Call drivers
Definition
Call drivers: types, examples, and how to manage them
Call drivers are the specific reasons customers contact a support team, whether by phone, chat, email, or social message. Every driver gets tagged against a fixed taxonomy at wrap. A well-run call-driver programme turns raw contact volume into a diagnostic tool that tells operations exactly what to fix upstream to lower cost per contact and lift CSAT.
Most contact centres track call drivers in a taxonomy, a 15-to-40-line list of reasons like “billing dispute,” “password reset,” or “delivery delay,” that agents tag every interaction against.
The taxonomy is the input for every workforce plan, self-service investment, and product feedback loop the operation runs. Without it, the centre operates blind: more headcount solves symptoms, never causes.
The discipline emerged at telco and utility call centres tracking IVR selections in the 1990s. It’s now standard practice at every BPO of scale, and increasingly at in-house support teams under a Chief Customer Officer, where the taxonomy feeds directly into the product roadmap.
Key takeaways
- A call driver is the specific reason a customer contacts support, captured through a 15–40 line taxonomy tagged on every interaction.
- Drivers split into external triggers (seasonality, marketing, outages) and internal ones (product launches, policy changes, service defects).
- Top-5 drivers typically account for 60–80% of total volume in most operations.
- Managed drivers feed the product roadmap; unmanaged ones inflate headcount indefinitely.
- Deloitte’s 2024 survey found operations tracking driver-level data cut cost per contact 15–25% within 12 months.
How it works
Call drivers work as a diagnostic layer on top of raw contact volume. Every interaction — voice, chat, email, social — gets tagged against a fixed taxonomy at wrap. Aggregated over a week, the tags surface which upstream problems generate the most inbound cost.
A working taxonomy has three columns: the driver itself, its parent category, and the root cause the ops team suspects. Agents pick one driver from a dropdown at call-close. QA reviews a 5–10% sample weekly to catch tag drift.
The taxonomy splits into two big families, external and internal, because they need different fixes:
| Family | Example drivers | Fix owner |
|---|---|---|
| External (market, event) | Seasonal spike, marketing campaign, economic shift, weather event | Marketing, forecasting, WFM |
| Internal (product, ops) | Product launch, service outage, policy change, defect | Product, ops, comms, engineering |
According to Deloitte’s 2024 Global Contact Center Survey, operations tracking driver-level data see cost per contact fall 15–25% within 12 months of a taxonomy refresh — driven almost entirely by shifting the top-3 drivers to self-serve.
Tag hygiene matters as much as the taxonomy itself. A 30-line list with disciplined tagging beats a 200-line list nobody uses. Most operations cap tagging at 30 drivers to keep it usable during a live call.
Examples
Real call-driver programmes vary by industry, but the top-5 categories usually account for 60–80% of volume: a Pareto pattern that repeats across retail, telco, fintech, and airlines. Four concrete examples show the shape.
Amazon (US, retail). Return-related drivers routinely account for 35–40% of contact volume in Q4. Amazon shifted “where is my return?” almost entirely to self-serve in 2019, cutting agent-handled return contacts double-digits year-on-year, per its investor day disclosures.
Southwest Airlines (December 2022). A weather-triggered scheduling meltdown drove more than 16,000 flight cancellations over ten days. Call volume peaked at roughly 11,000% of a normal weekday, per DOT enforcement filings, a textbook external-driver spike no forecasting model could absorb.
Klarna (2024). The buy-now-pay-later firm publicly disclosed that AI chatbots handle work equivalent to 700 agents, covering the top-5 self-service-eligible drivers (payment status, refund query, dispute intake, plan change, credit availability). CEO Sebastian Siemiatkowski said the shift saved US$40 million a year.
BT (UK, telco). Network-outage drivers spike when infrastructure fails. A controlled example: the January 2022 999 emergency-call outage, which drove sustained overflow into commercial call centres. BT publishes quarterly driver-mix data through Ofcom’s complaints reporting.
The pattern across all four: the driver-mix report doesn’t just size next month’s roster — it names which upstream team owns the fix.
Related terms
Call drivers sit inside a wider contact center taxonomy. These are the closest neighbours:
- Call center: the voice-only forerunner where driver-taxonomy discipline first emerged in the 1990s.
- First-call resolution (FCR): the metric most sensitive to driver-mix, so poorly managed drivers tank FCR fast.
- Average handle time (AHT): reveals which drivers are burning agent minutes and should be scripted or automated.
- Customer satisfaction: the outcome variable driver management is trying to move.
- Business process outsourcing (BPO): the delivery model that made driver-taxonomy a contractual requirement.
- Key performance indicator (KPI): driver-mix reports show up alongside AHT, FCR, and CSAT in every KPI dashboard.
FAQ
What’s the difference between a call driver and a call type?
A call type is a coarse channel/category label like “inbound support” or “outbound sales.” A call driver is the specific reason inside that channel: “password reset,” “billing dispute,” “refund query.” Call types are usually 3–5 total; drivers are 15–40.
How many call drivers should a taxonomy have?
Most well-run operations cap it at 15–30. Below 15 and the tags are too coarse to be actionable; above 40 and agents pick “other” or the closest match rather than reading the whole list. Gartner’s 2024 contact-centre guidance recommends 20–25 as a sensible midpoint.
Who owns the call-driver taxonomy?
The ops manager owns day-to-day tagging and QA sampling. The product and engineering teams own the fixes for internal drivers. In a BPO relationship, the client typically owns the taxonomy structure while the vendor executes and reports.
How do BPOs use call drivers to price contracts?
Most outcome-based BPO contracts price on cost-per-contact by driver family, with voice-heavy drivers at one rate and chat-eligible drivers at a lower rate. The BPO partner reports driver-mix monthly, and shared-savings clauses fire when the top-3 drivers shift toward self-serve.
How often should the taxonomy be reviewed?
Quarterly at minimum. New product launches, policy changes, and channel expansions all introduce new drivers within weeks. A stale taxonomy is worse than none, because agents force-fit inquiries into obsolete tags and the dashboard lies to leadership.
Do AI chatbots change call-driver management?
Yes, deeply. AI now handles the top-3 self-service-eligible drivers (status checks, simple reversals, FAQs) without human touch. What’s left is a smaller, higher-complexity mix that changes the WFM formula: fewer contacts, longer AHT, higher-skilled agents. The taxonomy still runs the show; the mix inside it shifts.
Ready to scope an outsourced contact-centre programme with driver-taxonomy discipline built in? Explore vetted BPO partners on the Outsource Accelerator hubs.
Independent
