Marathon Customer Care: An End-to-End Operations Playbook

Scope, Service Levels, and KPIs

Define the customer care scope across five phases: (1) Registration and lottery (T–9 to T–6 months), (2) Pre-race logistics (T–6 to T–1 months), (3) Race week (T–7 to T–1 days), (4) Race day (T–1 to T+1 days), and (5) Post-race (T+1 to T+30 days). For a 20,000–50,000-participant marathon, plan for 0.15–0.30 inbound contacts per registrant over the full cycle, with 40–55% of volume landing in the final 14 days. Expect a same-day spike of 3–7x baseline on bib assignment day and the final corral-change deadline.

Set clear SLAs by channel: email 12 business hours to first response (target 80% within SLA), live chat 60 seconds to answer (80/60), phone 80% of calls answered within 30 seconds (80/30), and social care responses within 60 minutes 07:00–22:00 local. Target first contact resolution (FCR) of 78–85%, customer satisfaction (CSAT) of ≥4.6/5, and post-race Net Promoter Score (NPS) of ≥60. Measure abandonment rate (<5% phone, <2% chat), reopen rate (<8%), and deflection from self-service (≥30%).

Forecast weekly volume using simple ratios if you lack historicals: in weeks T–8 to T–5, 1.5–2.5 contacts per 100 registrants; T–4 to T–2, 3–5 per 100; race week, 8–12 per 100; race day, 12–20 per 100 (mostly logistics, tracking, and family queries). Adjust ±20% for first-year events or when changing start logistics, baggage policy, or medical screening.

Channel Strategy and Tooling

Provide at minimum four channels: email (for documentation-heavy issues like deferrals), live chat (policy clarifications and quick itinerary help), phone (urgent travel, accessibility, safety), and social (Twitter/X and Instagram for public-facing updates). Use a unified help desk/CRM to keep a single ticket history across channels; configure mandatory fields for “runner ID/bib,” “race year,” “topic,” and “resolution code” to power reporting. Implement unique, channel-specific SLAs and business hours calendars to avoid false breaches.

Stack recommendations: ticketing/CRM (e.g., Zendesk, Freshdesk, Salesforce Service), telephony/IVR (e.g., Aircall, Talkdesk, Twilio), and an integrated knowledge base. Build 60–90 knowledge base articles covering policies, logistics, and troubleshooting. Tag articles by phase and surface them contextually on key pages (registration, results, tracking). Use macros and forms for repeat scenarios (name changes, wave/corral requests, expo hours, travel letters). Automate responses for the top 10 intents but leave easy agent override when nuance exists.

• High-value knowledge base articles: registration edits, payment receipts and VAT, time-qualification verification, lottery odds and timelines, deferrals/transfers (fees and deadlines), corral/wave seeding rules, medical certificate requirements (where applicable), expo hours and location, bib pickup proxy rules, gear check vs. no-baggage start, start transportation windows, pacer policy, wheelchair/handcycle procedures, course cutoffs and sweep buses, hydration/gel locations, on-course medical and AED coverage, family reunion area maps, real-time tracking FAQ, age-group awards timelines, results verification and protest process, medal/merch replacement policy, post-race photo partners and delivery windows.

Instrument deflection: show article suggestions on contact form submit and measure click-through to resolution. Aim for ≥25% form abandonment after article view (good deflection), with a satisfaction prompt to confirm usefulness. Maintain a status page for incidents (e.g., tracking outage) to reduce inbound; update with timestamps and next ETA.

Staffing, Scheduling, and Training

Use a simple demand model: for every 10,000 entrants, plan 4–6 FTE agents steady-state (T–8 to T–5), 6–9 FTE in ramp (T–4 to T–2), and 10–14 FTE plus 8–12 trained temps race week. On race day, staff 1 phone agent per 1,000 expected concurrent viewers of your tracking app peak (often 10–20% of entrants’ supporters), with an overflow BPO ready to absorb a 3x surge. Provide coverage 08:00–20:00 (Mon–Fri) off-peak, extend to 08:00–22:00 (Mon–Sun) in the final two weeks, and run 04:00–16:00 local on race day.

Implement skills-based routing: policy queries, registration billing, elite/athlete services, accessibility, and medical non-diagnostic (scripted triage only) as separate queues. Establish an escalation matrix: Tier 1 (all), Tier 2 (billing adjustments, policy exceptions), Tier 3 (race directorate, medical director, operations). Maintain on-call roles with contact windows and decision rights documented.

Training plan: 12 hours of LMS modules (policies, logistics, tone and de-escalation), 4 hours of scenario labs, 4 hours of side-by-side, and certification on 20 core macros. QA 5 random tickets per agent per week; target ≥90% policy adherence, ≥95% accuracy, ≤2% compliance errors (PII, PCI). Calibrate weekly during the last 30 days with sample reviews to keep decisions consistent under pressure.

