a health score for every eligible active retainer client. Mark clients with less than 60 days of history or missing evidence separately rather than forcing a score. That first pass becomes the basis for intervention and ongoing review.
Step 1: List Every Active Retainer Client (5–10 Minutes)
Pull your invoicing records and list every client with an active recurring engagement. Exclude one-off project clients.
Record the client name, monthly retainer value, engagement start date, and primary contact.
If this takes more than 10 minutes, build the list from recurring invoice line items in the last 30 days.
Aim for one entry per active retainer client, typically 6–12 entries.
Mark any client with fewer than 60 days of delivery history as “insufficient history.” Exclude them from the first scoring pass and add them in month three of the engagement.
Step 2: Gather the Signal Data (20–30 Minutes)
For each eligible client, collect evidence for all five signals. Start with activity from the last 30 days, then check the longer history required for payment patterns and consecutive months of results.
Payment records: Due dates, payment dates, and any pattern of increasing lateness.
Email or Slack threads: Changes in communication tone.
Project and delivery records: Call attendance, report responses, engagement with deliverables, and revision count.
Results tracking: Actual results against the targets agreed at onboarding.
Recent calls or messages: Mentions of expansion, future projects, referrals, scope reduction, or a pause. Check the last three calls or relevant threads.
Use the tools your agency already has. Allow 20–30 minutes for 6–8 clients. If the information is too scattered to gather in that time, collect what you can, record the gaps, and consolidate access before the next monthly scoring cycle.
The output is evidence for up to five signals per client. Write “insufficient data” where evidence is missing; do not assign a default score or Green tier.
Step 3: Score Each Client (15–20 Minutes)
Score each of the five signals from 1 to 3, add the scores, and assign a tier. Use the Client Health Scorecard in Toolkit 1 (PDF) or a simple list.
Score each signal independently before calculating the total. Do not work backward from the tier you expect.
If using AI, compare its proposed scores with your own assessment of the underlying evidence.
If scoring takes more than 20 minutes, choose the most representative observation for each signal. Where the pattern varies, use the original rule: average the observed pattern and round down, not up.
Where evidence is insufficient, mark the signal “insufficient data” rather than forcing a total or tier.
The output is a ranked list of eligible clients, each with a total score and tier. If every client scores Green, review the evidence and scoring criteria before accepting the result. Do not assume the first pass must contain an Amber client.
Step 4: Schedule Tier Actions (15–20 Minutes)
Open your calendar before ending the scoring session. Every Amber or Red client needs a dated next action.
Amber: Schedule an unscheduled 20-minute check-in within 5 business days.
Red: Book a founder-led review within 48 hours and block time to draft the written resolution offer.
Multiple Red clients: Address the lowest-scoring client first. If three or more score Red, review whether a shared delivery or communication problem needs attention.
The output is a calendar entry for every client scoring below 10. No Amber or Red client should leave the first pass without a scheduled intervention.
Installation Sequence
Step 1: List active retainer clients, 5–10 minutes. Output: Client list.
Step 2: Gather signal data, 20–30 minutes. Output: Evidence for five signals per eligible client, with gaps marked.
Step 3: Score clients, 15–20 minutes. Output: Score and tier for every client with sufficient evidence.
Step 4: Schedule tier actions, 15–20 minutes. Output: Dated interventions for every Amber and Red client.
Total estimated time: 55–80 minutes. The first pass identifies at-risk clients and schedules action before the next billing cycle closes.
How the Dashboard Works Across Three Agencies
Solo-Founder Digital Marketing Agency
Agency: $68K/month with 7 active retainer clients.
First pass: Six clients score Green. One $6,000/month client in month eight scores 7 (Amber).
Recorded signals: Results 1, communication 2, engagement 2, and payment 2. Expansion was left blank because no conversation had occurred.
Action: The founder schedules a results-focused check-in for the following week. The client confirms they have been concerned about results, and the founder presents a revised strategy.
Outcome: The client agrees to continue, retaining a $6,000/month account.
The four recorded scores total 7, but the blank expansion field makes the five-signal score incomplete. Under the scorecard’s rules, no expansion discussion scores 2; with that score, the total would be 9 (Amber). The record should be corrected before the final tier is logged.
