A big customer wants us in their procurement system

Every one of these goes to the customer’s procurement or supplier-enablement contact, not to your account manager. Send them in one email and expect the answers to arrive over a week, because the person who knows is usually not the person who wrote to you.

The questions are ordered so that the first two answers alone usually settle it. If they need only purchase orders and a price file is acceptable, you are not doing punchout and you can go and do something faster.

Map the answers onto this table. The left column is what the customer said, the middle is what they usually mean, and the right is what the work actually is. Where two rows fit, the cheaper one is worth proposing first, because procurement teams accept a hosted catalogue more often than suppliers expect and nobody ever asks.

The cost column deliberately mixes money and time, because on a deadline the time is the binding constraint. A connection you can stand up in two weeks and improve later beats a better connection that misses the date and takes you off the approved supplier list.

If the answers point at real punchout, three things have to exist and only one of them is software you can buy. Knowing which of the three you are missing is the difference between a six-week project and a six-month one.

The first is a session: their system sends a setup request identifying the buying organisation and the buyer, and your side returns a URL that signs that buyer in as that company. The second is a priced catalogue for that company, which means your storefront has to already show one customer their own negotiated prices. The third is the return trip, where the cart goes back as a structured message their system can turn into a requisition.

The second one is where suppliers get stuck, and it has nothing to do with punchout. A storefront that cannot yet show a signed-in customer their contract price will faithfully deliver list prices into the buyer’s procurement system, where they will be approved and become a purchase order at the wrong price. Solve pricing before you solve punchout, in that order, every time.

Practical note on sequencing. Nothing works until the buyer’s side has been configured and tested against your specific connection, and their enablement team schedules that. Ask for the test slot at the same time you ask for the deadline.

Run the commercial test first, because it decides how much of the deadline is yours to worry about. Add up the annual revenue of the accounts that have asked for this in writing. Not the accounts you think might ask. If that number is a meaningful share of your book, this is a retention project and it deserves the budget. If it is one mid-sized account and the connection plus its ongoing maintenance costs more than the annual margin on it, you are allowed to price the work, show them the number, and offer the cheaper connection instead.

Then sequence by time to live rather than by elegance. A hosted catalogue file in two weeks keeps you on the list while you build. A punchout gateway in six to ten weeks is realistic if the buyer’s enablement team gives you a slot. A native platform build is not a deadline answer for anyone.

One thing to settle before you turn on a second channel: if you already exchange EDI purchase orders with this customer, adding punchout gives you two order paths into one account, which is how the same order arrives twice. Decide which channel owns which product range before, not after.