The last two weeks have been mind-blowing. In succession Apple launched the iPad, announced iPhone OS 4 (used in iPhone, iPad and iPod touch devices), changes in Apple’s license agreement sent the developer world (particularly Adobe) into a tizzy, and then, as a topper, Twitter bought the maker of Tweetie, a Twitter client.

There is amazing power and tremendous risk in partnering and these four events clearly show the risks and rewards.

Let me be clear that this is no manifesto for going alone. The reality is, as a start-up, we are really made or broken by our initial luck in partnering, whether that’s using someone else’s hardware and operating system, an initial customer, or a strategic relationship with a big partner.

On the positive side, Apple’s iPhone OS and devices running it have brought smartphones to millions of people who never used one before, opening up purchases of third-party apps in a way that has never been done before, and that has created a number of newly wealthy developers (not me). On the other hand a change in two paragraphs in Apple’s license agreement has potentially wiped out a handful of other companies’ businesses.

Twitter has also brought something new to millions of customers, namely a new communication platform. And in one move it might have wiped out thousands of other companies’ businesses by making a choice and buying one of them.

Infinity Softworks has benefited in some ways, too. Almost unrecognized in the iPhone world, being there at launch coupled with Apple’s decision to ship no calculator with the iPad has raised powerOne revenues to breath-slightly-easier levels.

I have learned to ask clear questions about these relationships: Is the partnership short-term or long-term? Is the partner committed to the relationship? Is it likely that we will get in the way of their bigger strategic goals?

Sometimes I guess right. We had a lengthy and fruitful bundling relationship with Palm and other Palm OS licensees, but we were dumped when Palm used an alternative product instead. We were also making tremendous headway in education when Palm decided there was no future in handhelds (ha, ha), shifted focus to smartphones, and fired their education team.

Even when we fit perfectly, it’s not always easy to see how the relationship will work out. (It’s also not every day a company makes a billion dollar investment in a market and then backs out on a dime but that is just the epitome of how poorly Palm’s been run.)

If I’m on the web I’ve got to ask myself these questions about Google, Twitter and Facebook. If I’m reliant on Apple, I’ve got to ask these same questions of it.

And then I have to hope I’m right.

It is very rare that I have an experience that ends in such utter failure as the one I just had with Vonage.

I signed up for Vonage two years ago, liking the idea of getting a phone line that was less expensive than the landline offered by my local phone company and not quite ready to dump our phones for cell phones. But over the past two years, our need for a home phone became less and less, getting to the point where all of our time spent on the phone could easily be done within the cell minutes.

Unlike others on the web, I had a few connection issues but always had a good experience when calling their support or using their systems. The problem didn’t arise until I tried to cancel. The word deceptive practices is an understatement.

So I called two months ago to cancel, waited on the phone a short time for a representative. When he came on, he told me he could handle that for me, no problem. He politely offered a few options to keep me on, as I’d expect. No company likes losing a customer and they were actually good deals. He offered to cut my bill in half for a year, for example. But when I politely said no, he said he’d have the phone disconnected on December 3, giving me the interim time free in case I wanted to change my mind.

I confirmed all this, got a confirmation number and hung up, thinking I’d been taken care of.

December 3 came and went. We still had a dial tone. December 4 came and went. Still had a dial tone. December 5 I called to find out what was going on.

The first person looked at my account and immediately transferred me to a manager. The manager was cordial, apologized for the mis-communication and explained to me that by accepting the free month, they required me to call back again to disconnect! News to me, as I would not have accepted the free month if that was the case. She immediately disconnected the number, gave me a refund for the month (although she still charged me $.90 or so, which I can only assume was the 1-day fee for the month) and sent me on my way.

I don’t mind having to jump through a hoop to disconnect by making the phone call. Nor do I mind that they spend a bunch of my time trying to get me to stay on. What I do mind, though, is the deception. The key to customer satisfaction is pretty simple, I think. Say what you’re going to do, do it, then tell them it’s done.

Now, Vonage lost me. (And apparently I’m not alone.) If I ever need a home or business number again, it sure won’t be with them. And before, when I recommended them to friends and associates, now I’m telling everyone who reads my blog — thousands of people every month — not to touch them with a ten foot pole.

Starting a company is a giant science experiment. We develop a series of hypotheses and test them in as controlled a way as possible, making changes to the hypotheses as we go.

Generally, those hypotheses fall under a number of basic categories:

• Product X solves ______ problem.
• Product X has ________ customers.
• Product X will make money by ______.
• and so on.

If a company is successful in answering these questions, they tend to grow big. And those that can’t answer the basic ones don’t grow at all (or very little).

In the 12+ years I’ve run Infinity Softworks, I’ve observed more than a few stupid ideas get a ridiculous amount of money. I thought large amounts of funding at an early stage of a company’s life — even $500,000-$2 million — was the wrong approach. Too many companies spent beyond their means or used it to grow a business before they knew what business they were in.

But I now believe I was wrong. Early stage funding is not intended to grow the business; it’s intended to answer the hypotheses. And by focusing on answering the hypotheses, it positions the company for growth.

So how much money does a company need? The answer is enough money to run the hypotheses and prove that there’s a business that deserves more funding (or can grow from revenues). I underfunded Infinity Softworks with our latest endeavor with web, iPhone and FastFigures, and put ourselves in an interesting position as our first hypotheses proved false.

This leads to obvious second question: where does the money come from? Is it internally funded or does a company raise capital from angels and VCs and others? That answer is it depends. I have a good friend who runs Creative Algorithms, a husband and wife team creating apps for smartphones. They’re able to run their hypotheses out of their own pocket and, given the size of the business they are trying to grow, it wouldn’t make sense to raise funds. I, on the other hand, raised funds. The size of the experiments — coupled with the business potential — required it.

I believe strongly that the web — and the traditional PC and mobile app businesses that it touches — is going through an evolutionary process. Many of the old business rules are broken — things like how to market, how to charge, the value-price relationship (more expensive means more value). Ten years ago we could make a connection and say “such-and-such a product was marketed and sold this way so we will have success by doing the same.” But now that is no longer true. Everyone’s trying to figure out these new rules from scratch.

As we move FastFigures forward, soon overtaking the core capabilities we offered in our powerOne applications and then turning our attention to the new and exciting features we have planned, this disparity between the way we used to do it and the way to do it now are hitting home.

So I asked for help. And Albert Wenger at Union Square Ventures was kind enough to provide some clarity. My first question was regarding the so-called freemium model, where some features are available for free and some features cost money, and where to draw the line between the two. I was also struggling between companies that seem to go without charging customers and some that charge right away. Albert gave me excellent advice:

As a general rule I would say if your business has a network effect, i.e. more users make the service more valuable to each user, then you should worry primarily about adoption and about pricing later.  If that’s not the case then you want to get pricing in place right away.  It is best to charge for those features that will discriminate between business users who can charge to a business credit card and everyone else.

This is simple and straight-forward, dare I say an axiom for the internet age. If there are two customers — consumer and business — then businesses will pay subscriptions and consumers won’t. So the portion of the service that attracts consumers should be free for them to use and (potentially) monetized some other way, whether that’s advertising or virtual land or  transaction fees or something else.

I find this enlightening. It encapsulates so many of the web businesses I see more succinctly than I have ever seen it before. Thanks, Albert!