How to price software/Saas? | Hacker News Reader

How to price software/Saas?

network.businessofsoftware.org

50 pointsnreece17 comments

The best advice I have ever received: charge more. Whatever your first impression is, it is not enough.

You do not have enough confidence in your application. You see how much of a hack it is. You worry that someone could duplicate it all in a Jolt-fueled frenzy. And you're wrong.

2.5 years ago I was terrified of the prospect of taking money and asked for some folks opinions on whether my software was worth $15, $20, or $25. In my heart of hearts, I was sure it would not sell even at $15. Someone convinced me to charge $25, at the least. I wish I remember who -- the decision has made me over ten thousand bucks.

And I'll be raising prices one of these days. People will pay them -- that is all I need to know to justify a price increase.

I remember a great story from Sandra Kurtzig's book "CEO: Building a $400 Million Company from the Ground Up". She demoed her package to her first customer. He was wowed and asked , "How much?" She was worried about overpricing, but she said "$50,000" anyway. And when he didn't blink, she added, "per module." And when he still didn't object, "per year."

It really depends on who your customer is. Many managers have discretion up to $1000, so a few hundred dollars is not a hard sell. Above that gets into the corporate approval cycle. If you are selling enterprise software, you can get away with $2500 per seat, but then your selling cycle becomes months long.

If you are selling to consumers, maybe only a few dollars or a few $10's of dollars. Think TurboTax. Developers are the really cheap bastards, maybe $.05, But they can probably talk their managers into spending a couple of hundred dollars.

The perceptions of the customer makes a huge difference. A friend works at software company which sells software to corporate and government clients. Their pricing varies from $50,000 to $300,000 with the size and type of the organization they are selling to. They originally tried to sell to larger customers for the same price as smaller, but they didn't get any sales until they jacked up the price.

The price charged for our own product is now quadruple the "OMG! Nobody will pay this!" price I introduced the software at a few years ago. Every time the price has increased, sales and revenue have increased as well.

The best suggestion I've read (some time ago on the BOS forum) is to keep raising prices until you start receiving complaints about the cost. If nobody is complaining (not even the cheapskates), you're not charging enough.

Another bonus from higher prices: You no longer get those nickle-squeezers whose only criteria was finding the cheapest option. To generalize from my own experience, those folks are colossal pain to deal with.

If at least 50% of your prospects aren't complaining about the price, then you're not charging enough.

I didn't sell my first unit until I raised the price.

> Do you look to the market to price (external input) or do you primarily look internally (we think it's worth xyz and adds $x value therefore..)?

Whatever you do, don't look internally. Your estimate of what it's worth is always awful. Why? Because prices aren't measured on a linear scale. Just to take an example from every day life: books. A lot of people make impulse purchases of books in the $10-$20 dollar range. Let's call one of those people Bob. It doesn't matter much if Bob is short on cash, a $10 book is easily bought. Often, Bob won't read the book, or won't finish it because it may not turn out to be very good. And curiously, Bob will not save to buy a $30 book, that he's sure he's going to like, because it's simply priced outside of the arbitrary range. In our minds, good books are not allowed to cost more than bad books. If we see two John Grisham titles laying next to each other in the bookstore, and if they're priced differently, what do we assume?

a) one book is much better/enjoyable/exciting than the other book

b) one book is newer/more popular/hardcover/etc

Think about this. It makes no sense that (b) is the right answer. And yet it is.

Mostly, I think, you have to find the right price point by experimentation. Don't be afraid to raise the price when you release a new version, and don't be afraid to lower the price with holiday certificates. Then measure and refine. A lot of people are afraid of raising prices because they might lose customers that way, but compare that to all the money lost by giving away your product too cheaply.

Eric Sink wrote an excellent primer on software pricing and you can read it at http://www.ericsink.com/bos/Product_Pricing.html

That was a good read. He says: "We want to charge the maximum amount that our customers are willing to pay." At the risk of being pedantic, the more precisely correct answer is to charge each customer how much that customer is willing to pay. All other things being equal, that would be optimal pricing. All other things, of course, aren't equal.

That is why versioning exists. It lets you better approximate getting each person to pay how much they'd be willing to pay without all the hassle. He has that much right.

