Company Profile

Spendesk, a leading European spend-management platform, onboards mid-market finance teams. Every new customer requires policy configuration, accounting and HR integrations across 40+ connectors, and a company-wide employee rollout. Customer implementations historically ran 3 to 4 weeks, with the final week often blocked by a single missing piece of customer-side data.

The implementation team sits inside customer success. The same managers who own the rollout also own the renewal, which means every day the implementation slips, is a day of compounding risk on an account that's already paid.

Challenge

Implementation managers were spending most of their time pushing projects forward, not configuring. The day filled up with the work between the work:

• Chasing customers for sign-off, data, and integration credentials, often through three or four touches per ask

• Re-reviewing submitted configuration data for errors before go-live, because submitted didn't mean correct

• Coordinating across accounting, HR, and IT contacts on the customer side, with each stakeholder operating on their own calendar

• Updating the CRM and project tracker by hand after every interaction, because the next manager picking up the account needed to know what happened

Roughly a third of implementations stalled for two-plus weeks waiting on the customer. The customer wasn't deliberately blocking. They were busy. A request for HRIS credentials would sit in someone's inbox for ten days. A clarifying question about chart-of-accounts mapping would bounce between the finance lead and the accountant. None of these were single moments of dropping the ball. They were the slow accumulation of a thousand small inertia tax payments, each one taking a day off the implementation calendar.

Time-to-value dragged. Early renewals went into the pipeline at risk. And every stalled account took manager bandwidth that should have been moving the active ones forward.

What Tandem did

The team kept their existing CRM, project tracker, and ticketing systems. Tandem sat on top, doing four things:

• Closed the customer follow-up loop. Tandem ran an automated, escalating cadence the moment a customer went dark on a request, with each touch tailored to the specific outstanding item (credentials, sign-off, data file). If the customer responded with the wrong artifact, Tandem caught it and asked again with more specificity. If they didn't respond at all, the cadence escalated from email to Slack to a calendar-side nudge for the manager to step in personally.

• Validated configuration data on submission. When a customer uploaded their accounting mapping or integration credentials, Tandem checked them immediately. If a journal template was missing a required field, if a credential was expired, if a multi-entity mapping was missing one of the entities, Tandem flagged it back to the customer before the manager ever saw it. Errors that used to be caught by the manager during a manual review, sometimes days later, now got resolved at the moment of submission.

• Scripted the cross-team handoffs. Every handoff that used to happen in a Slack thread (accounting integration confirmed, HRIS data validated, ready for go-live) became a tracked task with an owner, a due date, and an automated reminder. The handoffs that used to fall through the cracks because nobody remembered who was supposed to pick them up now ran themselves.

• Kept the records clean automatically. CRM and project tracker updates happened in the background as Tandem completed actions. Managers stopped spending the last hour of their day updating Salesforce and Asana. Leadership got an accurate view of where every account stood, in real time.

The work that didn't move was the strategic work. Managers still ran the discovery calls, made the configuration judgment calls, owned the customer relationship. Tandem absorbed the coordination, the chasing, the validation and the documentation.

Results

The 57% reduction in median time-to-go-live is the headline. But the more durable change is in the distribution. Before Tandem, the gap between a healthy implementation and a stalled one was weeks. After Tandem, the long tail compressed. The implementations that used to drag for 35 to 45 days now close in 12 to 15. Managers stopped having one runaway account eating a quarter of their capacity.

Why leadership cares

Every implementation that closes thirteen days sooner recognizes revenue sooner and de-risks the first renewal. Faster go-live and fewer stalls mean more accounts live per manager and healthier NRR. It's the bird’s-eye view leadership actually asks for: accounts at risk, accounts to be implemented, and how fast both columns clear.

The second-order win is harder to put in a table. With each manager able to run 15 concurrent implementations instead of 9, Spendesk's next ten enterprise logos didn't require the next two hires. The team capacity scaled with revenue, not with headcount.

"A process that used to take our team hours per client now runs on its own. It fundamentally changed what our CSMs can spend their time on."

— Mathilde Hamy, Head of CS & Account Management, Spendesk