capacity ®
Loading your projects…

💬 ROI Guide

Use this with your customer to fill in the model together

Contact Centre Operation
Contacts per month
The total inbound contact volume across all channels — voice, chat, email and SMS — every month.
"How many contacts does your team handle per month across all channels?"
Typical ranges: 20K = small operation · 100K = mid-market · 300K+ = large enterprise. If they're unsure, ask for their WFM system report — don't let them guess from memory.
Average handle time (AHT)
The live contact time — from when the agent picks up to when the customer disconnects or the chat ends.
"What's your average handle time per contact? Does that figure include post-call wrap or just the live time?"
Typical: 4–6 min for simple contacts, 8–12 min for complex. Always ask whether wrap is included — many customers quote AHT without it.
Post-contact wrap time
The admin after each contact — writing notes, updating CRM records, setting dispositions — before the agent is ready for the next contact.
"After a contact ends, how long does the average agent spend on admin before they take the next one?"
Typical: 1–4 minutes. Most agents and managers underestimate this — it's often higher than they think.
Handling hours per day
Productive contact-handling hours per agent per day, after shrinkage is removed — breaks, training, team meetings.
"Of an agent's shift, how many hours are they actually available on the queue taking contacts?"
Typical: 4.5–5.5 hours. Total shift minus 45–60 min breaks, minus 30 min team meetings, minus any training time.
Repeat contact rate
The percentage of contacts that are a customer calling back about the same unresolved issue.
"Do you track how often a customer contacts you more than once about the same issue? What's your repeat contact rate?"
Typical: 15–25% for a well-run operation. High repeat rate = strong FCR improvement opportunity in the model.
Cost Base
Fully loaded agent cost
The total annual cost of one agent — salary plus all employer costs. Not just the salary.
"What's the average annual cost of one agent, fully loaded? That includes salary, employer NI, pension, benefits and any desk or equipment costs."
Typical: £28–34K fully loaded in UK contact centres · $45–55K in the US. Most customers quote salary only — add 25–35% for on-costs.
Annual attrition rate
The percentage of agents who leave each year — the single biggest hidden cost in most contact centres.
"What's your annual agent turnover rate? How many agents do you lose per year as a percentage of your headcount?"
Typical: 25–45% in UK front-line contact centres. If they say below 20%, check whether they're measuring resignation only — involuntary exits add several more points.
Replacement cost per agent
The cost of filling one vacancy — recruitment fees or job board costs, interview time, temporary cover.
"What does it typically cost to replace one agent? Include recruitment fees if you use agencies, and management time spent in the process."
Typical: £3,000–8,000 depending on whether they use agencies. £4,500–5,500 is a safe starting point if unsure.
New hires per year
Total annual hiring — both replacing leavers and any headcount growth.
"How many new agents do you hire per year in total — both to replace people who leave and any growth in the team?"
Onboarding cost per agent
Getting one new agent from day one to full productivity — training materials, trainer time, and the gap in output while they ramp.
"What does it cost to get a new agent from day one to fully handling a normal workload? Consider training time, materials and the difference in productivity while they're finding their feet."
Typical: £3,000–6,000. Onboarding time is usually 4–12 weeks for contact centre roles.
QA team size and cost
How many people are dedicated to manually reviewing contacts for quality — and what they cost.
"How many people are focused purely on quality assurance — listening to calls or reading chats and scoring them? What's their annual cost each?"
Why this matters: A typical operation has 3–8 QA analysts manually reviewing 1–3% of contacts. That's the contrast — "right now your team is sampling 2% manually. Capacity scores 100% automatically, every contact, every time."
Sector-Specific Value Drivers
These inputs are specific to the customer's industry sector. Each one captures a value lever unique to their business — churn, claims, returns, escalations — that sits on top of the universal cost drivers. Fill these in as you discuss the customer's specific priorities.
AI Impact Assumptions
Automation rate
The percentage of contacts fully resolved by AI — no human agent involved at any point. This is the single most important assumption in the model.
"If we look at your most common contact types, which ones could realistically be fully resolved without an agent? Walk me through your top five."
Build it bottom-up: if 35% of contacts are order status and those can be automated, you're already at 35%. Add password resets, balance queries, appointment scheduling. Arrive at the number from use cases, not a headline figure. Typical first year: 35–55%.
Handle time reduction
For contacts that still reach an agent, how much shorter does each one become when the answer is surfaced in real-time — no searching, no tabbing between systems.
"When agents need to look something up — a product spec, a policy detail, a procedure — how long does that take on average? That's the time real-time assist recovers."
Typical: 10–20%. 15% is a defensible conservative figure. Every percentage point, across all assisted contacts, compounds significantly.
Wrap time reduction
How much of the post-contact admin can be automated — the AI writes the summary, updates the CRM and sets the disposition while the agent is ending the contact.
"How much of your agents' wrap time is writing notes or updating systems? That's the part we automate from day one."
Typical: 40–60% reduction. This is one of the more concrete and deliverable promises in the model — wrap automation is relatively straightforward to implement.
FCR improvement
How much does first contact resolution improve — reducing the percentage of contacts that generate a callback because the issue wasn't resolved.
"If every agent had the right answer surfaced to them in real time, how many of those repeat contacts do you think would just stop happening?"
Tip: connect this directly to their repeat rate input. If repeat rate is 22% and you improve FCR by 25%, you remove 5.5 points of volume. That's a number you can do out loud in the meeting.
Attrition reduction
How much does agent turnover decrease when the job is demonstrably less frustrating — better tools, faster answers, fewer angry repeat contacts.
"Do you have any data on your exit interview themes? If the job was less stressful and agents felt better equipped, what do you think that does to retention?"
Typical: 20–30% reduction. This is the assumption customers challenge most — frame it as conservative and lead with the exit interview evidence rather than the number.
QA automation
What percentage of manual QA scoring is replaced by automated 100%-coverage analysis.
"If every contact was automatically scored against your QA framework, how much of your QA team's time would that free up for coaching rather than scoring?"
Typical: 70–85%. QA analysts shift from reviewing contacts to acting on the insights — a significant upgrade in how they spend their time.
Investment & Realisation
Capacity annual platform cost
The annual subscription — this syncs automatically from your quote if you've built one in the Quote Builder.
If this shows a sector estimate: go to the Quote Builder first, build the commercial for this customer, then come back. The licence cost drives the payback calculation — use the real number.
Implementation cost
Year-one professional services, integration work and knowledge onboarding. This is a one-off — it only appears in Year 1.
"We'll need to factor in the implementation investment in year one. This covers integration work, professional services and getting your knowledge base into the platform."
Internal delivery effort
The customer's own people's time on the programme — their project owner, IT resources and subject matter experts.
"You'll want to factor in your own team's time too — a project lead, IT support for integrations and a few subject matter experts reviewing AI responses. Typically 2–4 people for 3–6 months."
Important: never let this be zero. It's never zero. Customers who underestimate internal effort tend to have harder implementations.
Year-one benefit ramp
Benefits don't arrive at 100% on day one — this captures the gradual build as adoption increases and the AI learns the operation.
"We model a first-year ramp to keep this conservative. We're not assuming you get full benefit from month one — it builds as adoption increases."
Typical: 70–80%. Below 70% can undersell the case; above 85% can feel optimistic to a CFO. 75% is the number most customers accept without challenge.
Year 3 growth
Expected increase in benefit volume from Year 2 to Year 3 — reflects contact volume growth in the customer's business.
"Is your contact volume likely to be higher in Year 3 than today? If so, by roughly what percentage?"
Typical: 5% for a stable operation · 10–15% for a growing business. This affects the three-year cumulative figure meaningfully — adjust it to the customer's own forecast.
Capacity Capacity Builder

Your AI-powered deal toolkit

Build a quote, model the ROI, and generate a client-ready proposal — all in one place. Follow the three steps below.

Admin
1
Quote Builder
Build the commercial
  • → Pick a tier by agent seat count
  • → Enable AI Agent channels (Chat / Voice / Email / SMS)
  • → Add RTAA, CI, integrations and add-ons
  • → Apply Managed Services — base rate + 2% per solution bucket
  • → Discount within your authority (10% self-serve, 30% manager)
Watch: Core = $20K · Pro = $50K · Enterprise = $75K. Use agent seats as the primary sizing measure, not employees.
2
ROI Calculator
Build the business case
  • → Select the customer's industry sector
  • → Adjust headcount and volume to match the deal
  • → Platform cost syncs automatically from your quote
  • → Review value drivers across cost, revenue and risk
  • → Choose Conservative / Expected / Ambitious scenario
Watch: Scale the benchmarks to the customer's actual headcount before presenting. The sector defaults assume a specific size.
3
Proposal Builder
Generate the document
  • → Click Assemble Proposal to import quote + ROI data
  • → Add client name, logo, challenges and contact details
  • → Edit the executive summary and commercial sections
  • → Preview the full proposal before exporting
  • → Export as PDF or share via the deal package export
Watch: Always send ROI data from the ROI tab first — it populates the business case section of the executive summary.
Worked example — Pro deal, two solutions
From the companion guide
Line item Calculation Amount
Platform fee — ProFixed$50,000
1× inbound integration (read/ingest)~$10K each$10,000
AI Agent — Voice (300K min)300K min @ $0.18$54,000
CI Voice (24K hrs)24K hrs @ $1.75$42,000
ARR (before Managed Services)Sum of above$156,000
Managed Services — Pro base 15% + 2% (CI bucket)$156K × 17%~$26,520
Total deal (before discount)$182,520
The +2% MS rule: AI Agent (Voice) = 1 bucket. CI = 1 bucket. Total = 2 buckets → Pro base 15% + 2% extra = 17%. You can have many channels inside one bucket at no extra % — Voice + SMS + Chat is still one bucket.
Keyboard shortcuts
Save current section ⌘S
Export PDF (section-aware) ⌘P
Next tab in sequence ⌘→
Previous tab ⌘←
Discount approval thresholds
Rep — self-serve ≤ 10%
Manager — Mike / Neil / Dana ≤ 30%
Executive — Karaline > 30%
Tie discounts to a signature date. Capture the reason on the service order.

Return on investment analysis

Capacity · AI Value & ROI Calculator
Build the Business Case
Select sector
Benefit realisation stance Scales every value driver
Contact centre operation▼
How to discuss this section with your customer ▼

Start here to establish the scale of the operation. These numbers define everything that follows — the bigger the volume and the longer the handle time, the larger the cost base and the larger the opportunity.

"Before we get into the numbers, can you give me a sense of the scale? How many contacts per month across all channels, and roughly how long does the average one take?"

💡 If they don't know the exact numbers, estimates are fine — the model is directional. You can always refine it later. Don't let uncertainty about precise figures stop the conversation.

Cost base▼
How to discuss this section with your customer ▼

The cost base section is where you establish what this operation actually costs to run today — before any efficiency gains. These numbers are the denominator in the ROI calculation, so getting them right matters.

"What's the fully loaded annual cost of one agent? I mean salary plus employer NI, pension and benefits — not just the salary figure."

The attrition question often surprises people. Most leaders know it's high but haven't calculated the full cost. Walk them through the maths: if they have 200 agents at 30% attrition, that's 60 leavers a year at £5K each to replace — £300K before you've automated a single contact.

💡 QA team size is a powerful moment in the conversation. Ask them what percentage of contacts their QA team reviews manually. When they say 2–3%, respond: "Capacity reviews 100%. Every contact, every time, automatically."

Sector value drivers▼
How to discuss this section with your customer ▼

These inputs are specific to the customer's sector — the value drivers that go beyond contact handling costs. They capture the strategic levers: churn prevention, claims handling, returns reduction, SLA delivery. These are often the numbers that resonate most with a CFO.

"Beyond the contact handling costs, what are the big-ticket problems this operation creates or solves for the wider business?"

💡 Let the customer tell you which drivers matter most to them before you show the numbers. The ones they bring up unprompted are the ones they'll defend in an internal business case.

AI impact assumptionsWhat Capacity changes▼
How to discuss this section with your customer ▼

This is where you move from "what the operation costs today" to "what it could look like." These assumptions are the ones a CFO will scrutinise, so build them from the customer's own use cases rather than starting with a headline number.

"Rather than starting with a percentage, let's work through your most common contact types. Which of your top five contact reasons could realistically be fully resolved without an agent?"

Build automation rate bottom-up: if 35% of contacts are order status and those can be automated, you're already at 35%. Add password resets, balance queries, appointment scheduling. Arrive at the number from real use cases — the customer owns that number, not you.

💡 The handle time and wrap reductions are easier to defend than automation rate because they're about contacts that still reach an agent. Frame them as: "For every contact that does still reach your team, they spend less time hunting for information and less time on admin afterwards."

Investment & realisationLicence, delivery, ramp▼
How to discuss this section with your customer ▼

The investment section closes the loop — it puts the cost of Capacity into the model so the customer sees the net picture, not just the upside. Be honest about the full investment including one-off costs and internal effort.

"The model factors in three cost components: the annual platform fee from your quote, the implementation cost in year one, and your own team's time on the programme. Let's make sure all three are realistic."

The ramp rate is important to get right — it manages expectations. Saying 75% of benefits in year one is conservative and credible. Saying 100% looks like a sales pitch. The customer will use this number internally when building their business case.

💡 Never let internal effort be zero. A customer who enters zero here will struggle with their implementation and blame the product. Better to show a real number and demonstrate that the return still works.

MethodologyHow each line is calculated▼
Year one net benefit
—
Expected case
— after all Capacity costs
—
Benefit multiple
—
total benefit ÷ total year-1 cost
Payback
—
from go-live, at run-rate
Cost per contact
—
before → after

Where the value comes from

—
Largest contributorSmallest
Gross annual benefit—
Cost reduction
—
Revenue & retention
—
Risk & compliance
—
Three-year business case▼
 Year 1Year 2Year 3
Benchmarks behind these assumptionsCapacity & Creovai deployments▼
    All scenarios at a glance
    Saved Scenarios
    Directional model only. Built from operating inputs and benchmark impact ranges from Capacity and Creovai deployments. Not a quote or forecast. Validate every assumption before it goes to a customer.
    ✓ Saved

    Proposal Builder

    Each project holds one proposal, built from your quote and ROI. Configure the Quote and ROI tabs, then click Assemble Proposal in the sidebar to generate it.

    Rename Proposal

    Unsaved changes

    This project has changes you haven't saved.

    Discarding is permanent — anything you assembled or edited since the last save will be lost.

    New project

    For your projects list. Only you see this.
    Appears on the proposal and ROI covers. Logo and contact details come later.

    Client details

    Used on both the proposal and the ROI document.

    Sits under the client name on the cover. Optional.
    Leave blank until you know the recipient.
    —from your profile
    The cover version is rendered white, so single-colour marks reproduce best.