4-Person Content and SEO Agency
Agency: $75K/month with 8 active clients.
First pass: Two clients appear Amber and one $4,500/month client in month six appears Red.
Recorded signals for the at-risk client: Results 2, communication 1 after three weeks of one-word emails, engagement 1 after missing two of three calls, and payment 2 after a slightly late payment. Expansion was recorded as 0.
Action: The founder calls within 24 hours and learns the client has been interviewing replacement agencies for two months. The founder offers a scope adjustment, a reduced rate for 60 days, and a documented results improvement plan.
Outcome: The client agrees to a 60-day extension. The $4,500/month account remains at risk.
A signal cannot score 0 under the 1–3 scale, and the four valid scores total 6, not 4. If the 0 meant no expansion discussion, the correct expansion score is 2 and the total is 8 (Amber). If it meant scope reduction or pause language, the score is 1 and the total is 7 (Amber). The founder’s prompt call was appropriate, but the stated Red score is not supported by the recorded signals.
6-Person Performance Marketing Agency
Agency: $120K/month with 10 active clients.
Pattern: A quarterly review finds three clients from the same intake quarter scoring Amber at their six-month mark. All three score 1 on results vs. targets and use the same service line.
Action: The founder reviews that service line’s delivery process and finds a shared gap in reporting against benchmarks.
Outcome: The agency corrects the reporting gap at the process level rather than treating the three accounts as unrelated problems.
A cluster of Amber scores on the same signal calls for a shared-process review alongside individual client interventions.
Check That the Dashboard Is Installed
Before the first intervention, confirm:
Every active retainer client with sufficient history and data has a current health score. Clients without either are marked for follow-up.
Every Amber or Red client has a dated intervention on the calendar.
The weekly 20-minute dashboard review is booked as a recurring block.
If any item is missing, installation is incomplete. The first scoring session ends when every identified Amber and Red client has a scheduled next action. The next step is to check the scores against new evidence and review what the interventions produce over 90 days.
Test Your Client Health Dashboard Against Churn Risk
Calculate Your Preventable Churn Exposure
The calculator estimates exposure, not churn that is certain to occur. Revenue billed during a discovery gap is not itself lost revenue; it represents the period when an earlier intervention might have been possible.
Completed Example: Scaling-Band Agency
- Average monthly retainer: $5,000
- Active retainer clients: 8
- Assumed discovery lag without monitoring: 3 months
- Revenue billed during the gap per incident: $5,000 × 3 = $15,000
- Assumed incidents per year: 2
- Illustrative annual exposure: $15,000 × 2 = $30,000
- Illustrative monthly exposure: $30,000 ÷ 12 = $2,500
- Subscription cost: $12/month
- Illustrative exposure-to-cost ratio: $2,500 ÷ $12 ≈ 208:1Your Calculation
- Average monthly retainer: $[amount]
- Active retainer clients: [number]
- Estimated discovery lag without monitoring: [number] months
- Revenue billed during the gap per incident: $[retainer] × [months] = $[amount]
- Estimated incidents per year: [number]
- Illustrative annual exposure: $[amount per incident] × [incidents] = $[amount]
- Illustrative monthly exposure: $[annual exposure] ÷ 12 = $[amount]
- Subscription cost: $[amount]/month
- Illustrative exposure-to-cost ratio: $[monthly exposure] ÷ $[subscription cost] = [ratio]:1At $5,000/month, a 2–4 month discovery lag represents $10,000–$20,000 billed during the gap per incident. At $10,000/month with a 4-month lag, that figure is $40,000. These are exposure estimates, not guaranteed recoverable revenue. The 208:1 figure compares modeled monthly exposure with subscription cost; it is not a measured prevention return.
Simulate the Intervention Before You Build
Starting Scenario
Agency: $90K/month revenue, with 9 active retainer clients averaging $5,000/month.
At-risk client: In month eight of a $5,000/month engagement; quietly dissatisfied since month five.
Signals: Results slightly below target and communication shifting from enthusiastic to transactional. The client does not complain and pays on time.
Without the Delivery Dashboard
Month eight: The engagement continues without a conversation about results.
Month nine, week three: The client pays the invoice but drafts a cancellation email.
Month nine, week four: The email arrives. The founder calls, and the client says they are “not seeing the growth we expected.”
