Most problems start before the order
When an IT project runs late on hardware, the cause is rarely the warehouse or customs. It is usually a planning call weeks earlier, where someone needed a price fast, a supplier answered fast, and nobody checked the details. The gaps show up later: a line item that was never confirmed, two people expecting two different delivery dates, paperwork that doesn’t match what the site is ready for.
Moving fast is fine when the inputs are complete. A date you can rely on is worth more than a quick answer you have to revisit.
Write an RFQ a supplier can work from
A good RFQ answers the questions a supplier would otherwise have to ask: confirmed part numbers, which substitutes you would accept, quantities, packaging requirements, where it ships and when you actually need it on site. With that, you get a firm offer back. Without it, you get an estimate and a few more rounds of email.
Teams sometimes leave detail out to keep things moving. It works the other way round: every open question turns into a revision later, usually under deadline. A complete request also makes quotes from different suppliers comparable.
If you buy regularly, keep a template. Required fields are the ones a supplier needs in order to commit; optional ones can cover budget, preferred alternates or rollout order. New staff then start from your standard instead of their own.
Sourcing risk and shipping risk are different problems
Sourcing risk is about the product: is it available, is the channel legitimate, is it the right configuration, what warranty comes with it. Shipping risk is about getting it there: transport, customs, handling, staging and the last mile. Discussing both in one conversation tends to produce weak trade-offs, because they need different checks and usually different people.
Check them separately. Confirm the line items, the substitutes and the commercial terms first, then the route, the delivery order and whether each site is ready. That way you don’t find a sourcing problem after the shipping plan is already booked. Everyone can see the same timeline, but each decision has one owner.
Build the schedule from dependencies
Delivery dates have to allow for approvals, integration windows, site access and change control. A plan worked backward from the launch date without checking those looks fine on a slide and falls apart in practice.
Instead of one delivery date, agree three: the date you want, the latest date you can live with, and the date by which you decide on a fallback. If a part gets tight, you already know what happens next, and finance knows how firm the date really is.
Suppliers can plan around this too. With realistic dates, stock can be reserved and inspection scheduled in advance, and problems get flagged early. That matters most in multi-site rollouts, where one late shipment holds up everything after it.
Agree on inspection and a fallback up front
Decide before the order who checks what, at which stage and against which criteria. A typical check covers physical condition, serial numbers where they matter, packaging and paperwork. Some categories also justify a functional or configuration check before release. Match the depth to how critical the equipment is; not every order needs the full treatment.
When both sides agree on acceptance criteria in advance, a problem becomes a conversation about facts instead of an argument about who is responsible.
The same goes for fallbacks. For each critical item, write down what triggers a switch, which alternatives are acceptable and who approves them. Teams that do this rarely end up needing it, and the exercise usually improves their first-choice decisions as well.
Keep one tracker and a short weekly review
Procurement watches cost, engineering watches compatibility, operations watches site readiness and leadership watches the date. Without a regular update, each group ends up with its own version of events. One shared list of commitments and exceptions fixes most of this, and it can be a spreadsheet.
A weekly review with procurement, delivery and technical leads is usually enough, plus a quick mid-week check on anything urgent. Ask the same questions each time. Has any line item changed? Has any date moved? Is a substitution waiting for approval? Does the inspection level still fit? Does someone senior need to make a call this week? If nothing has changed, keep it short.
Sort open items by what they are waiting for: information, a decision, or someone to do the work. It stops the same discussion coming back every week. And look back over your last few orders now and then. You will quickly see where uncertainty tends to come in, whether that is missing detail, late approvals, unclear substitution rules or shipping assumptions nobody confirmed.