Service Parts Teams Rush Parts After Stockouts Their Control Towers Saw Coming, Research Finds

Service parts organizations mostly use their control towers to react after a stockout, even as those systems flag hundreds to thousands of alerts a week, according to Unilog.SC’s October 2026 Industry Research Report. Four of five global equipment makers surveyed were receiving between 300 and 2,000 control tower alerts weekly, with 15 to 50 percent of those alerts judged high priority. Managing the gap between the plan and actual network conditions took up 40 to 90 percent of planning teams’ time.

That gap between signal and response is now how analysts grade the category. Nucleus Research’s 2026 Control Tower Technology Value Matrix, published Sept. 1, describes control tower maturity as increasingly defined by “decision velocity,” the speed at which an organization moves from a relevant signal to a response it can execute. The firm notes that buyers are tying their control tower investments to measurable reductions in backorders and manual exception work.

The Unilog report, titled “Your control tower sees everything. That is the problem,” draws on a survey of five companies and a simulation built on the network of a global IT hardware company. Part of the reason companies respond late, according to the report, is that alerts are generated automatically while every response is still handled manually.

Timing carries particular weight in service parts. A typical part, as the report describes it, moves fewer than five times a year and costs more than $1,000. Those parts are stocked close to customer sites because service contracts measure acceptable waiting time in hours.

All five surveyed companies relied heavily on rushing parts from a central warehouse once a stockout had already happened. Only one did so extensively before the stockout, at the point when the control tower could already see the shortage coming.

In the simulation, two interventions produced most of the improvement. Moving parts between local warehouses ahead of a shortage raised same-day fill rates by about 3 points from a base near 95 percent, and sending parts early from the central warehouse raised them by about 4 percent. Neither expediting repairs nor reserving stock for premium customers improved results.

Actions taken closest to the customer had the greatest effect. Roughly 30 percent of transfers between local warehouses delivered 80 percent of the gains, while decisions made further upstream left too many steps between the action and the customer’s equipment. The report also found that notifying purchasing as soon as a failed part goes unreturned or proves unrepairable, so a replacement order starts sooner, improved fill rates at negligible cost.

Eyal Yossef, VP of supply chain solutions at Unilog.SC, said the purpose of a service parts control tower is to isolate the small number of problems “that will stop a customer’s machine” from the much larger volume of minor deviations, and to do so while a transfer between two stocking locations can still fix them. In his view, that means a short list with a clear owner, and decisions made before the week begins.

For global networks, the nearest stocking location is often in another country. The report notes that this turns a warehouse-to-warehouse transfer into a customs operation requiring an importer of record and clearance time. It argues that a service parts lead logistics provider operating both the forward stocking locations and the cross-border flow can determine which transfers will arrive in time to protect a customer commitment.

The report includes a set of ten service parts exceptions, each paired with a recommended first response. They range from forward stocking locations at risk of running dry before the next replenishment to customs holds on cross-border transfers, and the list also covers customers nearing their contract limits. As a starting point, the report suggests comparing how many alerts the control tower raised last week against how many a planner actually opened. It also recommends automating decisions on low-cost, fast-moving parts so planners can spend their time on the expensive ones.

The report also recommends ranking alerts by the customer contract at risk. Service contracts are written at the customer and system level, while day-to-day decisions are still made part by part. A low-stock alert at a single location, the report notes, does not show which customer has already used most of its downtime allowance for the quarter.