PureSpring Water

The Challenge

PureSpring Water’s route map hadn’t been redrawn in eight years. When the company launched its Houston delivery operations, the founder built 15 routes by hand, assigning customers to drivers based on zip codes and rough geography. Those original routes worked. Then the customer base grew from 600 to 2,500, the driver count expanded from 15 to 25, and new customers were tacked onto whichever route seemed closest. No one ever went back to rebalance the system.

James Whitfield joined PureSpring as Operations Director and immediately recognized a problem that everyone else had accepted as normal: the route distribution was deeply unequal, and it was driving people out of the company.

Senior drivers who had been with PureSpring since the early years held the original, geographically tight routes. These routes covered dense clusters of customers in central Houston and the Galleria area. Senior drivers were finishing their deliveries by 1pm, loading up for a second partial run, and clocking out by 2:30pm with full pay.

New drivers got what was left. Their routes stretched from Katy to League City, covering three times the geographic area for roughly the same number of stops. New hires routinely worked until 5pm, sometimes later, and they could see that the senior drivers were heading home hours before them.

The problems were compounding:

• Route finish times ranging from 1pm to 5pm—a three-hour gap visible to everyone. New drivers talked about it in the break room, in the truck, and during exit interviews.
• 40% annual driver turnover: PureSpring lost 10 of its 25 drivers every year. At a hiring and training cost of $3,500 per driver, that was $35,000 annually spent replacing people who left because the job felt unfair.
• Fuel waste from sprawling routes: New-driver routes covered significantly more miles than senior-driver routes for comparable stop counts. The company was spending an estimated $25,000 per month on fuel, with a disproportionate share going to the inefficient outer routes.
• No system for adding new customers efficiently: When a new commercial account signed up, the office manager added it to whichever route served that zip code. There was no analysis of drive time impact, no consideration of workload balance, and no optimization after the addition.
• The turnover problem was self-reinforcing: New drivers quit within 4 to 6 months. Their routes were covered by other drivers working overtime, which burned out additional people. When a replacement was hired, they inherited the same sprawling route and the cycle started over.
• Total cost calculations: $35,000 in direct recruiting and training expenses, plus an estimated $15,000 in overtime pay for coverage during vacancies, plus unquantified costs from missed deliveries, customer complaints, and management time spent on hiring. The route imbalance was PureSpring’s most expensive operational problem.

The Solution

James evaluated four route optimization platforms. His primary requirement was the ability to balance workload across 25 drivers not just by stop count but by total drive time. A route with 100 stops in a 5-mile radius is a fundamentally different workday than a route with 100 stops spread across 40 miles. He needed software that understood that distinction.

Route Optimization handled both dimensions. James could set constraints for maximum drive time, maximum stops, and service time per stop, and the optimizer would balance routes across all three factors simultaneously. The route schedule feature was equally important: PureSpring’s customers were on biweekly delivery cycles, and James needed templates he could reuse rather than rebuilding routes from scratch every two weeks.

A Complete Route Rebuild for 2,500 Customers

The rebuild started with a full data import. James exported PureSpring’s entire customer database, including delivery addresses, delivery frequency (weekly or biweekly), average delivery time per stop, and access instructions. He uploaded the CSV into Upper and ran the optimization across all 25 drivers.

The results looked nothing like the old route map. Upper redrew every territory based on actual drive time and stop density, not zip code boundaries drawn eight years ago. Some senior drivers gained stops in areas they’d never served. Some new drivers lost the long highway stretches that had made their days miserable.

James created separate templates for Week A and Week B of the biweekly cycle. Each template was saved and reloaded every two weeks, with adjustments only for new customers, cancellations, or schedule changes. The biweekly planning process that had previously required two hours of spreadsheet work now took 15 minutes of template loading and minor edits.

Equalizing the Workday

The most important metric wasn’t route length or stop count. It was finish time. James set a target: every driver should complete their route between 2:30pm and 3:15pm. No one finishes at 1pm while someone else is out until 5pm.

Upper’s capacity optimization feature balanced the routes to hit that window. Routes in dense urban areas had more stops but shorter drive times. Routes covering suburban and exurban areas had fewer stops but longer distances. The total workload, measured in combined drive time plus service time, was equalized across the fleet.

When the new routes launched, the finish-time gap dropped from over three hours to 45 minutes. The earliest driver finished at 2:30pm. The latest finished at 3:15pm. Every driver worked a comparable day, and the conversations about unfair routes stopped.

James also implemented a rule for new customer assignments. Instead of adding new accounts to routes by zip code, he ran the optimization monthly with all new customers included. Upper placed each new stop on the route where it had the least impact on total drive time, keeping the balance intact as PureSpring’s customer base grew.

Territory Zones and Route Consistency

Drivers in recurring delivery businesses build relationships with their customers. A water delivery driver who shows up every two weeks at the same office knows where to leave the jugs, which elevator to use, and which receptionist to check in with. James didn’t want to sacrifice those relationships by rotating drivers through random routes.

Upper’s territory feature let James define geographic zones that kept each driver serving a consistent area. The optimization balanced workloads within those zone constraints, so drivers kept their customer relationships while still benefiting from efficient stop sequencing.

The Impact

PureSpring Water’s route rebuild produced measurable results within the first quarter and compounding benefits over the following year. The improvements touched every part of the operation, from driver satisfaction to fuel costs to customer growth.

Driver turnover dropped from 40% to 12% annually. In practical terms, PureSpring went from losing 10 drivers per year to losing 3. At $3,500 per hire, that saved $24,500 in direct recruiting and training costs. Including reduced overtime for vacancy coverage, the total annual savings reached approximately $28,000.

The fuel savings came from eliminating the sprawling, overlapping routes that had accumulated over eight years. With geographically optimized territories and efficient stop sequencing, PureSpring reduced total fleet mileage by 22%. At their fuel consumption rate, that translated to $5,500 per month in savings.

Customer growth accelerated once the route infrastructure could support it. James added 300 new commercial and residential accounts over six months, each one placed on the optimal route by Upper’s algorithm. The additions were absorbed without extending any driver’s workday past 3:15pm and without hiring additional drivers.

PureSpring’s management team approved a fleet expansion plan for the following year, adding 5 more drivers and extending coverage to the Beaumont-Port Arthur corridor. The decision was possible because the existing operation was stable. Turnover was under control, routes were balanced, and the system could absorb growth without creating new imbalances.

Performance Metrics

Driver satisfaction, which James measured through quarterly anonymous surveys, rose from 3.1 to 4.4 out of 5. The improvement was almost entirely driven by one factor: route fairness. When drivers feel their workload is comparable to their peers, other aspects of the job become more tolerable. When they feel they’re working harder than everyone else for the same pay, nothing else matters.