In a prior IT leadership role, I took responsibility for Professional Services delivery at a software company, Sales was signing multi-year customer contracts, but invoicing did not begin until customers went live. The implementation queue was so congested that credible start windows for some new customers stretched 12 to 18 months into the future. Business already sold could remain unbilled while the delivery backlog continued to grow.
$13.8M in new ARR activated through completed implementations in my first full year, four times the previous annual record of $3.4M.
Signed recurring revenue awaiting implementation fell from about $22M to $8M, despite continued new sales. The following year was on track to exceed its $15.9M new-ARR target when I left.
Why starting more projects slowed delivery
A newly signed contract was routinely treated as an urgent kickoff, even when the same specialist teams were still committed to existing customers. Sales made promises without a reliable view of delivery capacity. Active projects were interrupted or delayed to accommodate the next priority, increasing work in progress without moving enough customers into production.
The delivery practices compounded the problem. Project managers followed different methods, requirements and build documentation varied, and no shared capacity model connected project scope, staffing, customer readiness, and likely go-live dates. Some implementations reached production with unresolved defects or requests that kept the delivery team occupied for months afterward. That stabilization work consumed capacity needed for the next customer.
The constraint was the way work entered, moved through, and exited the delivery organization. Adding more starts or another provider to the same system would not have cleared it.
Make capacity and priority decisions explicit
I rebuilt the function around a common project-management office, a defined implementation lifecycle, documented requirements, and repeatable technical build standards. I also built a working capacity model covering the signed backlog, project stage and type, weekly specialist assignments, expected start and go-live dates, stabilization periods, and available capacity.
That model became the basis for regular decisions with Sales, Finance, and technology leadership. I stopped offering start dates that delivery could not support. When leadership wanted a customer accelerated, the decision included which existing commitment would move and the consequences for its go-live and revenue timing. Sales could then prepare customers for realistic start windows instead of discovering capacity constraints after making the promise.
Inside delivery, we reduced competing assignments for scarce specialists, made implementation work less dependent on a small group of senior staff, and used project execution to develop broader team capability. Requirements and build standards reduced avoidable rework. An external implementation provider was evaluated, but the nonstandard process would have made outsourcing slower and more expensive while still consuming internal expertise. Fixing the delivery system came first.
A go-live started revenue. A completed handoff released capacity.
This was a mixed portfolio of software implementations and upgrades. Some deployments involved complex integrations with customer systems and operational workflows, requiring technical staff, business analysts, and customer teams to resolve field mappings and process changes. Production go-live activated the contracted ARR, but the delivery obligation continued until the customer could be supported reliably.
We tracked stabilization as a distinct internal project phase, prioritized failures in contracted functionality, and separated later enhancements from unfinished implementation obligations. The final transfer to Technical Services was a planned, customer-visible event with readiness conditions, a date, and a receiving owner. That created a credible point at which implementation capacity could return to new work without abandoning existing customers.
Results that appeared within months
By midyear, the faster pace of customer go-lives was evident enough that leadership raised the annual target.
Average implementation duration fell from about 140 to 105 days across the project portfolio. Average quarterly go-lives increased from five to 16. Post-go-live stabilization declined from more than 120 days to roughly 30 to 45 days.
Those operating gains produced the commercial result: $13.8M in new annual recurring revenue activated in my first full year, compared with the earlier $3.4M record. The signed but unimplemented recurring-revenue backlog declined from about $22M to $8M, even as new contracts continued to arrive.
When I departed, the next year was on track to exceed its $15.9M target, with the remaining customer go-lives already underway.
For a business owner or Operating Partner, the exposure is straightforward: contracted demand creates value only when the organization can deliver what was sold. The ability to complete implementations, activate revenue, and release capacity determines whether growth can be converted into operating results.
Related Plain Ops work: When the constraint is unclear, the IT Operations X-ray can establish what is limiting delivery. When the cause is already understood, defined-outcome work or Execution Leadership can help implement the operating changes. Discuss the situation.
