Fog Creek Software
Discussion Board




Pricing Products

Could you give some advice is working out how much to charge for a software product?

Matthew Lock
Saturday, February 21, 2004

That's a really hard question but a good one!

Your goal, I assume, is to maximize income. Assuming software with a COGS ("cost of goods sold") of zero, income = revenue = price * quantity. Lowering the price theoretically increases the quantity sold, while raising the price reduces the quantity sold but can increase the overall revenue.

One theory says that if you imagine a chart with revenue on the y-axis and price on the x-axis, the chart is a mountain-shaped bump. You have to figure out if your current price is on the left side of the mountain, in which case you should raise it, or on the right side of the mountain, in which case you should lower it.

The only way to answer that question is to test different prices. One good way to do that is to set a high initial price and compare the response rates to various discounts. Direct marketing is very good for this, or you can have a few "one time sales".

Meanwhile, there are several sneaky tricks to keep in mind.

(1) As a small software company you are NOT going to compete on price. Don't say to yourself "we have the ability to make this thing cheaper than anyone else, so we should compete on price." Large companies have all kinds of economies of scale that you don't have, and you're not going to be able to sustain a price advantage. Find a different advantage.

(2) Pricing sends messages. Expensive products (like Perforce, $795/copy) "seem" like higher-quality products than cheap ones (Vault, $199/copy) even if they aren't. People believe that "you get what you pay for." A new high price may INCREASE your unit sales if the old price sent a message of "Cheap!"

(3) There are natural limits. Home/consumer users won't spend more than about $20 in cash, or about $50 on a credit card. At most corporations the low level business managers have the right to spend some fixed amount of money on their credit cards without getting any permission, and that amount is usually either $300 or $500. If you set your price at $600 you are arbitrarily losing a lot of these sales because now the low level manager needs to go through purchasing and get approval. For years UserLand Frontier has been priced at $899 ... just out of reach of everybody with a corporate American Express, and too low a price to justify a salesperson. The sales of that product were not very good. If Frontier were my product I would have kept it priced at $899 (to send a message) but had all kinds of discounts to get the actual cost below $500. Now it fits on your credit card *and* you think you're getting a good deal.

(4) With software sold in corporations, as soon as your price gets up in the $3000 level, the amount of approval it needs is so absurd that you are not going to sell products without a salesperson making a few visits. Hiring the salesperson, sending them out to make presentations, hotels, airfare -- now it costs $50,000 to get the sale done just in sales closing costs. That's why you see a lot of software products at $100,000 and a lot under $3000, but anywhere in-between and it's impossible to make sales. (It's sort of funny -- these big corporations create so much bureaucracy around purchasing in order to protect themselves against losing money, but they just force the vendors to spend a fortune on salesforces, which results in vastly higher prices for the big corporations).

(5) The kind of people who give advice to startups will tell you again and again that almost every startup makes the same mistake: pricing their product too low.

(6) Recently a company with a $29 consumer product asked me for pricing advice. $29 on a consumer product is a money-losing proposition. The cost to acquire a software consumer is around $100. And consumers are only willing to pay around $30-$40 for software.  So it seems like consumer products are money-losing propositions right? Mostly, but not always. The most successful company at the $30-$40 price point is Intuit with Quicken and TurboTax. When Intuit sells you a copy of Quicken, they're going to make you register (perhaps using a rebate offer). Then they're going to try to sell you Turbo Tax every year. Then they're going to sell you preprinted checks for your inkjet printer. Then they're going to charge you $10 for state tax returns. Over the lifetime of a customer they'll make a couple of hundred dollars, so acquiring a customer for $100 is worth it. That's why Intuit can do consumer and you can't: they're really a stationary company!

Anyway, to the company with a $29 price point, I suggested:
(a) immediately change the price to $79 (crossed out) limited time offer only $29! You don't want to send a "cheap" message.
(b) Figure out how they're going to sell a continuous stream of products to those consumers worth at least $100 in the future.

That's enough for now...

