Reducing response time across seven workflows without adding headcount
Manufacturing / Home & interiors e-commerce · Program management across seven-plus cross-trained support workflows · Nov 2023 — Jan 2026 · via Hugo
Seven-plus workflows were being covered by five people, and response times were long enough to threaten the client relationship.
Delay came from context-switching across workflows rather than from workload volume — a specific attention and interference problem.
Restructured workflow allocation and workforce planning, and designed a Stroop-informed intervention plus process improvements to reduce switching cost.
Response time cut from 6h 43m to 2h 25m while quality finished at 99%, with no added headcount.
Overview. The client manufactures and supplies handmade tile in the United States. Hugo ran their front- and back-end customer support operation from launch in November 2023 through steady state, and I served as Project Manager on the program from its early weeks until January 2026.
I owned delivery for a team of one Team Lead and four Support Specialists across sample requests, quotes, showroom checkout, mock-ups, refunds, chat and back-office material-library workflows. Support was cross-trained rather than workflow-siloed, so coverage and quality had to be managed as one program rather than seven separate queues.
- Directed a team of one Team Lead and four Support Specialists against client-defined SLAs.
- Acted as the primary point of contact for client escalations and day-to-day communication with the client's operational leads.
- Owned billing accuracy, headcount allocation, and shift/workforce scheduling across the team.
- Ran recurring client calls and authored the program's Weekly and Executive Business Reviews.
- Diagnosed recurring performance issues and designed the training intervention that fixed the biggest one.
The program ran against tight, client-defined SLAs — a 5-minute average handling time on email and 1 minute on chat, with 2-hour response times on the general inbox and 1 hour on urgent tickets — while cross-training agents across seven-plus workflows inside two separate platforms.
A recurring quality issue emerged during ramp: newer agents struggled to quickly and accurately identify the correct tile colours and products in the system, which slowed handling time and created rework risk.
| Workflow | What it covers |
|---|---|
| Sample creation | Colour chip and size samples in default finish, requested by clients or reps |
| Library updates | Notifying clients of new product designs and tile style changes |
| Showroom checkout | Creating sample orders after in-showroom client visits |
| Quote creation | Formal commercial quotes on behalf of sales reps |
| Paint sheets (mock-ups) | Digital mock-ups from hand-painted, numbered colour indicators |
| Returns & refunds | Processed per company policy |
| Website chat support | Live chat inquiries against a 5-minute AHT target |
| Material library leads / audit | Back-office fulfilment and project data-integrity checks |
Billing, headcount and workforce management
I owned headcount allocation against the contracted minimum of 3 agents plus 1 QA per shift, tracked and approved overtime against client requests, and maintained the daily shift timetable across workflow assignments.
Coverage ran on a 3pm–2am WAT schedule aligned to the client's Pacific-time business hours, with no missed shifts against the minimum headcount requirement.
Client calls, WBRs and EBRs
I authored and presented the program's recurring Weekly Business Reviews — weekly snapshots, week-over-week metric trends across sample orders completed, response time and handling time, commentary on volume drivers, and next steps — plus Executive Business Reviews for broader stakeholder visibility.
The colour and product recognition problem: a Stroop-based intervention
When QA and client feedback both flagged that newer agents were slow and error-prone at matching the correct colour and product in the system, I designed a training intervention based on the Stroop effect — incongruent colour-name and product exercises, hands-on tile-sorting practice, and mnemonic aids to build faster, more accurate recognition.
I sequenced the rollout — content creation, review with the client, training content in the learning platform, then bi-weekly performance tracking — and assigned clear owners to each stage so the fix was measurable rather than anecdotal.
| Metric | Wk 1 | Wk 2 | Wk 3 | Wk 4 | Wk 5 |
|---|---|---|---|---|---|
| Sample orders completed | 120 | 135 | 122 | 132 | 66 |
| Avg. response time | 6h 43m | 5h 23m | 3h 15m | 3h 32m | 2h 25m |
| Avg. handling time | 5m 28s | 5m 46s | 10m | 7m 15s | 8m 16s |
Program close-out — final week I ran
- 01Volume: 1,007 outbound messages that week, up 45% week-over-week.
- 02Handling time: email handling down to 3m 39s, a 23% weekly improvement, driven by a default ticket-acknowledgment workflow that let agents work the queue oldest-to-newest.
- 03Quality: 99% quality score across communication, process compliance, email handling and client coordination, holding steady.
- 04Back-office throughput: material-library audits jumped to 975 in the week, up from 307, at roughly 30 audits per hour.
- 05Team: closed out with a fully cross-trained 4-agent roster and zero quality flags that week.
- Avg. response time, start to handoff6h 43m → 2h 25m
- Email handling time at handoff3m 39s
- Quality score at handoff99%
- Back-office audits in final week975
- Workflows held in steady state7+
- Cut average ticket response time from 6h 43m to 2h 25m over a five-week window, well inside the client's 2-hour SLA target.
- Sustained coverage against a 3-agent / 1-QA minimum headcount requirement with no missed shifts.
- Delivered a repeatable, trackable fix for a recurring quality issue instead of a one-off correction — client feedback specifically cited noticeable improvement in product and colour accuracy after rollout.
- Kept the program in steady state across seven-plus concurrent workflows with a lean team of five, from pilot through January 2026.
More case studies
Full index- 01US digital pharmacy clientDiagnosing the root causes of customer and compliance failures, separating agent performance from systemic issues, and building the governance to improve both safely at scale93% CSAT vs 90% target
- 02EdTech learning marketplaceReducing recurring customer dissatisfaction through frontline capability84% drop in issue-recognition DSATs
- 03Conversational commerce SaaS platformDesigning and scaling a new customer-support channel90% onboarding pass rate vs 60% target