Quick summary

• After-call work (ACW) accounts for 15-20% of total handle time at most contact centers. Automating it is the fastest win.
• Agents switching between 4+ applications per contact add 15-30 seconds per interaction, according to Genesys research on desktop fragmentation.
• Knowledge base search failure is the second leading cause of long handle times. The fix is usually tagging and structure, not more content.
• Setting AHT targets without FCR guardrails consistently backfires. Teams that reduced AHT while tracking FCR together improved both metrics.

---

Contact center managers are under pressure to reduce handle time. The math is straightforward: lower AHT means more contacts handled per agent per hour, which means lower cost per contact.

The methods used to get there are not always straightforward. Some of them work. Some of them lower AHT while destroying FCR and CSAT, which costs more money than the AHT reduction saved. Here’s what actually moves the number without causing downstream problems.

Start with after-call work

After-call work is everything agents do after a contact ends: updating CRM records, filing disposition codes, sending follow-up emails, scheduling callbacks. It’s invisible to customers but shows up fully in AHT.

At most contact centers, ACW is 15-20% of total handle time. For a team with a 6-minute average, that’s 54-72 seconds per contact just on wrap-up.

The good news: most of this can be automated or significantly reduced.

Modern platforms like Genesys Cloud CX and NICE CXone can auto-populate disposition codes based on call categorization. Five9 has a “disposition timer” that limits how long agents can sit in ACW before the system auto-closes. Dialpad’s AI transcription can generate call summaries that update CRM fields automatically.

If you’re on a platform that doesn’t support these features natively, Salesforce and HubSpot both have integrations that can log call details via API without manual agent input. The setup takes a few weeks but the ROI is fast.

Realistic expectation: reducing ACW from 75 seconds to 30 seconds is achievable within 90 days at most contact centers. That’s 45 seconds per contact, which compounds significantly across a team.

Fix desktop fragmentation first

The second biggest time sink is agents switching between applications. The average contact center agent uses 4-6 applications per contact: the CRM, the ticketing system, the knowledge base, the payment or order system, and sometimes a separate call tool.

Each application switch takes time. Genesys published research in 2024 showing that agents accessing 5+ systems per contact spent an additional 15-30 seconds per contact compared to agents who could access the same information from 1-2 systems.

At 150 contacts per agent per day, that’s 37-75 minutes of time lost to application switching per agent per day.

Two ways to address this:

First, agent desktop unification. Salesforce and Zendesk both offer unified agent workspaces that pull together CRM data, ticket history, knowledge base, and call controls into one view. The integration work is real but the payoff is measurable.

Second, browser extension approach. If full integration isn’t feasible, a single browser tab with pinned tabs for each tool is faster than hunting through the taskbar. Sounds simple. Most teams haven’t done it.

Knowledge base search failure

When agents can’t find the answer in the knowledge base quickly, they do one of three things: put the customer on hold to ask a colleague, make up an answer (which creates repeat contacts), or spend extra time on the call while they search.

The most common knowledge base problem is not a lack of content but a lack of findability. Articles exist but agents can’t locate them during a live contact because:

• Articles are tagged with internal terminology that doesn’t match how customers describe their problem
• The search function requires exact keyword matches rather than fuzzy search
• Article titles describe the solution rather than the customer’s question

Freshdesk, Zendesk, and ServiceNow all have tagging systems. The fix requires going through your top 50 most-searched terms and ensuring each has a corresponding, accurately-tagged article. This takes a few days of work, not months.

One test: record 10 calls where agents put the customer on hold. Listen for what the agent searched in the knowledge base. Compare that to what the article was actually titled. You’ll find the gap immediately.

What doesn’t work: AHT targets without FCR guardrails

This one is worth stating directly because it’s the most common mistake.

When you set AHT targets and track AHT as a primary metric without simultaneously tracking FCR, agents optimize for AHT. They close tickets faster. They transfer more. They give partial answers that don’t fully resolve the issue. They get the customer off the phone without solving the problem.

The customer calls back. That second contact now adds to both AHT (another call) and FCR (which goes down). Net result: lower AHT per individual contact, higher total contact volume, lower CSAT.

Teams that reduced AHT successfully almost always tracked FCR alongside it and set a minimum FCR floor. If FCR dropped below a threshold (usually 70%), AHT targets were temporarily deprioritized until FCR recovered.

Tools that help by call volume

Under 500 contacts per day:

• Freshdesk Contact Center ($0-$69/agent/month) with connected knowledge base
• Dialpad ($15/user/month) with AI transcription for ACW reduction

500-2,000 contacts per day:

• Genesys Cloud CX ($75/agent/month) for unified desktop and ACW automation
• Talkdesk ($85/agent/month) with CRM integration and AI assist

Over 2,000 contacts per day:

• NICE CXone (custom pricing) with AI-assisted disposition
• Genesys (enterprise tier) with full digital transformation features

Frequently asked questions

What is a realistic AHT reduction target for the first 90 days? 10-15% is achievable for most teams if you address ACW and knowledge base search in that window. More than 20% in 90 days is possible but usually requires technology investment (unified desktop, AI transcription).

How do I convince leadership to stop using AHT as the primary metric? Show the correlation between AHT and repeat contacts. Pull the data for your top 10 agents by AHT and compare their FCR. If the lowest-AHT agents have the lowest FCR, that’s your argument. If they have the highest FCR, you have a genuine efficiency model worth spreading.

Does AI really reduce handle time? For specific use cases, yes. AI that generates call summaries and populates CRM fields automatically saves 30-60 seconds of ACW. AI that surfaces knowledge base articles during a call (rather than requiring agents to search) saves search time. The claims about AI handling entire contacts without agents are real in some contexts but require significant setup and ongoing tuning.

What’s the fastest win if I have a small team and no budget for new tools? Fix your knowledge base tagging and structure. It costs nothing but time. Identify your top 20 most common customer issues and make sure each has a clearly titled, easily findable article. Track whether agents open the knowledge base more often in the next 30 days.

How long should agents be in ACW after a contact? Industry norm is under 60 seconds for straightforward contacts. Complex contacts (order changes, technical escalations) may legitimately need 2-3 minutes. If average ACW exceeds 90 seconds, it’s worth investigating what agents are doing during that time.