Subscription Pricing Starting Strategy: start high and lower, or start low and raise?

With a subscription based web application aimed at businesses: Is it better to set the initial price at the high end of what you expect people would pay and lower it if it becomes necessary, or should you start with a lower price and then try to increase it once more features are added?

The main dynamic that I've having an issue with is:

• High price -> fewer customers but "more" profit.
• Lower price -> more customers but "less" profit.

I know I can do some spreadsheet jiggery and come up with some graphs and such but I'd really like some practical advice from people that have done it before.

Then again does price matter this much when selling to businesses?

Christian, I was a VP of marketing for major corporations for over a decade and can assure that starting high and going low is much easier than trying to raise your prices after the fact.

The exception is if you announce up front that this is a "one-time offer" an "introductory price" or some other qualifier. If you do that, then you must follow through and, at least for a time, sell the product only at the higher price.

Your credibility, your brand, it at stake.

Final thought, if you continue to add value to your product then starting with a low price and moving up (the value chain) can work. That's what Honda did years ago. They entered the US auto market with low end products and then, over years, began building larger cars with more features, which allowed them to move up the price/value curve.

A couple of blog posts that you might find useful related to pricing.

• QuickSprout: Customer Are Always Right, Except When They Are Wrong
• OnStartUps: How To Price Software Without Rolling The Dice

Overall, Steve is right on. You can never raise your price on the same product. Having a tiered pricing structure is much better (Take a look at 37 Signals Pricing for Highrise ). Have a look at the QuickSprout article. In a nutshell, they found that the low price offering garnered more customers but less revenue.

Pricing does matter B2B. What you really need to understand is the value that your product brings to the customer and price is accordingly. Don't worry so much about the number of customers. Worry more about creating a competitive product that customers want to buy. Loyal, return customers are far better than just low end customers that suck up support bandwidth for minimal return.

Raising prices isn't that hard if you're sensible about it. Specifically:

1. Warn existing customers before it happens.
2. Grandfather existing customers with old prices so they don't care.

New customers won't know it was ever less.

Generally lower prices are more honest at first because the product doesn't do much yet and probably has bugs. But it's not dishonest to charge more first either.

A nice middle-ground is to technically charge high but have deals all the time. That way you can just stop having deals whenever you feel like it.

Here's a very good reading on pricing.

“Don’t Just Roll The Dice: A usefully short guide to software pricing” It discusses at one point about finding the best pricing that will get you the most revenue.

I personally think that increasing the price while adding more features would be the way to go. But it also depends on what competition you have, I guess. Maybe the competition wants to compete on price. Then you'd have to take into account other factors than the number of features.

Joel Spolsky says in his interview from "Founders at Work" that they raised the price of FogBugz quite a few times and each time the number of customers increased.