Outcome: The client has already committed to a new agency. The agency loses a $5,000/month account. If the client otherwise would have stayed at least 12 more months, the modeled remaining relationship value is $60,000.
With the Delivery Dashboard
Month five: Results score 1, communication scores 2, and the total falls to 8 (Amber). The result signal also triggers a targeted check.
Month five, week three: The founder checks in: “I want to make sure what we’re delivering is hitting the mark for you.” The client shares their concern.
Following 60 days: The founder adjusts the strategy, introduces weekly progress updates, and sets a results review.
Month seven: The score reaches 11 (Green).
Month 12: In this modeled outcome, the client renews. The agency retains the $5,000/month account and preserves $60,000+ in potential relationship value, assuming at least 12 additional months.
The simulation shows how earlier detection creates a chance to intervene. It does not guarantee the client would renew or that the full modeled value would be recovered.
Two Futures: Six-Month Client Trajectories
These modeled scenarios extend through month six, not just 90 days.
Without the Delivery Dashboard
Month 1: Relationships appear normal. With no signals tracked, the founder judges client health by call energy and the absence of complaints.
Month 3: One client has sent one-word email replies for six weeks. Another misses a call and reschedules without explanation. The founder treats both as routine and focuses on new business.
Month 6: The first client sends a cancellation notice; the second reduces scope by 40%. The modeled impact is $8,000–$12,000/month in lost and reduced retainer revenue. Both showed behavioral changes from month two, but neither pattern was tracked.
With the Delivery Dashboard
Month 1: The first scoring pass identifies an Amber client. A check-in surfaces scope creep and delivery pressure, which the agency addresses before the concern compounds.
Month 3: After two consecutive monthly Green scores, the original Amber client returns to standard review. A new account in its second month shows low engagement; the founder logs the signal and schedules a check-in, but waits for sufficient delivery history before assigning a formal health score.
Month 6: Six monthly scoring cycles have been completed. In this modeled outcome, two Amber interventions have been executed and both concerns resolved without churn. The founder has a documented health trajectory for eligible clients to use in renewal, expansion, and difficult delivery conversations.
This is a simulation, not a promised result. The dashboard makes changes visible and prompts timely action; it does not guarantee retention.
Check Progress Over the First 90 Days
Day 30
Score every eligible active retainer client.
Complete the required Amber and Red interventions.
Save the scores as the baseline for future assessments.
Week 8
Complete the second monthly scoring pass.
Compare each client’s score with the first pass and apply the two-point drop trigger where needed.
Confirm the weekly 20-minute review is running.
Day 90
Complete the third monthly scoring pass.
Add the first 90 days of data to the Health-Score-to-Churn Predictor Calculator in Toolkit 3 (PDF).
Log any churn alongside the client’s preceding score trajectory.
If an Amber client has not returned to Green within 60 days, review the intervention. Did it address the low-scoring signal, or was it a general “how are we doing?” call? Adjust the action to the concern the score identified.
If Scores Do Not Match Client Behavior
Recheck the individual signals. Score communication sentiment across the full 30-day pattern, not the most recent exchange. Three weeks of terse replies are not erased by one friendly email.
Identify the error. Was a signal scored too generously, or was the tier applied incorrectly? A total of 8 is Amber, not “basically fine.”
Change one scoring criterion at a time. If results vs. targets is subjective because no target was defined, document a specific written benchmark for the client. Set benchmarks at onboarding for future clients.
Retest in the next 30-day scoring cycle. Keep the original scores and note the change so you can distinguish an actual improvement from a change in scoring.
What the Delivery Dashboard Helps You Notice
The clients who take the most account-management time are not necessarily the ones most likely to leave. A vocal client makes concerns visible. A quiet client may receive less attention even as engagement drops. The five-signal score gives the founder a reason to check both.
Early Signal 1: Results Decline While Communication Stays Positive
Results vs. targets scores 1 for two consecutive months.
Communication sentiment has previously scored 3.
The client may still be engaged while losing confidence in the results. Present a specific recovery plan before waiting for them to raise the shortfall.
Early Signal 2: Every Signal Stays Neutral
All five signals score 2 for three consecutive months.
The total remains 10, which is Green under the score thresholds, but no signal improves.