Policies: Refunds, Transfers, and Deferrals

Publish a one-page “money and entries” summary with unambiguous dates, times (including time zone), fees, and edge cases. Example baseline: no cash refunds after 30 days from purchase; name transfers allowed until T–30 days for a $25–$50 fee; deferrals allowed until T–14 days with a $35–$75 fee, granting 1x entry for the next year with the requirement to pay the then-current registration price. For lotteries, clarify whether deferrals bypass the lottery and whether charity/third-party entries are eligible.

Define medical and military exceptions: permit late deferrals up to T–48 hours with documentation from a licensed practitioner or official orders. Clarify that medical advice is not provided by customer care; agents triage to the medical team using a non-diagnostic script and capture only necessary details (symptoms avoided; focus on capability to participate and documentation).

Publish processing timelines: transfers completed within 2 business days, deferral codes issued within 24 hours, refunds (where applicable) posted in 5–10 business days. Note all non-refundable items (processing fees, charitable donations) and taxes/VAT handling. Maintain an audit trail for policy exceptions, with approvals recorded by name and timestamp.

Crisis Communications and Race-Day Support

Establish an incident command structure (ICS) with clear comms pathways. For severe weather or course alterations, use pre-approved templates for SMS, push, email, website banner, and social pinned posts. Maintain a mass-notification system capable of sending at least 60,000 messages in under 5 minutes with delivery status reporting. On race day, set hotline hours 04:00–16:00 local and publish them in advance in the runner guide and tracking app.

Situate information desks at the expo and near each start village entrance; staff with policy-trained agents and equip with printed quick-reference sheets. Use a color-coded status protocol (Green/Yellow/Red/Black) aligned with medical and race operations, and ensure customer care scripts match the public status. Log all race-day announcements with exact timestamps to synchronize agent messaging and reduce rumor amplification.

• Race-day quick references (example details): Hotline +1-555-0137 (voice/SMS enabled); Accessibility line +1-555-0142 (TTY 711); Tracking status page status.examplemarathon.com; On-site Info Desk “Start Village Gate C” 05:00–09:00; Family Reunion Info “12th St. & Park Ave. (Northeast Corner)” 09:30–15:30; Lost & Found Desk “Finish Chute West Exit” 09:30–16:00; Media inquiries [email protected] (not for runner support).

After-action: within 72 hours, publish a debrief note if there were material disruptions, with a simple timeline, what occurred, and what will change for next year. Internally, produce a 1-page KPI report (volume by channel and topic, SLA, FCR, CSAT, top 10 macros used) and a backlog list of content or policy fixes.

Data Privacy, Accessibility, and Language Support

Handle PII under GDPR/CCPA: collect only what’s necessary (name, DOB, email, phone, emergency contact, and verified time-qualification where applicable). Retain support ticket data for 24 months; purge payment tokens immediately after settlement if you are not the processor. Restrict access by role and enable SSO/MFA for all tools. For payments, keep customer care out of card handling; all changes route through secure self-service links (PCI DSS SAQ A posture).

Offer at least three languages if your entrant base is international; as a baseline for North America/Europe, cover English, Spanish, and French. Use certified interpreters for phone and chat when needed; publish coverage hours. Provide accessibility: TTY relay for hearing-impaired runners, large-type and WCAG 2.2 AA-compliant web pages, and clear wayfinding text for blind/low-vision runners using screen readers. Ensure agents can schedule guide runners or mobility assistance where permitted, with a documented cutoff (e.g., T–14 days).

Create a Data Subject Request channel ([email protected]) and document SLA: acknowledgement in 72 hours, fulfillment within 30 days. Maintain a deletion workflow that removes linked attachments (medical notes, IDs) and confirms completion to the requestor. Log all privacy decisions for audit with date, operator, and disposition.

Budget and ROI

Use a per-registrant model to budget: $2.50–$3.50 per registrant for core care (tools, staffing, translation, QA), plus a 20% contingency during race week. Example for 30,000 entrants: base $90,000 (at $3.00), contingency $18,000, knowledge base and content $8,000, and overflow BPO reserve $15,000 at $25/hour. Tooling for omnichannel typically lands at $0.20–$0.45 per contact depending on volume and contract.

Track ROI through deflection and churn prevention: every 10% increase in self-service resolution commonly reduces agent hours by 8–12%. Preventing 300 refunds at an average $160 registration value protects $48,000 gross; clear policies and fast responses directly influence that outcome. Reinforce the business case by reporting monthly on cost per contact by channel (email $3–$5, chat $4–$6, phone $7–$12, social $2–$4) and highlighting improvements from content and automation.