If you have a product that can be used in many different ways, do not let your customers’ creativity run wild. Be very prescriptive and don’t be afraid to say no to a bad integration. It can be an unrecoverable mistake.

In 2006, I started TrialPay as an alternative payment system – “pay by trying something else.” I first used the concept for my own shareware products, then at SiteAdvisor, and after incorporating the company and speaking at the Shareware Industry Conference in Denver (yes, there was such a thing) I signed up WinZip (one of the top downloaded products in the world at the time) and some “big” shareware players. These nimble customers followed my guidance and put us as an option on the downloads page (“get WinZip for free with TrialPay!”) and soon we were processing millions of dollars of sales.

We raised a Series A on the progress; it was time to go to the big leagues. Match.com, Yahoo (big back then!), Skype, Electronic Arts, Intuit, so many companies with low conversion from free->paid for their products. Big name VCs like Battery and Index and Ron Conway who were making intros left and right. What could go wrong?

Everything. I met with the head of Yahoo Personals who loved the concept but said it would be hard to integrate (since we were effectively a payment option) — “scarce engineering resources.” But here’s an idea! Rather than changing the website to put TrialPay on it, let’s send an email to customers as a test. Rather than integrating as a new tender type, let’s give customers a coupon code. Rather than showing this in the US, let’s pick a region with more flexible testing schedules. Make everything as easy and fast as possible. I liked the way this person was thinking!

So what happened?

TrialPay was featured in a standalone email message (with very muddled messaging, given Yahoo legal had to sign off) to something like 1000 expired users in foreign countries. Maybe a few of them proceeded, which was free money for Yahoo! But in the exec’s mind, “TrialPay didn’t work” and it was an uphill battle to say “no, that’s because you put us in the wrong place!!!”

This story played out many times before I got the message. BigCo proposes an integration that’s easy FOR THEM but virtually designed to fail. Stupid me says yes because I want the logo/customer. Or maybe smart me says yes because I don’t want them to work with our competitor!

We churned through multiple big company logos — while meanwhile watching little, unheard of customers with millions of dollars of revenue apiece.

The lesson is to CLEARLY PRESCRIBE how a customer should use you. As in:
Option A
Option B
Option C

if they want Option D or E, just say no. Do not be accommodating. Be firm.

This is, incidentally, the setup for account management. Give your account managers a bingo board with all of the desirable integrations/placements and pay them for filling out that bingo board.

Do not be afraid to say no. I won’t say there’s no recovery from a “we tried Startup X, it didn’t work” but it’s very hard.

And now as a VC, I like things that are “set it and forget it” — often very boring systems as a result.