How to run a 3PL software trial or pilot

A demo shows you what the software can do. A pilot shows you what it does with your own receiving, your own pickers and your own rate card. For a small 3PL the second one is the only test worth paying attention to, and it fits in two weeks if you plan it.

This page is a script for that two weeks: what to agree first, what to run each week, and how to decide at the end. It assumes one client and one building. The cost side of a purchase is covered in 3PL software pricing models.

Agree the ground rules first

A trial that starts on a sales call and a login often ends with a vague "it looked fine". Write these down before anyone is given access:

• Dates. Start day, end day, and a review meeting already in the calendar.
• Cost. Whether the pilot costs anything, and if it does, what it costs and whether that comes off the first invoice. Get it in writing.
• Data. What you will load, who owns it, and what happens to it if you do not continue. Ask for a full export at the end whatever you decide.
• People. Which two or three of your staff will use it, and who at the vendor answers questions during the two weeks and how fast.
• Scope. What you are not testing. Leaving out EDI, a second warehouse or a returns flow on purpose keeps the pilot honest.

Pick the client and the stock

You have two choices, and neither is wrong.

A pretend client. You invent a client with a rate card, 20 or 30 items and a few inbound deliveries. Nothing can go wrong for a real customer. The weakness is that nobody takes pretend orders seriously, so the floor staff may rush through them.

A real client, in parallel. You keep running the client with your current method and mirror the same movements in the new system. It is double work, but you get real items, real mistakes and real invoices to compare against. Choose a client with steady, simple orders and tell them what you are doing.

Do not put a real client's live operation only in a system you have not tested. That is a launch, not a pilot.

Week one: stock in, stock found

Week one tests the foundation: can you receive goods and then find them.

1. Set up the client. Time how long it takes to create the client, enter the items and add the locations. Note every field you did not understand.
2. Receive a delivery against an expected quantity. Then receive one that does not match it, one with a damaged carton and one with an item you did not expect. See how each one is recorded and who can see the record.
3. Put the stock away and scan or confirm the locations the way your staff would on a normal day.
4. Walk the floor. Pick 15 or 20 items at random from the system, go to the location it names and check the quantity. Write down every miss.
5. Run a short count on one aisle and see how a difference is entered and approved.

At the end of week one you should be able to say whether the system tells you the truth about what you hold. If it does not, the rest of the pilot matters less.

Week two: orders out, invoice sent

Week two tests the work that makes money.

1. Create orders by hand and by import, whichever your client really uses. Include a single item order, a multi line order and one that cannot be filled in full.
2. Pick and pack with the staff who do it every day, not with the person who is most comfortable with software. Watch where they hesitate.
3. Ship, and check the paperwork. Does the carrier label print, does the tracking reach the client view and can the client see the order status without calling you?
4. Make a change mid order. Cancel one, change an address on another. Real clients do this every week.
5. Run a dry billing. Take the activity for the two weeks from the system and price it by hand against the rate card, then compare the result. Anything that happened on the floor but is missing from the system is a billing gap.

Set the measures before you start

Decide what a pass looks like on day one, not on day fourteen when you are tired and the vendor is friendly. Examples that work for a small operation:

• Every carton received can be found by location.
• The random location check in week one has no unexplained misses.
• Every order shipped with the right items and the right label.
• The dry invoice matches your hand calculation, or every difference has a known cause.
• A new staff member can pick an order after a short walk through, without help from the vendor.

Keep the list short. Five measures you will really check beat twenty you will not.

Keep a log

One shared page, with a line for each problem: what happened, who hit it, what the vendor said and when. After two weeks you will see which problems were learning curve, which were missing features and which were things you can live with. Learning curve problems fade. Missing features do not.

Run the review

At the end, sit down for the meeting you already booked, with your log and your measures.

• Which measures passed, which failed, and why?
• What would you have to change in how you work to use this system?
• What does the next step cost, in the same terms you were quoted, with double your clients?
• Did the vendor answer quickly when you were stuck? That is a preview of support after you sign.

If it passed, ask for the final price in writing and have the export of your pilot data sent over, so you know the exit works before you need it. If it failed, you have lost two weeks, not a contract. If you are still working out what you need before you try anything, start with what a small 3PL needs first, and for the stages after a pilot, see the first 90 days.

Questions people ask

How long should a 3PL software trial last?

Two weeks is enough to push one client through receiving, putaway, picking, shipping and a dry billing run, if you plan the days in advance. A trial with no plan can run for a month and still prove nothing.

Should I use real stock in a trial?

Use real stock only if the trial runs alongside your current method, so a mistake in the new system cannot stop an order. Otherwise use a pretend client with a small set of real looking items.

What should I agree with the vendor before a pilot?

Who owns the data you load, what happens to it afterwards, whether the pilot carries no charge, in writing, who answers your questions during the two weeks and what the next step costs if you continue.

How do I know a 3PL software pilot passed?

You set the measures before day one: for example every received carton found by location, every order shipped with the right items, and a dry invoice that matches the one you calculate by hand.