An operations team signs off on a new sortation system after a strong vendor demo. Six months after go-live, the sorter runs well, but orders still ship late. The real delay was upstream in picking, and the new equipment simply made the line in front of it longer.
Stories like this usually trace back to questions nobody asked before the purchase order was signed. The seven questions below will help you test your plan, bring the right people into the room, and choose automation that fixes the problem you actually have.
1. What Problem Are We Trying to Solve?
Start by naming the constraint in one sentence. Automation is a tool for a specific problem, and a vague goal like “modernize the warehouse” leads to vague results.
Most warehouse automation projects target one or more of these constraints:
- Labor availability: You cannot hire or keep enough people to meet demand.
- Throughput: Orders per hour fall short during peak periods.
- Accuracy: Mispicks, short shipments, and returns are costing money and customers.
- Space: Your SKU count is growing faster than your building.
- Safety and ergonomics: Workers lift, bend, and walk all day. Across private industry in 2023 and 2024, overexertion and repetitive motion caused about 946,000 cases serious enough to require time away from work or restricted duty, more than any other cause, according to the Bureau of Labor Statistics.
Rank these by what they cost you each week. If your managers describe the main problem differently, settle that disagreement first. It will shape every decision that follows.
2. Is Our Current Process Worth Automating?
Automating a broken process makes the wrong work happen faster. Before you buy equipment, make sure the workflow itself is sound.
Picking is a good place to look. In the 2026 Automation Study from Peerless Research Group and Modern Materials Handling, only 12 percent of respondents had fully automated picking, and 33 percent ran it manually with no plans to automate. In most buildings, people still walk to every pick location, so slotting, routing, and batching rules decide how far they travel each shift. Slotting software was also among the tools respondents most planned to add over the next two years.
Check these areas before you commit:
- Slotting: Are your fastest-moving items stored closest to packing and shipping?
- Pick methods: Would batch, zone, or wave picking reduce walking?
- Data quality: Are item dimensions, weights, and location records accurate?
- Operator input: What workarounds does your team use every day, and why?
Low-cost fixes may solve part of the problem. At a minimum, they give you a clean baseline to measure automation against.
3. What Does Our Order Data Say About Volume and Growth?
Your data should decide the size and type of system. Pull 12 to 24 months of order history and study it closely.
Focus on these details:
- Order profile: How many lines per order and units per line? How many orders are single items?
- SKU velocity: How much of your volume comes from a small group of items?
- Handling units: Do you ship eaches, cases, totes, or pallets, and where do they split or combine?
- Peaks: What does your busiest hour on your busiest day look like?
- Growth: What volume is realistic in three to five years?
A system sized for an average Tuesday will struggle the week before a major holiday. Many planners size equipment to a high-volume “design day” and then decide how overtime or temporary labor will cover the rare days above it. Whatever method you choose, make it a deliberate decision.
4. How Much Flexibility Will We Need?
Match the rigidity of the system to how certain your future is. Fixed systems such as conveyor, sortation, and automated storage and retrieval systems (AS/RS) can deliver high throughput and storage density. They are also hard to move or resize once installed. Flexible options such as autonomous mobile robots (AMRs) can often be added in smaller steps and adapted when the layout changes.
Ask these questions to find the right balance:
- Facility timeline: How many years remain on your lease, and how does that compare to the expected payback period?
- Business changes: How often does your product mix, customer base, or sales channel shift?
- Volume swings: Could volume double, or drop, in the next few years?
- Phasing: Can the system be expanded in stages as demand grows?
A stable, high-volume operation in an owned building can justify fixed equipment. A fast-changing business in a leased space usually benefits from options that can move with it.
5. How Will the System Connect to Our Software?
Equipment only does what software tells it to do, so integration deserves as much attention as hardware. Three layers are usually involved:
- Warehouse management system (WMS): Manages inventory, orders, and tasks.
- Warehouse execution system (WES): Balances work across people and machines in real time.
- Warehouse control system (WCS): Directs the equipment itself, such as conveyors and sorters.
Some operations need all three. Others combine them. Either way, get clear answers to these questions:
- Can your current WMS send the data the new equipment needs, as fast as it needs it?
- Who owns and supports each software layer?
- How will the system handle exceptions like short picks, unreadable labels, or damaged cartons?
- What performance data will come back for reporting?
Write down which system makes each decision. Integration gaps that look small on paper often become the biggest delays at go-live.
6. What Happens When Something Goes Wrong?
Every system stops at some point. Plan for downtime before it happens, so a breakdown costs you a few hours and your orders still leave on time.
Cover these points in your plan:
- Uptime targets: What availability is promised, and how will it be measured?
- Manual backup: Can you still ship orders if the system is down for a full shift?
- Maintenance and parts: Who handles preventive maintenance, which spare parts are kept on site, and how fast will a technician respond?
- Training: Are operators trained on recovery steps as well as the normal workflow?
- Single points of failure: Does one sorter or one conveyor line feed your entire shipping area?
Your team will learn the most about a system in its first bad week. Prepare them for it.
7. What Will It Really Cost, and How Will We Measure Success?
The purchase price is only part of the investment. A realistic budget includes:
- Equipment, installation, and software licenses
- Integration with your WMS and other business systems
- Building changes such as power, floor flatness, and fire protection
- Maintenance, service contracts, and spare parts
- Training, plus the drop in productivity during ramp-up
- Room for future expansion
Next, record your baseline before anything changes. Useful metrics include cost per order, lines picked per labor hour, order cycle time, pick accuracy, dock-to-stock time, and system uptime. Compare these numbers again once the system has stabilized. Without a baseline, you cannot show whether the project paid off or where it needs tuning.
Start With the Questions
The team in the opening story would likely have found its picking bottleneck by working through the first three questions. A few weeks of analysis could have pointed its budget at the right part of the building.
Put these seven questions in writing. Answer each one with your own data, involve the people who run the floor, and share the answers with every vendor or integrator you talk to. Good partners will welcome the clarity, and you will be able to compare proposals on equal terms.