A stable score does not answer whether the client sees enough value to continue. Use the next conversation to ask what is working, what is not, and what they need from the engagement. An expansion conversation may open that discussion, but do not treat the neutral score as a reason to pitch more work.
The point is not that silence proves dissatisfaction. It is that silence alone cannot prove satisfaction. The score makes changes and flat patterns visible so the founder can check what they mean before a cancellation arrives.
The next section covers the failure modes that can stop the dashboard from working after its first cycle.
Prevent Score Inflation After the First Cycle
A client health dashboard can fail when scores start reflecting the founder’s optimism instead of the evidence. The single point of failure is replacing observable signals with a general feeling about the relationship.
How Score Inflation Happens
Month 1: A client scores 8 (Amber) based on behavioral evidence.
Month 2, communication: The client has a good call, so the founder scores communication sentiment at 3. The full 30-day written pattern still warrants 2.
Month 2, results: Performance remains below benchmark, but the founder scores it at 2 because the team has a good explanation ready.
Month 2, classification: The total rises to 11 (Green), and the founder closes the Amber intervention.
Month 4: The client cancels. The score history shows that the warning signals were still present when the client was reclassified.
The Health Tier Governance Protocol in Toolkit 2 (PDF) guards against this failure by tying each score of 1, 2, or 3 to an observable data point. A good call does not erase a month of terse messages, and an explanation does not change results against the agreed benchmark.
Results vs. targets: score is determined by the documented performance benchmark, not the founder’s sense of whether “it’s trending in the right direction”
Communication sentiment: score is determined by the 30-day written communication pattern, not the last call
Engagement: score is determined by attendance records and response rates, not the impression of commitment
Payment: score is determined by days-late data, not whether the client “always gets around to it”
Expansion signals: score is determined by whether expansion language has appeared in writing, not whether the founder “gets a sense they might expand”
The anchoring rule makes score inflation structurally difficult because the scoring criteria are observable rather than interpretive.
Failure Mode Analysis
Failure Mode 1 — Score inflation over time
Early Signal: The founder describes a client as “probably fine, just a tough month” while assigning them a Green score. The justification language indicates subjective overriding of signal evidence.
Recovery: Return to the anchoring criteria. For each signal that was scored above 1, identify the specific observable data point that justifies the score. If no observable data point exists, rescore at 1.
Correction Timeline: One scoring cycle with strict anchoring applied. The month where anchoring is reinstated often produces the most accurate — and most uncomfortable — client health picture the founder has had.
Failure Mode 2 — Amber interventions executed as generic check-ins
Early Signal: The Amber intervention call covers general relationship topics without addressing the specific low-scoring signal. The client says the call was “great.” The health score does not improve in the following month.
Recovery: Rebuild the intervention conversation around the specific signal that triggered the Amber designation. If results vs. targets was the low-scoring signal, the intervention is a results conversation — not a general “how are we doing?” call.
Correction Timeline: Immediate. Redesign the next intervention conversation before scheduling it. The signal that triggered Amber is the agenda.
Failure Mode 3 — Weekly review cadence drops after month one
Early Signal: The weekly 20-minute review block is skipped “because nothing changed this week.” The review is being treated as optional rather than as a fixed governance ritual.
Recovery: The weekly review is non-negotiable regardless of perceived status. The point of the review is to catch the single-signal flares that appear between monthly assessments — which by definition appear in weeks where the founder assumes “nothing changed.”
Correction Timeline: Immediate. Reschedule the review as a recurring block with a specific agenda: payment status, communication tone, any single-signal flare in the last 7 days.
Failure Mode 4 — Red protocol deferred beyond 48-hour window
Early Signal: A Red client is identified on Monday. The founder’s schedule is full. The founder plans to “deal with it next week.” The 48-hour window closes. The client cancels on Thursday.
Recovery: The 48-hour window for Red clients is not a guideline — it is the intervention threshold. A Red client who does not receive a founder-led contact within 48 hours is a near-certain churn outcome. Block the time immediately regardless of schedule.
Correction Timeline: Immediate. The Red protocol supersedes all non-emergency schedule commitments for the 48-hour window.
Fix the Four Dashboard Failure Modes
FM1, score inflation: The founder says a client is “probably fine” while assigning Green. Anchor each signal score to observable evidence and check the correction in the next scoring cycle.
