Kill the honeymoon: How to run a vendor selection that actually works

So how do you avoid the honeymoon trap and choose a solution that still makes sense once the real work begins?

The biggest trap is jumping straight into feature checklists. It feels productive, but it’s upside down. Start with the business outcomes you want and work backward:

• Do you want to raise service levels and stop losing sales?
• Free up working capital locked in inventory?
• Automate tedious planning work to give planners time back?
• Model scenarios (tariffs, lead times, footprint changes) fast and often?

Document these outcomes in plain language and get executive alignment early. Then map each proposed requirement to one of those outcomes. Separate the true must-haves from the “nice-to-haves,” and kill a few darlings. If a requirement doesn’t directly support an outcome, it isn’t valuable, but noise.

Bonus: translating outcomes into a simple business case forces clarity. Will this affect P&L (e.g., gross margin, freight, scrap) or the balance sheet (e.g., working capital)? If you can’t trace it to one of those, pause. You’re likely adding complexity without impact.

Every vendor can say, “Yes, we handle seasonality”, but the question remains: can they handle your seasonality? A checklist won't answer if your pain is forecasting in a heavy seasonal business where small low-season shifts cascade into painful high-season misses. You have to see it in the tool.

As Benjamin Obling, COO at PERITO IBP, Roima, says, “Meetings should be at least 50% of the time in the tool.” If a session is mostly slides and talk, you get a podcast – not proof. Set a house rule: if less than half the meeting is in the product, stop the meeting and rescope.

You’re selecting a system to run your business, so here’s how you can design the “show, don’t tell” moments:

1. First, give vendors your three hairiest scenarios. Ask them to walk through setup, data inputs, decisions, and outputs live.
2. Ask them to explain failure modes. For example, “What breaks first? What do planners see? How do we recover?”
3. Request a short, recorded demo per scenario so evaluators who missed the meeting can review what actually happened, not what was promised.

Once you’ve narrowed the field to two or three serious candidates, resist the temptation to pile up glossy proposal decks. The absolute truth doesn’t reveal itself in polished documents; it shows up when things resemble everyday reality. That’s why a strong selection process should run on four parallel tracks.

The first track is functionality. Instead of letting vendors talk in generalities, hand them 10-15 of your toughest, most representative use cases and see how the system actually handles them. These could be specific SKUs, sites, lead times, and constraints. This lets you sit in the driver’s seat and navigate the tool.

The second track is proof of concept. Give the vendors a slice of your real data; the messy and imperfect stuff you deal with daily. Then watch what they can build in a handful of workshops. The goal isn’t perfection; if they can get you to 80%, you’ll quickly learn how they collaborate, how fast they move, and how they problem-solve. Think of it as a preview of life after the honeymoon.

Next comes IT and integration. Many projects stumble here, so cut off the shortcut early: no “temporary Excel uploads.” Get your IT team and the vendor’s experts in the same room and map the entire data flow from end to end. Let them sort out authentication, connectors, refresh cadence, error handling, and monitoring. Decide upfront where transformations will happen and how the system will be operated on day two, not just day one.

Finally, there’s finance. Big promises like “cutting dead stock” look good on slides but are not a strategy. Finance needs to test the hard numbers. How will the solution impact working capital, gross margins, and cost assumptions, and how will integrated business planning outputs connect to budgeting and forecasting? Finance must see how the solution supports their needs.

It’s in these four conversations that cultural fit really emerges. Not in a tick-box that says “yes, we’re a good match,” but in the way people listen and problem-solve once the discussions get concrete.

A strong vendor selection process doesn’t feel like a honeymoon. It feels like work. Data gets tested. Demos get challenged. IT opens the hood, and finance pushes on the numbers.

That tension is healthy. Your goal shouldn’t be dazzled in week three, but to still be glad about your choice in year three.

AI has made it absurdly easy to produce 50 pages of requirements and 50 pages of vendor replies. That gives more paperwork than necessary, so keep your RFI short and pointed:

• What outcomes are you buying? (3-5 bullets.)
• What are the 10-15 critical capabilities tied to those outcomes? (Describe with examples, not generic terms.)
• What three scenarios must work well? (Define success criteria and acceptable shortcuts.)
• What constraints matter? (Security, hosting model, integration patterns you already use.)

In the RFP stage, concreteness is what matters. Replace “Answer these 200 questions” with “demonstrate these 12 flows.” Record the sessions and use the recordings as the source of truth for scoring.

The harsh truth is that failed implementations rarely happen because of technology alone. They happen because the selection process stayed in the honeymoon phase too long, where everything looked perfect on slides but was never tested against reality.

By anchoring requirements in business outcomes, insisting on “show, don’t tell,” running four parallel tracks, and keeping RFIs and RFPs lean, you massively increase your chances of making the right call. You’ll cut through the noise of glossy promises and see how a solution will perform in your world, with your data and under your constraints.

And that’s the point. Vendor selection isn’t about who tells the best story in week three but about who you still want to work with in year three.

Take the time, run the hard conversations, and kill the honeymoon. Your business and your future supply chain resilience will thank you for it.

Roima Intelligence helps manufacturers and supply chain companies turn technology selections into lasting results, not just promises. Contact us.