Why most pilots fail (and the one trait of the ones that don't)

The mistake: spraying instead of concentrating.

The instinct is to pilot Positive Pay across as many business clients as possible to "see how it goes." That's backwards. A spray pilot produces noise. You get a handful of clients half-using the product and no clear read on whether the program works.

A community FI ran a 3-business pilot. One succeeded and became the in-house success story. Two un-enrolled. The variable wasn't the product. It was pilot design and business selection.

The three rules of a real pilot.

Rule 1: 1 to 3 businesses, no more. Concentration beats spray. Every business in the pilot needs to be fully invested and fully supported. You cannot give that level of attention to ten.

Rule 2: Screen for a dedicated exception reviewer. Same trait from Lesson 2.3. The single biggest predictor of pilot success is whether the business has assigned someone to own decisioning. Screen for it before you select a pilot business, not after.

Rule 3: Set a 90-day evaluation window. Less than 90 days and enrollment, decisioning, and exception handling haven't settled, so you can't read the signal. More than 90 days and you've effectively bought the product without negotiating leverage.

How to pick pilot businesses.

Look for three things: meaningful check or ACH volume (so there's something to protect), a recent fraud scare or a risk-aware owner (so they're motivated), and operational maturity (so they can assign an owner). A business with all three is your ideal pilot candidate. A high-volume business with no owner and no risk awareness is a pilot that will fail and teach you nothing.

How to read the results.

At the 90-day mark, look at decision rate first. If your pilot businesses are decisioning their own exceptions, the program works and you can scale. If they're not, don't scale yet. Diagnose whether the problem is training, the platform, or business selection, then fix it before you roll out to your full book. Scaling a broken pilot just multiplies the breakage.

Do this

If you're about to pilot, name your 1–3 pilot businesses and confirm each has a designated exception reviewer before you go live. If you can't confirm it, you've found your first problem.

What's next.

Lesson 2.5 covers the friction that quietly kills onboarding: file formats.

Self-check

3 quick questions

What's the most common pilot design mistake?

A Piloting for too short a time

B Spraying across too many businesses instead of concentrating

C Choosing businesses that are too small

D Not involving the vendor

Correct. Spray pilots produce noise. 1–3 businesses with full support gives you real signal.

Not quite. The most common mistake is spraying across too many businesses. 1–3 fully-supported businesses gives you real signal.

How long should a real PP pilot run?

A 30 days

B 90 days

C 6 months

D 1 year

Correct. 90 days is long enough for patterns to settle and short enough to maintain negotiating leverage.

Not quite. 90 days: long enough that enrollment and decisioning patterns settle, short enough to maintain negotiating leverage.

What should you do if your pilot businesses aren't decisioning their own exceptions at day 90?

A Scale anyway, it'll improve at volume

B Don't scale yet — diagnose and fix first

C Cancel the program

D Switch vendors immediately

Correct. Scaling a broken pilot multiplies the breakage. Diagnose whether the problem is training, the platform, or business selection, then fix it.

Not quite. Don't scale yet. Diagnose whether the problem is training, the platform, or business selection, fix it, then scale.

← PreviousTraining your business clients Next lesson →Removing file-format friction