FM2, generic Amber intervention: The check-in does not address the low-scoring signal. Set a conversation agenda around that specific concern before the call. Fix this immediately.
FM3, skipped weekly review: “Nothing changed” becomes a reason to miss the review. Keep the recurring 20-minute calendar block and use it to confirm whether anything changed. Fix this immediately.
FM4, delayed Red response: The founder says, “I’ll deal with it next week.” Move non-emergency work and book the founder-led review within 48 hours. Fix this immediately.
What Happens Without an Early Warning System
Month 1
Client relationships appear stable.
No behavioral signals are tracked while the founder divides attention between delivery and new business.
Month 3
One client has been deteriorating for two months.
They still pay on time and attend calls, so the founder misses other changes.
The intervention window has been open for 60 days.
Month 6
The client cancels, citing “results not meeting expectations.”
A post-mortem improves delivery but adds no detection layer. The next at-risk relationship can follow the same pattern.
With the Delivery Dashboard Installed
Month 1
The first scoring pass identifies one Amber client and one Green client whose score is falling.
The Amber client receives a check-in. The founder discusses the Green client’s results trajectory before its tier changes.
Month 3
After three scoring cycles, both clients have stabilized in this modeled scenario.
The Health-Score-to-Churn Predictor Calculator in Toolkit 3 (PDF) has its first 90 days of agency-specific data.
If all clients in that limited history with communication and results scores of 1 later churned, the recorded rate would be 100%. Treat it as a pattern to investigate, not a reliable prediction.
Month 6
After six scoring cycles, the founder reviews each client’s score before renewal.
Green conversations draw on documented delivery evidence. Amber conversations begin with the unresolved concern and recovery plan rather than a general pitch.
Keep the Dashboard Working Under Pressure
Stress Point 1: More Clients, Less Scoring Time
At 8 active clients, a monthly scoring session takes about 60 minutes. At 15, the estimate rises to 90–120 minutes.
If delivery pressure causes the founder to rush or skip scoring, coverage weakens as client volume grows.
Protect the scoring session. At higher volumes, have each account manager score their own roster, then have the founder review and validate the scores.
Stress Point 2: Informal Monitoring After a Red Crisis
After a difficult Red intervention, a founder may track every interaction closely and feel the relationship is improving.
Keep the formal score grounded in observable signals, regardless of that impression.
If the next score moves to Amber or Green based on evidence, record the improvement. If it remains Red, continue the Red protocol.
Handle Client Health Scoring Edge Cases
A Client Has Been Green for 12+ Months
Continue the five-signal check monthly. A long Green history may support a renewal or expansion conversation, but tenure does not exempt the client from monitoring.
The Agency Has Project Clients, Not Retainers
At the project midpoint and completion, check results vs. scope, communication sentiment, and payment timeliness. If any two signals fall below the acceptable level, schedule a proactive check-in. Keep this check separate from the five-signal retainer score and its tier thresholds.
No Results Benchmark Was Set at Onboarding
Assign a provisional 2 to results vs. targets for the first pass, rather than treating it as evidence of satisfactory results. On the next call, ask: “I want to make sure we’re measuring the right things for you. Can we confirm what a successful month looks like by the numbers?” Document the agreed benchmark before the next scoring cycle.
Two Account Managers Score the Same Client Differently
Create a one-page guide defining observable evidence for 1, 2, and 3 on each of the five signals. When scores differ, use the lower score until the account managers review the evidence and resolve the discrepancy.
Complete the First Pass in Under 90 Minutes
Step 1, Client List: 5–10 minutes. List active retainer clients.
Step 2, Signal Data: 20–30 minutes. Gather payment, communication, delivery, results, and scope evidence.
Step 3, Score Clients: 15–20 minutes. Use the Client Health Scorecard in Toolkit 1 (PDF).
Step 4, Tier Actions: 15–20 minutes. Put every required Amber and Red intervention on the calendar.
Total: 55–80 minutes. The aim is to identify the first at-risk client and schedule action before the next billing cycle closes.
If Setup Stalls
“I don’t have reliable results data for some clients.” Assign a provisional 2 only where no benchmark was documented. Agree on a written benchmark with each affected client in the same week. Do not treat the first score as fully evidenced.
