AHT / TMO
Average Handling Time measures operator time consumed by a handled contact. In the lab it includes talk/work time plus after-contact work, allowing cost per contact and capacity to be estimated.
DOC ROI · CALL CENTER MANAGEMENT & SLA
Turn inbound service cost into measurable customer value and ROI.
Capture the customer voice, classify the reason for contact, measure AHT/TMO, First Call Resolution and Service Level, calculate capacity and operating cost, then convert recovered demand into relevant outbound value.
Academic foundation
Call-center intelligence is a service-level and value system
The call center starts as an inbound operating cost. The management objective is to control service quality and cost, recover the customer relationship and then activate relevant outbound opportunities without losing traceability.
01 · CaptureCustomer voice + operational time
02 · ClassifyKeywords → subcategory → category
03 · Service levelAHT/TMO · FCR · SLA · abandonment
04 · RecoverSeverity · owner · action · follow-up
05 · ValueCost avoided + outbound opportunity
06 · ROIEconomic impact vs required investment
First Contact Resolution
FCR measures the share of contacts solved without a repeat contact for the same issue. Low FCR increases customer effort and creates avoidable operating cost.
Service Level
The lab compares the percentage of contacts answered inside the agreed threshold with the target SLA. Service level is a commitment, not merely a dashboard metric.
Cost & capacity
Operators, productive hours, occupancy and hourly cost determine available capacity, FTE requirement, cost per contact and total handling cost.
Inbound → Outbound
Once the service issue is controlled, eligible contacts can become relevant follow-up opportunities: advice, retention, repurchase or cross-sell. Recovery comes before commercial activation.
ROI
The lab combines efficiency savings and expected outbound contribution, then compares that value with implementation and outbound-contact cost.
Decision logic: customer contact → structured classification → service-level evidence → recovery → cost/efficiency → responsible outbound activation → measurable ROI.
Functional laboratory
Call Center Intelligence Lab
Use the example only to understand the model. For your own work, download the template, complete the campaign and contact fields, and import the file back into the lab.
Empty workspace
Contacts0loaded records
Actual AHT / TMO—handled-contact average
Actual FCR—first-contact resolution
SLA attainment—within answer threshold
Cost / contact—handling cost
Estimated ROI—efficiency + outbound
01 · Campaign Setup i
Start by describing the operating conditions of the call-center campaign: how many people are available, how much productive time they have, what one hour of operation costs, and what service targets the organisation expects. These assumptions provide the benchmark for the rest of the lab. The raw contacts remain unchanged; Campaign Setup simply tells the model what “good performance” should look like and what resources are required to deliver it.
What you configureStaffing, productive time, AHT/TMO, FCR, Service Level, abandonment, investment and outbound assumptions.
What it producesA reference model for capacity, cost per contact, actual-versus-target gaps and ROI calculations.
Beginner ruleIf you are learning, load Don Espadín first. For professional use, replace every assumption with your own operation.
Campaign operating model
These values are the assumptions used for capacity, service-level cost and ROI calculations.
h
days
%
€
sec
%
sec
%
%
ROI & outbound assumptions
The inbound operation is primarily a cost centre. These fields estimate how efficiency improvements and responsible outbound activation can create measurable economic value.
€
%
%
sec
€
%
AHT / TMOTime consumed by handled contacts.
FCRResolution without repeat contact.
SLAAnswer performance vs commitment.
ROIValue created vs required investment.
02 · Raw Inbound Dataset i
This is the evidence layer of the lab. Each row represents one inbound contact exactly as it was captured by the operation: the customer’s words, channel, handling time, waiting time, First Contact Resolution and any existing category supplied by the source system. We keep this layer separate because good analysis must be traceable back to the original facts. Nothing in Taxonomy should overwrite these source values; Step 03 adds a classification model and Step 04 creates a new, derived dataset from this evidence.
InputYour completed DOC ROI template, your own CSV/XLSX file, a manual contact or the Don Espadín example.
PreserveOriginal customer language and operational measures. This is the audit trail behind every later conclusion.
Next stepUse Taxonomy to decide how these contacts should be grouped into repeatable business reasons for contact.
02 · Raw DatasetWhat happened? Original facts and customer language.
