What is average handle time?
Average handle time (AHT) is the average length of a customer contact, from the moment an agent connects to the moment the agent finishes the work that contact created. Amazon Connect's metric documentation defines it as the average time, from start to finish, that a contact is connected with an agent, including talk time, hold time and after contact work. Microsoft's Dynamics 365 documentation describes the voice version as total talk time, plus total hold time, plus total wrap-up or after-call work, divided by the number of calls handled. In plain terms, AHT equals talk time plus hold time plus after-call work, divided by calls handled. For example, a team that handled 400 calls in a week with 2,000 minutes of talk, 400 minutes of hold and 600 minutes of after-call work would have an AHT of 3,000 divided by 400, or 7.5 minutes. Platforms differ in the details, so compare AHT only against your own past numbers, measured the same way.
Amazon Connect describes after contact work (ACW) as the time an agent spends on a contact after the conversation ends, which some businesses call call wrap-up time. Notes, follow-up emails and case updates all land here.
Where does phone support time actually go?
On calls about something on a screen, much of the clock goes to building a shared picture of that screen rather than fixing anything. Common time sinks:
- Describe-and-guess loops. The agent asks what the customer sees, the customer describes it in their own words, the agent guesses, and the cycle repeats.
- The wrong screen. One misread instruction sends the customer into a different menu, app or website, and the next few minutes go to finding the way back.
- Version and device differences. The agent's instructions match their own phone or the training material, not the customer's model, software version or text size.
- Paraphrased error messages. “It says something went wrong” leads to more questions than the exact error text would.
- Repeat calls. A call that ends with “I think it worked” but did not shows up later as a second full call, which never appears in the first call's handle time.
- After-call notes. The agent reconstructs from memory what the customer saw and did.
How does screen sharing cut handle time?
Screen sharing does not make anyone talk faster. It removes the part of the call spent describing a screen. Once the agent can see the customer's screen, directions get shorter and more specific, and the agent can confirm each step by watching rather than asking.
| Time sink | Without seeing the screen | With screen sharing |
|---|---|---|
| Finding the right screen | Several rounds of describing and guessing | The agent sees where the customer is and gives one direction |
| Device and version differences | Instructions do not match what the customer sees | The agent adapts to what is actually on the screen |
| Error messages | The customer paraphrases; the agent asks again | The agent reads the exact message |
| Confirming the fix | “Does it work now?” | The agent watches the result happen |
| After-call notes | Reconstructed from memory | Written from what the agent saw |
It does add a step: sending the invitation and waiting for the customer to approve. On phones, the first session may also involve installing an app. That is why screen sharing pays off on calls that stall, not on every call.
Tactics that reduce handle time
- Pick the call types where seeing helps
Look for calls that already involve a screen and run long: app sign-in problems, account setup, enrollment or application portals, e-signatures, device settings.
- Offer the link early
Set a simple trigger, such as the second time an agent has to ask “what do you see now?” Waiting until the call is already frustrating wastes the minutes you meant to save.
- Script the invitation
Give agents one or two sentences that explain what the customer will receive, that they approve it themselves, and that they can pause or stop at any time. Consistent wording builds trust and cuts hesitation.
- Plan for private steps
Tell customers to pause sharing before passwords, codes, payment details or Social Security numbers. Planning for this keeps those moments short instead of awkward.
- Guide with landmarks
Even when watching, say exactly where to tap: the position, the color and the label. See how to explain technical steps over the phone.
- Confirm the fix on screen
Before ending the call, watch the customer complete the task once. This is the cheapest protection against a repeat call.
- Write notes from what you saw
Note the exact screen, error message and fix, so recurring problems become training material.
What can go wrong when you chase lower handle time?
Handle time is a cost measure, not a quality measure. Pushing it down by rushing customers, transferring hard calls or skipping confirmation tends to produce repeat contacts, which cost a full call each. Track resolution alongside AHT.
- Counting the wrong calls. If screen sharing is used mostly on the hardest calls, those calls may still run longer than average. Compare like with like.
- Ignoring the first-time install. A customer installing an app for the first time adds setup to that call. Later calls with the same customer are faster to start.
- Losing the customer's trust. An unexplained link can look like a scam. Explain it before you send it.
How do you measure the impact before and after?
- Choose and tag call types. Pick two or three call types to test and make sure agents tag them consistently.
- Record a baseline. Capture AHT for those call types over several weeks before the change, split into talk, hold and after-call work.
- Track the companion metrics. Measure repeat contacts within a window you define (for example, the same customer calling about the same issue within 7 days), transfers, escalations and customer satisfaction.
- Run a pilot. Have some agents use screen sharing on the chosen call types while others continue as usual during the same weeks, so seasonal swings affect both groups.
- Measure the sharing step itself. Track how often agents offer a link, how often customers accept, and how long it takes from offer to seeing the screen.
- Allow a learning period. Compare the pilot's later weeks, not just the first few days.
- Judge by time per resolved issue. Total handle time across all contacts for an issue, divided by the issues actually resolved, captures both shorter calls and fewer callbacks.
| Metric | What it tells you | Watch for |
|---|---|---|
| AHT, split into talk, hold and after-call work | Where the minutes go | One part can rise while another falls |
| Repeat contact rate | Whether fixes stick | Short calls that cause callbacks |
| First contact resolution | Resolution from the case side | Definitions differ by platform |
| Transfers and escalations | Whether the first agent can finish the job | Hard calls pushed elsewhere |
| Share acceptance rate | Whether customers say yes | Invitation wording that causes hesitation |
| Customer satisfaction | How the call felt | Small sample sizes |
Where does GuidedView fit?
GuidedView is built for the calls described above. While the customer is on the phone, the agent sends a link by text or email, the customer approves, and the agent sees the customer's iPhone, iPad, Android or computer screen live in a browser. The customer does the tapping; the agent talks them through it on the same call. Customers can pause before private steps, and sessions are not recorded.
We do not publish handle time results for GuidedView. Measure it on your own calls with the method above. Each GuidedView account is $29.99 per month, customers need no account, and there is nothing to install on your website or deploy through IT, which makes a small pilot easy to run. See pricing or GuidedView for call centers.
Questions
What is a good average handle time?
There is no universal number. AHT depends on the call type, the industry and how your platform measures it. A password reset and a benefits enrollment will never share a target. The useful comparison is your own baseline for the same call type, measured the same way, before and after a change.
Does average handle time include hold time?
In common definitions, yes. Amazon Connect includes talk time, hold time and after contact work, and Microsoft's Dynamics 365 voice metric adds talk, hold and wrap-up time, then divides by calls handled. Check your own platform's definition, since details vary.
Can screen sharing make calls longer?
Sometimes. On very short calls, sending a link adds a step, and a customer installing an app for the first time adds setup. Agents may also use the clearer view to fix more while they are on the call. That is why you should measure by call type and track repeat contacts, not just AHT.
Should agents be measured on AHT alone?
No. AHT alone rewards ending calls quickly, which can drive repeat calls. Pair it with first contact resolution or repeat contact rate and customer satisfaction, so a shorter call only counts as a win when the problem stays solved.
Sources
Third-party details were checked against these pages as of November 2, 2026. Products change, so confirm current pricing and features with each vendor.