“My clients communicate verbally, not in writing.” Send a short post-call email: “Following up on today’s call: [three-sentence summary of what was discussed and agreed].” Record the response or lack of response as one communication data point; do not infer sentiment from silence alone.
“I have 14 active clients and scoring feels unmanageable.” Split the list by retainer value. Score the top 50% by revenue in week one and the remainder in week two, while keeping the monthly scoring cadence for both groups.
AI Velocity Prompt
Use this in Claude or ChatGPT with the client information you are permitted to share:
Assess client health for [client name] using the evidence below.
- Agreed targets and results for the last two months: [targets and results]
- Written communication from the last 30 days: [email and project threads]
- Call attendance, call notes, and report responses: [engagement evidence]
- Payment due dates and payment dates for the last 90 days: [payment history]
- Expansion, reduction, or pause discussions: [details or none]
Score each signal from 1 to 3 using the agency’s scoring criteria:
- Results vs. agreed targets
- Written communication sentiment across the full 30 days
- Engagement level
- Payment timeliness
- Expansion or scope reduction signals
Return the assessment in this format. Do not use a table.
Results vs. targets
- Score: [1–3 or insufficient data]
- Evidence: [one specific observation]
- Targeted check needed: [Yes/No]
Communication sentiment
- Score: [1–3 or insufficient data]
- Evidence: [one specific observation from the full 30-day period]
- Targeted check needed: [Yes/No]
Engagement level
- Score: [1–3 or insufficient data]
- Evidence: [one specific observation]
- Targeted check needed: [Yes/No]
Payment timeliness
- Score: [1–3 or insufficient data]
- Evidence: [one specific observation]
- Targeted check needed: [Yes/No]
Expansion/reduction signals
- Score: [1–3 or insufficient data]
- Evidence:A full-period review can reveal a communication pattern that a good call last Tuesday might overshadow. Check the AI’s proposed scores against the underlying records before assigning a tier.
Score inflation defeats the monitoring system when a reassuring impression replaces observable evidence. Keep each score tied to what the client did, not what the founder hopes it means.
Running This System in Your Current Condition
Contraction: Revenue Declining or Below $60K/Month
When time is constrained, use a monthly three-signal check instead of the full five-signal score:
Results vs. targets
Payment timeliness
Communication sentiment
Score each from 1 to 3, for a maximum of 9. In this compressed version, Red is below 4, Amber is 4–5, and Green is 6–9. Keep these thresholds separate from the full dashboard’s 15-point scale.
If payments slow across multiple clients at once, investigate whether they share a source of budget pressure rather than assuming each relationship has the same problem. Discuss options such as reduced scope, phased billing, or results-linked pricing where appropriate.
Track the share of active clients scoring Amber or Red. If it exceeds 30%, review delivery quality, client targeting, and market fit for a shared cause, not just individual account issues.
Stability: Revenue Steady at $60K–$150K/Month
Stable revenue does not show whether individual relationships are weakening. Keep the monthly score and weekly review in place so a cancellation is less likely to be the first clear signal.
A consistent six-month record in the Health-Score-to-Churn Predictor Calculator can reveal patterns in your own client base. Treat early patterns as evidence to investigate, not a calibrated prediction until you have enough history. At an average $5,000/month retainer, keeping a client for 12 additional months represents $60,000 in retained revenue; the dashboard creates opportunities to intervene, not a guarantee of retention.
Expansion: Revenue Above $100K/Month, Team Growing
As account managers take on clients, scoring consistency becomes the risk. Before a second account manager starts scoring, create a written guide with observable criteria for 1, 2, and 3 on each of the five signals.
Quarterly: Give account managers the same client case to score independently. Compare differences, resolve them against the evidence, and update the guide. Aim for scores within ±1 point across scorers.
For every Red designation: Have the founder review the evidence immediately and lead the intervention. Verification should happen inside the 48-hour response window, not delay its start.
The guide limits both unnecessary escalations and overly generous scores that hide a genuinely at-risk client.
The Delivery Dashboard in the Agency Operating System
I Spend All Day in My Inbox Answering ‘Where Is This’ Questions — Professional Project Governance creates the structured project data required for accurate client-health scoring. Use this when delivery data is scattered or unreliable.
