Five habits behind a dependable service visit

When dozens of stores report the same symptom within minutes, a dispatch desk's instinct is to work each ticket. That instinct sends technicians to sites where nothing can be fixed at the counter. Since 2024 our system flags any cluster of more than 25 tickets on one device model within thirty minutes, and the shift lead must either open a single incident or record why not. An incident lead then owns communication with the client, the remote team tests a rollback or workaround, and dispatch holds non-urgent visits. I review every cluster that was dismissed, because a false alarm costs an hour of attention while a missed one can cost thousands of stores a weekend of card sales. The rule is deliberately simple, so that a dispatcher on a night shift can apply it without first asking anyone's permission.

Utilisation is the figure our sponsor watches most closely, and it is easy to flatter. If drive time, parts collection and waiting for a store manager are excluded, a technician can look underused at 70 per cent and be exhausted in the same week. We now plan on a working week that includes travel, and we cap scheduled hours at fifty, with anything above that needing a regional manager's approval. Planned utilisation fell from 81 to 74 per cent in 2024. Technician attrition fell from 27 to 19 per cent over the following year, though wage changes contributed and I do not credit the schedule alone. The trade-off is real: fewer billable hours per person means employing more people, and I defend that cost in every budget round.

A rollout of kiosks or new terminals is judged by the client on the first morning after installation, when staff either can or cannot use the equipment. I build rollout plans from the number of trained crews we employ, each site's trading hours and the client's own freeze periods around promotions and holidays. Where a client's date demands more crews than we have, I set out the options in writing: a smaller committed scope, a later finish, or added capacity with a stated quality risk. I also keep a revisit rate for every rollout, measured as sites needing a second visit within thirty days. It is the most honest measure I know of whether an installation crew was rushed, and I show it to clients before they ask for it.

Our Querétaro centre handles routine remote troubleshooting in English and Spanish during the day for clients across the United States. Indianapolis keeps night escalations, mass-incident coordination and the decisions to dispatch. I drew that line on a practical test: work moves nearshore when it follows a written runbook, and stays onshore when it requires judgement about sending a person, interrupting a store or spending a client's money. The saving is meaningful, roughly 38 per cent per remote-resolution hour, but it disappears if an analyst escalates everything because the runbook is thin. Before any further work moves, the receiving team runs it alongside Indianapolis for six weeks and we compare first-contact resolution. Some work has moved back after that comparison, and I report those reversals to the board alongside the savings.

A four-hour response commitment sounds clear until someone asks when the clock starts, whether a remote fix counts and what happens when the client's own supplier caused the fault. I work with our commercial and legal colleagues on the operational definitions behind each service level before a contract is signed. We now specify how a mass incident is declared, how credits apply when several causes overlap and what reporting the client receives during an outage. Credits matter less to most clients than candour during the event. I would rather pay a credit we owe than defend a technically correct exclusion that store managers regard as an excuse. Commercial colleagues own pricing and terms; my responsibility is to make sure operations can deliver and measure each commitment exactly as written.