Joel Spolsky
Fog Creek Software
Saturday, February 21, 2004

>One theory says that if you imagine a chart with revenue on >the y-axis and price on the x-axis, the chart is a >mountain-shaped bump.

...just like the Laffer curve!

Steve
Saturday, February 21, 2004

This is exactly the question that I have been struggling with for the last few months. What about a subscription based service, Joel? Assuming you kept your customers and you charged in the $30-$50 range per year, you would need at least 3 years to make up that $100 in acquisition cost. Any thoughts on how consumers react to that pricing model?

Paul Lawler
Saturday, February 21, 2004

"Common knowledge" in the industry is that consumers won't accept subscription-based software unless there's some "service" aspect to it (like an online data component) or something that changes every year (TurboTax).

Common knowledge might well be wrong, but I haven't seen a whole lot of evidence to convince me it's worth the risk.

Joel Spolsky
Fog Creek Software
Sunday, February 22, 2004

The Laffer curve is no laughing matter.

(I couldn't resist.)

Warren Henning
Sunday, February 22, 2004

Why does it cost $100 to attract a new user? Is this figure based on advertising costs?

Frederik Slijkerman
Monday, February 23, 2004

(PS you're more likely of getting your question noticed if you post it in a new topic)

There are lots of different calculations you can do to get to the $100 figure.

One of them is $1 cost for a direct mail piece, 1% typical response rate.

You can also look at CPMs for ads...

It's just a rule of thumb.

Joel Spolsky
Fog Creek Software
Tuesday, February 24, 2004

Nothing between $3000 and $100000?  Apparantly you've never had to buy typical EDA applications that EEs and CompEs spend most of their day using.  I would bet that all of the Schematic Capture, PCB layout, and CAD programs I've ever used professionally have run between $5K and $10K each.  And that's usually for a one year lincense - you get to pony up again in a year if you want to get the bug fixes and brave the new features.

Kelly
Friday, February 27, 2004

What's the NPV of a stream of $10,000/year products for, say, uh, twenty years?

Joel Spolsky
Fog Creek Software
Friday, February 27, 2004

It is interesting what Intuit is now doing to "enhance" their revenue stream.  We use Quicken 2000 for our bills and just received a notice that effective April 8, 2004, our version of Quicken will no longer support online banking, but if we upgrade to the new version, it will.

So one week before tax deadline, we are being "forced" to upgrade to we lose functionality in the program that we have already paid for.

Joe Booth
Friday, February 27, 2004

As part of a successful, small company that survived the bubble, our products have the unique advantage that each customer usually purchases more than 1 license of our product(s). As a result, we priced their initial purchase at a high number (say $100), then provide a 50% discount for all subsequent purchases. The initial purchase covers the CAC and everything else is gravy.

If you're lucky enough to sell a product that people *need* and not just *want*, then you don't have to worry too much about price. They'll buy it at almost any cost.

In these scenarios, the margins are ridiculous.

I-405
Friday, February 27, 2004

But what if your product is multiuser enabled, how much to charge per user since the range 3000-100.000 is out of question?
If you charge 300/user, you already cross the line.

Emil
Saturday, February 28, 2004

+ and you have 10 user;
10users * 300 = 3000
Sorry for 2 posts.

Emil
Saturday, February 28, 2004

The company I work for pulls down about $10million / year selling software and services at $15k/year.  It works for us. 

Our competitors are also priced in this range.


Saturday, February 28, 2004

I know of one software company that sells their products on a subscription basis - Stardock. They seem to be doing fairly well at a price point of $50 to start and $35 per year thereafter. It helps that they are continuously improving their products, and that if you drop out, you get to keep using that which you have paid for previously (the $35/annum number gets you new product releases and the latest upgrades).

At any rate, they've fished me out of over a hundred bucks over the past 3 years. A price I am happy to have paid for the continuing support and customer experiance I've gotten out of them.

Silvercat
Friday, March 05, 2004

$3MM/year, average sale: $18k.  Profit > 20%.