One Unhappy Client Is Taking Up 80% of My Mental Energy — The Client Conflict Protocol guides recovery or resolution when a Red client does not stabilize. Use this when an intervention fails to repair the relationship.
High-Paying Clients Feel Ignored as We Get Busier — Strategic Account Management defines elevated relationship management for strategic clients flagged by health scores. Use this when high-value accounts turn Amber.
Catching Unhappy Clients Before They Cancel — The Feedback Engine captures concerns clients will state directly, complementing behavioral health signals. Use this when you need direct client feedback alongside score data.
Tracking All Client Projects Without Losing Your Mind — The Delivery Dashboard supplies the project-level data that health scores interpret into churn risk. Use this when project activity lacks a usable client-health view.
Where Are You in the Installation Sequence?
If project governance is not in place, start there. The health score needs structured delivery data.
If project governance is running and you have 6+ active retainer clients, install the Delivery Dashboard next.
Your Delivery Dashboard Fix Starts Now
At Week 8, you’ll be able to say:
“I can name the health score of every active retainer client. I know which relationships are Green, which are Amber, and which have required intervention in the last 60 days.”
“The last client cancellation was not a surprise. The health score showed deterioration six weeks before the exit, and an intervention was made. The client stayed.”
“The weekly 20-minute review is a fixed calendar block. It has not been skipped in eight weeks. I have a documented health trajectory for every client.”
Three time-boxed actions:
In the Next 30 Minutes
List every active retainer client.
For each, note when they last gave unprompted positive feedback: recently (3), not recently (2), or cannot remember (1).
Use this informal pre-score to decide which relationships to assess first. It is not part of the formal five-signal health score.
This Week
Run the full five-signal score for your top five clients by retainer value.
Use those scores as the first formal baseline, then score the remaining eligible clients.
Before Next Month
Add a recurring 20-minute weekly dashboard review to your calendar.
Use it to track new signals and open interventions between monthly scoring sessions.
Delivery Dashboard Progress Milestones:
Milestone 1: Every active retainer client has a current five-signal health score. No client assessed as “probably fine” without a score attached.
Milestone 2: Every Amber and Red client has a scheduled intervention. Zero at-risk clients without a next action on the calendar.
Milestone 3: First Amber intervention completed. The specific low-scoring signal was the agenda of the intervention conversation. The outcome — stabilized, still declining, or escalated to Red — is documented.
Milestone 4: Second monthly scoring cycle complete. Trend data is now available for every client. The two-point drop trigger has been applied to any client showing rapid deterioration.
Milestone 5: 90-day review complete. Health-Score-to-Churn Predictor has its first data cycle. At least one correlation between a specific score pattern and an actual outcome (retained or churned) is documented.
If you take one thing from each section:
The signals that precede client cancellation are behavioral, not verbal — they are measurable if a measurement system exists and invisible without one.
A client health score makes the state of every client relationship legible in a number — removing the subjective impression that allows deteriorating relationships to go undetected until cancellation.
The first health score pass produces the first at-risk client identification — the session ends when every Amber and Red client has a scheduled intervention on the calendar.
The clients who cancel without warning are rarely the loudest ones — they are the quietest, and their silence reads as satisfaction until the five-signal score reveals what is actually present.
Score inflation — giving clients higher health scores than the evidence warrants to avoid discomfort — is the single mechanism that defeats the entire monitoring system; anchoring scores to observable data points is the only fix.
But if you remember only one thing:
The Delivery Dashboard does not prevent all client churn. It eliminates the churn that happens because the founder had no system to see the warning signs that were already present. The founder who runs this system once has a detection capability. The founder who runs it monthly has a retention advantage that compounds with every scoring cycle.
Delivery Dashboard Checklist
Reference this before closing your first scoring session each month.
☐ Every active retainer client with 60+ days history has a current five-signal score
☐ Each score is anchored to observable data, not last call impression
☐ Every Amber client has a targeted check-in scheduled within 5 business days
☐ Every Red client has a founder-led call booked within 48 hours
☐ Weekly 20-minute review block is set as a recurring calendar event
A health score with no scheduled next action is a label, not a governance system — every Amber and Red client leaves this session with a calendar entry.
FAQ: The Delivery Dashboard System
Q: What exactly are the five signals in the Client Health Score?
