Pricing for the ISV

I recently read through an article about how IBM i ISV’s are apparently wrong in the way that the charge for their products. While I do have some sympathy and do recognize that some pricing practices are not appropriate in today’s global competitive market, I also feel the argument is incomplete.

As an ISV we develop product in the hope of being able to sell it but that comes as an upfront cost before you even make the first sale. The cost of developing a product is not insignificant, so you need to make sure that the market is big enough to cover those costs and more (yes profit is key for survival).

Here are some of the arguments made within the article.

“Provide straight-forward, flat pricing with no haggling”

The poster goes on to state that it should be a single price regardless of the size of the system or the activity on that system such as active cores or number of users etc.

Well I have yet to walk into a customer and not be expected to haggle on the price! Its human nature to want to get a better deal than is originally placed in front of you, it makes you feel better when you get that better price and can take it to your manager.

Don’t play favorites? I have already blow that one out of the water above, some customers demand more just because of who they are. A large blue chip company brings with it more opportunity to up-sell other products and they tend to have professional negotiators so they tend to get the best deal! But they do tend to be happy to pay more for the right solution and because they have bigger operations the cost of the purchase is spread over a much wider base. Maybe they are not favorites but they certainly get more attention.

If I walk into a small client who only has a small 1 core system and less than 20 users what do you think he is going to say when he finds out he is paying the same price as the big guy down the road with 64 active cores and 3,000 users?? I am pretty sure he is not going to feel like he was dealt a good hand!

I do agree that the price has to be fair and I do not get involved with adding additional cost just because the system has multiple LPAR’s or more active cores, the price is set by the IBM tier group for all our products. This should reflect the capabilities of the system to handle much more activity and therefore spread the additional cost over a much larger base.

“Freeze software maintenance when a customer purchases your software”

Nice idea but totally impossible to meet! If the developers I employ would accept a pay freeze at the time I employ them and the suppliers I use (that’s everyone who adds cost to my overhead) would freeze their costs maybe I could do the same. In reality its never going to happen. There are too many influences that affect the ongoing cost of support, that cost has to be passed on to the users of the software. The users always have the option of terminating support, they can stick with what they have as long as they want. However having said all of that, we have not raised our maintenance costs to our customers for many years, we are just making a lot less money than we should..

The question about including the first year of support with the initial price is mute, add them together and they are a single price. Some companies like to allocate license fees and maintenance to separate accounts so they like to see it broken out. We don’t stop improving the product on the day we sell it, its a continuous cycle so if you need support or want to take advantage of a new feature we just added maintenance is an acceptable cost.

“Make it easy for customers to upgrade their hardware”

If a client stays within the same IBM tier they should not pay additional fees to move their application, however if they move up a tier group they should. This all comes back to the discussion above about how the initial cost should be set.
We do not charge for moving the product to a new system in the same or lower tier, we don’t add a fee for generating the new keys either, but you must be up to date on maintenance or expect to pay something even if it is just a handling fee.

IBM charges for additional core activations which we do not agree with but when you look at the capability of today’s systems and what activating an additional core can do for adding more users its not that simple anymore. What I certainly do not like with IBM’s fees are that we are billed for the extra cores PLUS we have to by additional user licenses etc as well if we add more users! That is just gouging at its best!

“Don’t make your customers pay for capabilities they don’t need”

Its easy to say modularize your application in such a manner as to allow the clients to pick and choose what they want. The reality is some options just can’t be left out because of dependencies on other options. Another problem is clients now have to make a decision as to what the are going to purchase, how many times have you bought a product with more options than you need just because the price point for the additional features is so minimal? The client is not paying for something he does not need he is paying for a product that meets his requirements and maybe more at a price that is acceptable. If the price is wrong your competitor will make the sale not you.

Purchasing decisions are not always made for the right reasons, we are human and we make decisions based on our own set of principles. Even companies which have purchasing policies in place that should reduce the effect of human emotion it will still be a part of the sale.
Trying to predict a clients choice is near to impossible even if you have a relationship with the decision maker, other factors will always come into affect. All you can do is put forward what you feel is a fair and acceptable price and be prepared to haggle. Trying to put a set of rules such as above into the process is only going to end badly!

Chris…