Change Everything By Changing One Thing

As usual, I came across another great post by Seth Godin regarding pricing models.

Netflix revolutionized the movie rental business simply by charging by the month rather than by the rental.

Apple changed the music industry with iTunes by letting me download only the songs I like for just 99 cents each rather than pay for the whole album.

Nokia is changing the music industry once more with their “Comes With Music” phone. It comes with unlimited music downloads from the 4 top labels for a fixed subscription fee rather than me paying for each download.

Adobe is now offering a free online version of Photoshop that lets me do almost everything I can do with the professional full priced version.

In all these examples … changing the pricing model changed everything.

What if EDA companies priced their tools like a subscription? A fixed fee gets you any tool they have as long as you only use one license at a time.

Or pay-per-use? Need to do some regression sims? Just pay for the number of sims you do.

Or what about tiered pricing? Full price for the 1st license and less for each additional license?

Sure, these are pretty whacky ideas for an industry as mature as EDA. But maybe that’s what the industry needs.  Someone to change everything by changing just one thing … their pricing model.

harry the ASIC guy

Tags: , ,

7 Responses to “Change Everything By Changing One Thing”

  1. James Colgan says:

    Hi Harry,

    I believe that the business of EDA is where the real challenge is for the industry moving forward. We need some creative thinking and some perceived risk taking to inject some vitality back in. I say “perceived”, because as you point out, while these types of models are new to EDA, they’re not new per se.
    Interestingly, three out of the four examples you use are new entrants into a particular market. This could be where the shake up occurs. Maybe we’ll have one “Adobe” that is bold enough to venture into a new business model with one of its products, but it would be interesting to see if there are a couple of startups out there that are looking to differentiate themselves via their business model AND their technology.

    – James

  2. James Colgan says:

    I forgot the shameless plug – if anyone out there is interested in pursuing different models leveraging web technology (SaaS/Cloud Computing) drop us a line. Xuropa has the technology and experise to help.

    – James

  3. JohnB says:

    I’m not an EDA business guy (just a customer), but do we have to figure this out new for ourselves? Are there better precedents in other industries? You’ve cited a few, but they are consumer-oriented and hard for me to think about in EDA.

    How are other industries more successful with highly specialized software? Would Mechanical CAD (Autocad), Mathematica, and SAP be worthwhile considering? Are they more successful and what do they do differently?

  4. Jeremy Ralph says:

    With, our SaaS register automation tool, we are doing it based on a user subscription year. We have also been thinking about doing per-unit pricing. With online web applications, I think per user pricing is pretty normal. Heck, we’ve even been offering a free promotion, whereby qualified customers get a free user for 6months for production use.

    For SpectaReg Onsite, it seems the big companies have trouble with anything that is not licensed per concurrent seat, FlexLM style. We approach these situations with an open and flexible mind.

  5. harry says:


    Thanks for the link to Paul McClellan’s EDA Graffiti blog. I had never read it before, but many of the posts on there, including the one you pointed out, are very interesting. I’ll definitely put it in my RSS reader. Good stuff!


  6. Dave Murray says:

    Hi Harry,
    I agree we need a new approach to EDA licensing. In Duolog, we looked long and hard at how our tools could be made available to a wide range of users and we have implemented a licensing model we call “Complexity Based Licensing”.
    We were determined not to limit the tools’ capabilities by having “Lite” or “FPGA” versions with some important features disabled. Instead, we’ve based our licensing model purely on the complexity of the systems that the tool can process. Take, for instance, our register management tool called ‘Bitwise’ ( Bitwise comes out of the box with all features enabled – no limitations. When licensing, we simply enable the tool to work with the complexity of designs that the user is capturing. In the case of a register management tool, the natural unit of complexity is the bitfield. IP or FPGA developers will generally be capturing systems with no more than 300 – 500 bitfields. Sub-system Integrators will assemble systems with, perhaps, 3,000 – 5,000 bitfields while those integrating larger systems will require unlimited complexity.

    This allows a chip design company to pick and choose the best set of licenses for their projects. For example, a large chip design project may require ten IP licenses, two Sub-System and just one Unlimited license. We have been getting very positive feedback on this model because it allows customers to maximize the ROI from the tool by only paying for the complexity that their users require. With 12-month licenses starting at just $5,000, the ROI is clear and immediate. The register bitfield has a documentation view, an RTL view, several verification views and several software views so if you can automate all of these views your ROI is instinctively tied to your bitfield. Why would we limit the ROI by reducing the feature set? In addition to the built-in features, Bitwise comes with an extensive suite of generators (Documentation, RTL, C, SystemC, SystemVerilog etc.) that our users can modify, or add their own, without limit or cost.

    We are implementing a similar complexity-based scheme on our other tools such as Spinner, our I/O fabric generator. In this case, the ROI, and thus our licensing metric, is tied to the number of pins in the design.

    – Dave

Leave a Reply