In Stay Sharp Episode 145, Why Nobody Can Predict Their Software Costs Anymore, co-hosts Juliann Grant and Jonathan Scott sit down with Paul Empringham, the VP of CAD Simulation and PLM at NYBS Consulting. Paul’s previous episode, What Leaders Need to Know About Software Licensing, focused on tips for license management and optimization. This time, he’s talking about what the future holds for enterprise software licensing.
What Lies Ahead
Paul ties the huge push to get customers on cloud or platform as a service (PaaS) to software vendors’ desire to have customers using the latest versions and upgrades they’ve spent a lot of money on. He explains, “Providing a SaaS or PaaS technology allows the vendor to deal with all of the upgrades in the background. They should make it invisible to the end user.” While there are a lot of positives to PaaS, a downside to users is that with everything controlled by the vendor in the cloud, it’s more difficult for a company to understand their software usage, something Paul explained in the previous episode is vital to optimizing your licenses. Not to mention, you might be paying for entire packages of functionality you’re not using, without a lot of insight into that reality.
Paul Empringham, the VP of CAD, Simulation, and PLM at NYBS Consulting
Paul also predicts unexpected costs ahead with all the changes in what’s being provided and how people are working with data, for example, AI and APIs. The idea of a token per use—or token burn—does the opposite of incentivizing engineers to fail fast and try more, as opposed to the “all you can eat” version which means you can try and fail all you want. Jonathan says, “If every time I try, it costs me something, I stop wanting to innovate.” Paul agrees, noting that some companies are burning through tokens and budget with junior engineers playing with AI tools. He says, “You want your seniors using it effectively and setting the framework and the roadmap and the guardrails around how they should be using them.”
The Future of Licensing is Tokens
A lot of vendors are already moving to a token system, typically a token pool. In that structure, a company has a collection of tokens, and each user accessing an advanced feature uses a specific number of them, which are then put back into the pool when users close the feature. It’s a revolutionary licensing change, Paul says. “Before, people were reluctant to buy these additional really expensive licenses because they’d only use them for a few days of a project, whereas now you don’t have to. It’s allowed customers to really have free flow across all of these advanced features.” And it allows companies to use whatever product their teams need depending on their workflows that day, without worrying about having the correct license or not.
But there’s a challenge, Paul says. “For the first time in our lives, I don’t think we can predict what our software’s going to cost us, and that’s never happened before.” Adding to the unpredictability are different approaches and confusion around tokens and AI tools. Typically, AI has one-use tokens you burn up and software has token pools—but some engineers might use AI assistants (token burn) inside digital engineering platforms (token pooling). “We might be in this situation where engineers are going to be constrained and have a cap on what they can spend on AI assistance within CATIA. Who knows?” Paul says.
“Companies just need to understand that license models always change,” he offers. “It’s inevitable. As industry changes, as the way we want to consume these things change, the license model will change. You just have to have the best visibility of your usage and what you’re trying to achieve as much as possible.” If you’re not already tracking license usage and user consumption, start doing so, because the companies that understand their usage data will be able to move at speed. There are tools that can help you, even by tracking open browser tabs if your software is in the cloud. He says, “Most companies out there are going to have a hybrid approach. They’re going to have perpetual licenses, subscription licenses, and people working on a cloud platform. It’s confusing. It’s difficult. They just need to have as much information as possible.”
Learn More About What Licensing Will Look Like
The full podcast contains more discussion and details on a variety of topics, including the lesson Uber learned about tokens for AI usage, how AI is going to increase the speed and cadence of software releases from vendors, why Paul says token economics is already replacing token maxxing, and more.
Be sure to check out Stay Sharp Episode 145: Why Nobody Can Predict Their Software Costs Anymore and join us each week for a new podcast.
Follow the Razorleaf Podcast, Stay Sharp in Digital Engineering on:
