When a leadership team asks me whether they should replace phone support with asynchronous chat, the real question is rarely about preference. It’s about money, capacity, and customer experience. The word “replace” hides a lot of trade-offs: cost per contact, resolution time, customer satisfaction, and the impact on workforce design. In this piece I’ll walk you through a pragmatic, repeatable way to calculate the exact break-even point — the moment when an asynchronous chat channel becomes cheaper (or cost-neutral) than voice — and the practical considerations you need to factor in beyond the raw numbers.
Why a precise break-even matters
High-level claims like “chat is 50% cheaper than phone” are tempting, but they’re dangerously incomplete. Costs depend on your team’s productivity, average handle times, occupancy targets, technology fees, and whether asynchronous workflows let agents handle multiple conversations concurrently. A precise break-even calculation gives you:
Key metrics you must collect
Before any calculation, gather accurate baseline data from your current phone support and projected/asynchronous chat setup. Here are the essentials:
Core formula: turning metrics into cost per contact
The simplest way to compare channels is cost per handled contact. Here are the two formulas I use:
Phone cost per contact
Phone_cost_per_contact = (Agent_hourly_cost / Occupancy) * (AHT_phone_hours)
Chat cost per contact (asynchronous with concurrency)
Chat_cost_per_contact = (Agent_hourly_cost / Occupancy) * (AHT_chat_hours / Concurrency)
Notes:
Worked example
Below I show a sample calculation for a mid-sized support team. Replace these numbers with your actual data to get a tailored result.
| Metric | Phone (baseline) | Asynchronous chat (projected) |
|---|---|---|
| Contacts per month | 20,000 | 20,000 |
| AHT (seconds) | 600 (10 minutes) | 300 (5 minutes active handling) |
| Concurrency | 1 | 3 (agent handles 3 active chats on average) |
| Agent fully loaded hourly cost | £18 | |
| Occupancy | 0.80 (80%) | |
| Platform & telephony cost per month (per agent equivalent) | £30 (phone trunking, CCaaS) | £50 (chat platform, messaging fees) |
Step-by-step compute:
In this example, asynchronous chat is substantially cheaper per contact. But remember: platform fees, automation, and differences in resolution rates change the picture.
Adjust for resolution rate and contact deflection
Cost per resolved issue matters more than cost per contact. If chat has a lower first-contact resolution (FCR), you might see more repeat contacts and higher overall cost. Adjust as follows:
Effective_cost_per_resolution = Cost_per_contact / FCR
If chat FCR = 70% and phone FCR = 85%, you’d divide each channel’s cost per contact by its FCR to compare on a per-resolution basis.
Model sensitivity: what moves the needle
Run a simple sensitivity analysis to see which levers have biggest impact:
Operational realities to include
Numbers alone don’t capture operational risks. Here are practical items I always check:
How I recommend running a pilot
A pilot is the fastest way to validate your model. I typically run a 6–8 week pilot with:
Use this live data to recalculate the break-even and test whether assumed concurrency and automation gains materialize. If not, iterate on workflows: add macros, micro‑templates, suggested replies, or AI-assistants to reduce AHT and increase FCR.
Quick checklist before you decide
Replacing phone support with asynchronous chat can deliver meaningful cost savings, but only if you test assumptions about agent concurrency, automation, and resolution rates. Use the formulas and steps above to build a transparent model you can share with finance and operations, then validate it with a focused pilot. If you want, I can create a simple spreadsheet template based on your real numbers — send me the inputs (AHTs, agent cost, occupancy, platform fees, concurrency) and I’ll return a populated model you can use in stakeholder conversations.