A: The five signals are results vs. targets, communication sentiment, engagement level, payment timeliness, and expansion or scope reduction signals. Each is scored 1–3 based on observable behavioral data from the past 30 days. The maximum possible score is 15. The scoring is anchored to what clients do, not what they say on calls.
Q: How are the health tiers defined?
A: Green is a score of 10–15, indicating a healthy relationship that maintains the standard delivery cadence. Amber is 6–9 — at least one signal has deteriorated and proactive communication is triggered. Red is below 6, meaning the relationship is at risk and a founder-led intervention is required within 48 hours.
Q: Why can’t I rely on regular client calls to catch dissatisfaction early?
A: Calls produce only what clients choose to share. Clients who are dissatisfied often maintain cordial call behavior until they exit — they feel guilt, conflict-avoidance, and sunk-cost awareness, so they stay quiet. The five signals track behavior independent of what a client reports verbally: payment timing, revision count, engagement rates, results trajectory, and expansion language.
Q: How long does the first scoring session take?
A: The full first session runs in 55–80 minutes across four steps: listing active clients takes 5–10 minutes, gathering signal data 20–30 minutes, scoring each client 15–20 minutes, and scheduling tier actions 15–20 minutes. Using AI assistance for signal review can compress the data-gathering step to 20–30 minutes from a potential 60–90 minutes done manually.
Q: What is the weekly 20-minute review and why does it exist alongside monthly scoring?
A: The monthly score requires assembling full behavioral data across five signals — that happens once a month. The weekly review is a lighter governance check that catches single-signal flares between monthly assessments: a late payment this week, a terse email thread, a missed call. It is a fixed calendar block, not an optional check-in.
Q: What are the Proactive Intervention Triggers?
A: Three specific thresholds force action before the client raises a concern. First, any single signal dropping to 1 triggers a targeted conversation on that dimension regardless of the total score. Second, a two-point drop in total score within 30 days triggers an immediate mini-review even if the client is still technically Green.
Q: What is score inflation and how do I prevent it?
A: Score inflation is the gradual drift toward assigning clients higher scores than the evidence supports to avoid the discomfort of an Amber or Red designation. It is the primary failure mode that defeats the monitoring system.
Q: Can AI tools help with running the health score?
A: Yes. Pasting 30 days of email threads, call notes, project comments, and payment status into Claude or ChatGPT and asking for a 1–3 score with evidence per signal takes 20–30 minutes versus 60–90 minutes manually.
Q: What happens if a client has been Amber for three consecutive months?
A: Three consecutive months at Amber without returning to Green is a structural failure of the standard Amber protocol — the relationship has been managed without resolution for 90 days. The trigger rule requires escalating to the Red protocol regardless of current score.
Q: What if I have fewer than 6 active retainer clients?
A: The Delivery Dashboard is built for 6 or more active retainer clients. Below that threshold, direct relationship management is sufficient to maintain visibility across the full client base.
⚑ Found a Mistake or Broken Flow?
Spotted a math error, unclear framework, or broken link? Use this form to flag it — helps me keep the articles accurate and useful. Report a problem →
› More to Explore: Quick Navigation · Service Agencies
➜ Help Another Founder, Earn a Free Month
If the Delivery Dashboard just showed you which clients are closer to canceling than you realized, share it with one founder stuck watching retainers leave without warning.
When you refer 2 people using your personal link, you’ll automatically get 1 free month of premium as a thank-you.
Get your personal referral link and see your progress here: Referrals
Get The Delivery Dashboard Toolkit
You’ve read the system. Now implement it.
Premium gives you:
Ready-to-use PDF toolkit—every template, diagnostic, and formula pre-filled, zero setup, immediate use
Plug-and-play AI diagnosis sessions—drop into Claude, Gemini or ChatGPT, answer a few questions, save hours of guessing, get your exact next move
Audio key points—concentrated frameworks you can absorb in minutes, implement while you move
Unrestricted access to the complete library—every system, every update
What this prevents: Missing a $10,000–$20,000 churn incident that a 55-minute scoring session would have caught.
What this costs: $12/month.
Download everything today. Implement this week. Cancel anytime, keep the downloads.
Already upgraded? Scroll down to download the PDF, audio, and your AI session.