The part he neglects to mention is why three versions is a magic number. He hints at the reason but it should be stated more often. People do not like complexity so having -too- many choices becomes overwhelming. Three lies in a sweet spot where people tend not to be overwhelmed (small, medium, large).

There is a more important reason, though, and that is the aversion to extremes. People are averse to choosing any extreme without good cause. Three is the minimum number to take advantage of this. If you had just two versions, the best way to force people to upgrade to the better and more expensive version is by simply introducing a new super-expensive third option. This causes angst in people who might otherwise buy the cheapest.

Hence why everything comes in small/medium/large, etc. This is also why the "default" position is always the middle, and you have to ask for the others specifically. Use words for the lowest like "basic" or "starter" etc. In general, versioning in threes is done to force "most" people into the middle option and let you profit more on your "power users" in the high-end (for example, I am a known "power-user" at Wendy's).

Here are several articles i've collected over time about pricing:

http://dl.getdropbox.com/u/113637/softwarepricing/pricing%20...

http://dl.getdropbox.com/u/113637/softwarepricing/How_Do_You...

http://dl.getdropbox.com/u/113637/softwarepricing/Econometri...

http://dl.getdropbox.com/u/113637/softwarepricing/VBSP.pdf

http://dl.getdropbox.com/u/113637/softwarepricing/PricingMet...

http://dl.getdropbox.com/u/113637/softwarepricing/SW_Pricing...

First, look at your variable costs. These are the things that go up as more people use your product. In SaaS, that means bandwidth and storage. Price accordingly. I think $0.25 per GB for both storage and bandwidth is a reasonable assumption (yes AWS can give you lower, but you have to deal with request pricing and S3 vs. EC2 persistent storage, etc.). Don't price yourself lower than that. If you want to give people 5GB of storage space and 50GB of transfer, you need to charge them at least $15. Those are real costs.

Any free version will be crippled because storage and bandwidth cost money. That's ok. Give a few MB away for free with unlimited time (unless you have a SaaS that isn't going to be crippled by low storage allowances). That way, people can sign up without a credit card and really use the system without any hassle. If they really want it, they'll need more storage and pay. Most of the accounts will probably end up being browsers who don't really want/need your service, but that's ok.

Take into consideration your market. Is this a market with lots of users? If so, go for lower margins, you'll make it back in scale. If this is a niche, that niche will pay up and you'll have fewer users so you'll have to charge more.

Most importantly, make it obvious. Don't make it so that someone looking at your rates wonders why the pricing scheme is the way it is. Make it obvious how to sign up and terminate. Make it hassle-free. It makes you look professional. Don't do paypal or some other form of payment simply because it's easy/cheap. It's a hassle. It looks unprofessional. Make it easy for users to get to what they want - hopefully it's your product, but if not they're still people that deserve the respect of ease of use.

Thanks! Didn't know the BoS site had a forum as well.

I'm personally against the idea of selling software. It doesn't have any real value because it costs nothing to reproduce. Sure, the development costs money, and you need to live, but there are other more morally correct ways of obtaining cash than selling something that should be free.

For more, see fsf.org

Is it moral to charge for food? Water? Heat? For tutoring or teaching? If I've learned something in my years and I want to make the world better by passing it on, through software or any other means, I'm morally obligated to find a way to sustain it. I personally also feel morally obligated to educate my children so that they can do the same.

The best model man has yet developed for allowing people's talents to be used to their fullest is business.

I disagree with your implication that I am somehow immoral for selling software. In fact, I am a little annoyed.

(Edit: but I did upmod because it's an interesting POV that deserves to be heard.)

Stallman sold copies of Emacs for a pretty penny above reproduction costs.

It just so happens that he give the source alongside.

Freedom 0-4 does not say you cannot sell your product.

Being against selling software because it costs nothing to reproduce implies that selling other more "tangible" products is only priced to recoup reproduction cost. Pick any product that you aren't opposed to selling and think about what percentage of the price is there to recoup production cost.

Value is in the eye of the beholder.

If people are willing to pay 100 dollars for FooBarSoft version 12, it means that they value it at least 100 dollars.

Reply on news.ycombinator.com