→
03 · TaxonomyHow should we classify the evidence consistently?
→
04 · Classified DatasetWhat does the contact mean and what operational fields are added?
Raw inbound dataset
Source evidence before cognitive classification. Scroll inside the table to inspect the full dataset; the source remains intact when the taxonomy is applied.
Raw data first.If your file already contains Category / Subcategory, the lab discovers those values and uses them to propose the initial taxonomy. If they are empty, DOC ROI proposes labels from the customer statement using the built-in keyword rules. In both cases the source fields remain visible so learners can compare “what came in” with “what the model produced”.
03 · Taxonomy · Configure the Classification Engine i
Taxonomy is the rulebook that converts many different customer phrases into a stable set of business reasons for contact. A category is the broad container, such as Logistics or Payment; a label / subcategory is the precise operational reason, such as Delivery delay or Duplicate charge. Select only the labels that make sense for your operation, add your own when necessary, or remove irrelevant ones. The lab does not change the classified dataset while you edit: the new configuration becomes operational only when you press Apply Taxonomy & Reclassify.
Why it existsWithout common labels, two agents can describe the same problem differently and management cannot aggregate demand, recurrence or cost.
How it classifiesExisting source classifications are respected when active; otherwise customer text is evaluated against the active detection keywords.
What to manageActivate, deactivate, add or remove labels. Then apply the taxonomy once you are satisfied with the classification rulebook.
Taxonomy configuration engine
Not configuredEdit the classification rulebook here; no downstream result changes until you apply it.
How this step works1) Import raw contacts. 2) The lab discovers existing Category / Subcategory fields or proposes labels from customer language. 3) Select, add or delete labels. 4) Press Apply Taxonomy & Reclassify. Only then is the classified dataset rebuilt.
Categories i0configured containers
Active labels i0used by classifier
Potential coverage i—before applying
Current coverage i—last applied result
Categories & labels
Categories are containers. The selectable chips are the operational labels/subcategories. The tiny × removes a label from the current taxonomy.
Add your own category
Create a business family and its first operational label. Detection keywords are used when the source file does not already contain that classification.
Classification rule: imported Category / Subcategory values are respected when their label remains active. Otherwise the text is evaluated against the active keyword rules.
04 · Classified Dataset · From Evidence to Decision-Ready Data i
This is where the difference between Step 02 and Step 04 becomes visible. The Raw Dataset tells us what happened; the Classified Dataset keeps those facts but adds a consistent interpretation that managers can aggregate and act on. After applying Taxonomy, each contact can receive keywords, category, subcategory, a possible secondary category, severity, recovery SLA and ownership. These new fields are the operational layer used by the SLA, ROI, outbound and reporting sections. If you change the taxonomy later, reapply it so this derived dataset is rebuilt with the new rules.
Preserved from Step 02Case identity, original customer statement, channel, AHT/TMO, waiting time, FCR and other source evidence.
Added in Step 04Keywords, category, subcategory, secondary signal, severity, recovery SLA and accountable owner.
Why it mattersOnly structured, decision-ready data can be grouped, routed, costed, prioritised and converted into repeatable management evidence.
Raw evidenceCustomer voice + operational facts. Never overwritten.
+
Active taxonomyControlled labels and classification rules selected by the user.
=
Classified evidenceOriginal facts + structured operational interpretation.
Classified dataset
Decision-ready output generated only after Step 03 has been applied.
Contacts0
Classified0
Unclassified0
Coverage i—
Waiting for classification.Configure and apply the taxonomy in step 03.
05 · Service Level Agreement & ROI Evidence i
Step 05 asks whether the operation is delivering the service promised and what the performance gap costs. Actual AHT/TMO, FCR, answer Service Level and abandonment are compared with the targets defined in Campaign Setup. Time and staffing are converted into contact capacity and operating cost, while the model estimates the value of closing AHT and FCR gaps. The outbound opportunity is kept separate and only starts after recovery, so the final ROI can distinguish efficiency value from commercial contribution.
Operational evidenceActual versus target AHT/TMO, FCR, Service Level and abandonment.
Economic evidenceHandling cost, cost per contact, capacity, required FTE and avoidable efficiency cost.