You overgeneralized just a bit.

Dave
Monday, March 08, 2004

"$29 on a consumer product is a money-losing proposition. The cost to acquire a software consumer is around $100."

You'll forgive me for pointing out the exception that perhaps arguably proves the rule, but Ultra-Edit seems to do quite well at $30, and my UE version 7 has never lacked in a way to make me feel like upgrading to [insert latest version here, 10 at this writing].  Panic's Transmit ftp client is only $25, yet they seem to be eeking out a living as well, and there's sure as heck never been a good reason to pay twice for Transmit other than the port to OS X; eternal free upgrades are part of Panic's "charm".

So your answer is obviously painted with too broad a brush.  Perhaps it costs less to reach savvy computer users, which both of these examples nearly require by definition (all things considered), but there obviously are people charging less than $100 for consumer software and doing better than breaking even -- without subscription services or other add-on sources of revenue.

R Bailey
Monday, March 08, 2004

It's a shame to have to sell a product that's worth $995 to some people who may care about lots of advanced features for $99 because you don't want to alienate 90% of your customers, who only want the basics.  You can release separate editions for each of them (see Windows XP Home/Pro).  Software is very amenable to this type of segmentation.  You can even release free trial versions which are limited in functionality or effective use dates.

When you think about it this way, you don't have a single "mountain", but a range of peaks for different segments.  Your promotional or discount strategy may also be very different for different segments.  The actual price point, sales and marketing strategy, and other pricing tactics will vary by situation, but effective segmentation and segment pricing will likely greatly increase your profitability.

Reuben Swartz
Tuesday, March 09, 2004

Began working on this topic this year.  Was I surprised at the misleading info from the accounting community!  People (including me until a while ago) think they should throw in all their costs.

Setting MINIMUM price happens on a TOTALLY different spreadsheet than your profit/loss page.

To price correctly you must include all your AVOIDABLE costs and include NONE of your fixed costs or past work that already been paid for.

So, if you build a custom app for a client and then use that R&D to create a product for many more clients, don't put in your costs of the past-work.  It's been paid for (sunk) - let it go!

Include sales costs (new wages, new printing, new monthly web hosting etc.) if you must buy them for this (new) project.  If any of it has to be bought anyway, regardless of this particular product, then it's not avoidable and therefore is NOT included.  Include the avoidables only.

Once you know your MIN price, you then have to look at the market.  Totally agree with Joel's remarks on pricing and perception - too low is as bad as too high.

Tim Thomas
Wednesday, March 10, 2004

This is terrific information.  My company (at this moment, "my company" = "me after hours") is getting ready to put a product aimed at small/mid-sized businesses running SQL Server out there, and I've been debating on how much to charge per license.  (License here means "SQL Server to be monitored" -- the product does SQL monitoring and baseline analysis for non-DBA types.)  After reading all this (and some other research as well), I'll probably go with a $200 per license price point, with a 3-license bundle for the low, low price of $499.

I'd be curious to hear other specific examples of how you guys came up with prices...

John
Saturday, March 13, 2004

Joel, of all people, I'm surprised that you think anyone is fooled by those "for a limited time only" deals.

anonymous
Sunday, March 21, 2004

For a limited time deals have a substantial impact on sales.  I've seen it time and time again.

Note that sometimes it may not be the impact you want.  If you get into a pattern of having reduced price periods, many customers will simply wait for the next one, causing your cash flow to have major ebbs and essentially setting your price at the discounted level.

The latter isn't always a bad thing if you plan for it.  People associate a higher price with quality and capability.  If you can get it in their mind that your product is a valid competitor in the $10,000 segment, but available for a mere $7,500, then you can benefit.  But after a while, you'll be seen as a $7500 product.

John
Friday, March 26, 2004

What about value-based fees (a la Alan Weiss) ?

Does anyone have experience pricing software based on the product's value to the client?