Interpret carefullyThe ROI is a teaching/business-case estimate based on the assumptions entered in Step 01, not an audited accounting result.
Service Level Agreement & ROI Evidence
Management deliverable: operational baseline, service commitments, cost model, improvement opportunity and outbound value case.
SLA commitments
| Metric | Actual | Target | Gap | Status |
|---|
Capacity & cost model
Handling cost = Σ(AHT seconds ÷ 3,600 × operator hourly cost)
Contact capacity = operators × hours/day × days × occupancy ÷ target AHT
Contact capacity = operators × hours/day × days × occupancy ÷ target AHT
Efficiency opportunity
Teaching assumption: AHT opportunity values the time difference between actual and target. FCR opportunity estimates the cost of repeat contacts that could be avoided if the target resolution rate were reached.
Inbound → Outbound opportunity
Recovery before revenue: an outbound action should only be activated after the original service need is controlled and when the follow-up is relevant to the customer context.
Estimated ROI—
Load data and apply campaign assumptions.
Management interpretation
DOC ROI · CALL CENTER INTELLIGENCE LAB
Service Level Agreement & ROI Evidence
No dataset loaded.
Actual AHT—
FCR—
SLA attainment—
Estimated ROI—
Handling cost—
Efficiency value—
Outbound revenue—
Outbound contribution—
Service commitments
- No data yet.
Economic case
- No data yet.
06 · Inbound → Outbound Activation i
The call center begins as an inbound service cost, but recovered demand can also reveal relevant commercial opportunities. This step identifies the contacts that are eligible for a follow-up only after the original need is controlled, then applies the outbound assumptions from Campaign Setup: eligible base, outbound handling time, conversion, average order value and gross margin. The result separates expected revenue, gross-margin contribution and the operating cost of making those outbound contacts.
GateRecovery, relevance and eligibility come before any commercial contact.
Value modelEligible contacts × conversion × order value creates revenue; gross margin converts revenue into contribution.
Cost modelOutbound AHT × operator hourly cost quantifies the additional effort required to activate the opportunity.
Inbound → Outbound Activation
The commercial opportunity starts only after the service need is controlled.
Outbound activation gate
Rule: no commercial activation before recovery. Relevance, permission and customer context come before conversion.
Economic opportunity
Activation steps
01 · RecoverClose the original service need.
02 · QualifyCheck permission, relevance and eligibility.
03 · SegmentUse customer value, reason for contact and context.
04 · OfferDefine the next best action, not a generic sales push.
05 · ContactSet channel, timing, owner and outbound AHT.
06 · MeasureTrack conversion, revenue, contribution and ROI.
07 · Management Reporting · Read the Complete Business Case i
The final step brings the complete operating story together. Read service quality first—AHT/TMO, First Call Resolution, Service Level and abandonment—then examine inbound cost and capacity, and only afterwards read outbound revenue, contribution and ROI. The classification charts explain why customers contact the operation, while Taxonomy Health checks whether the classification model is usable. Print this section when you need a single management record of the assumptions, operational evidence and economic result created by the lab.
Service objectiveAre we resolving demand at the expected time, quality and answer commitment?
Economic objectiveWhat does inbound demand cost, what efficiency value is available and what contribution can responsible outbound create?
Decision objectiveWhich demand drivers, service gaps and ROI levers should management prioritise next?
Management reporting
Final operational and economic view: service, efficiency, inbound cost, outbound value and ROI.
Service & efficiency KPIs
The indicators that define whether the operation is resolving demand efficiently and at the promised service level.
Inbound cost & capacity
Translate time and staffing into operating capacity and economic evidence.
Outbound value & revenue
Measure only the commercial activation that follows service recovery and customer relevance.
ROI bridge
Separate the return created by operational efficiency from the return created by outbound activation, then read the combined business case.
ROI logic. Efficiency ROI compares AHT/FCR efficiency value against the configured improvement investment. Outbound ROI compares expected gross-margin contribution against outbound operating cost. Combined ROI uses both value streams against improvement investment + outbound operating cost.
Demand & classification
Use the distributions to understand why customers contact the operation and where risk or workload is concentrated.
Category distribution
Severity distribution
Channel distribution
Top subcategories
Taxonomy Health
0