That is, the price is some factor in ROI. For example, if my product can save/make a company $1M in a year, then why not charge $500,000 for the software, and the client gets 100% ROI in 6 months?

If you're not selling value, then what are you selling? Time?

Value-based fees seems to apply well to professional services (like public speaking, mentoring, focus groups, etc.) but what about software projects?

Mark A. Richman
Tuesday, April 06, 2004

Hello!!! There is a word for saying the product costs $100 but that it is on special for $29 -- fraud.

False advertising charges have been brought against such marketing. You need to at least set the price to $100 for a while before you can "discount" it.

Doug
Tuesday, April 06, 2004

"Hello!!! There is a word for saying the product costs $100 but that it is on special for $29 -- fraud."

Take a look at the FogBUGZ pricing screen. Note how you are eligible for a 30% off "competitive upgrade" if you are currently using any bug tracking system, even one you wrote yourself. That is what Joel is talking about here.

Competitive
Friday, April 09, 2004

Reuben Swartz raises an interesting point(above):

"You can even release free trial versions which are limited in functionality or effective use dates."

I'm currently writing a business plan, and I've been toying with the idea of distributing free trial versions of the product. The one thing that is making me shy away from the idea is the lack of 'buy-in' from potential customers. My product will probably take customers 10-20 hours of effort to learn to use. I'm worried that a customer wouldn't make that investment in a product unless they'd already invested some money in it.

Is this fear justified?

Nick Hynes
Sunday, April 11, 2004

Let me ask a corollary to the original question: How did you decide to price CityDesk at $299?  I just checked out the starter edition with the idea of maintaining my personal website and weblog with it (currently using blogger for the weblog and maintaining the rest of the site manually).  I love the product, but $299 is just too steep to ease the maintainence of a personal website a bit.  I'm an acknowledged cheapskate, but I would think the pricepoint for users like me is around $90-$100.  Plus, I would think that getting a bunch of weblogs on CityDesk would provide some good advertising as well.  Are you just not interested in this market?  At $299, I'm guessing you're after the enterprise market pretty much exclusively, and not interested in the personal-site market at all.

Joe Ganley
Monday, April 12, 2004

Hi!
In the sale of business software, as in the whole trade, exist two agents; the seller and the buyer.
The two have their characteristics.
For each software exist X01 potential buyers [in  certain market].
The need of buyer to buy the software is X02. This need varies from 0% to 100%.
The usefulness of the software for the buyer is X03 [does installing the software reduce the costs?].
Desire [emotional] of buyer, to install advanced  software is X04 [this factor is the most important; the buyer buys motorcycle BMW or Harlay-Dafson, when a bicycle would be enough!].
On another side the seller:
Number of the competitors is Y01.
Quality of the software, in relationship the buyer's real and emotional needs, is Y02.
Quality of the software, in relation to the competitors, is Y03.
Sale effort [investment in publicity,...] is Y04.
Effort of sale of the competitors is Y05.
Price of the seller software [ours] is Y06.
Price of competitors is Y07.
Capacity to assist the requests is Y08. The requests need to be assisted [volume of the requests is not sale].
All know existence of the factors above. Nobody, a priori, knows their values.
Profit and Losses depend upon these factors.
The academic display is always subjects to critics.
But  is always exact if we have exact data. And that is  hopelessly pursuit.
0tto, oz@oz.pro.br
Tuesday, 2004/05/25

konstantin otto christof botschkowski
Thursday, May 27, 2004

Consider VALUE-BASED PRICING.  First understand what criteria the customers are using to make the purchase (features, compatibility, etc) and which criteria are most important.  Spread 100 points across the key criteria.  Conduct a survey, interview using, run focus groups, etc. 

Next, score (1-10) your product and other competitive products against the criteria to see how much value each product is delivering. 

Finally, compare (plot) the prices charged for these products against the valus associated with them.  This will give you a general idea of how to price your product based the product's value from the customer's perspective.

Steinar Knutsen
Wednesday, July 14, 2004

*  Recent Topics

*  Fog Creek Home