"Off the shelf" SaaS products are dying. Long live the open product framework

TL/DR: SaaS is the model of the past, not the future

For decades, companies have put up with bloated generic software that never quite “nailed it”, because coding and product development was hellishly expensive.

Not any longer.

"Off the shelf" professional SaaS products are going to become obsolete, thanks to AI coding tools making bespoke development more cost-effective, faster and more available to every organisation than before.

The future is open product frameworks - modular, transparent systems customers can adapt themselves or with their choice of delivery partner. Closed software is now unacceptably rigid, insecure and restrictive.

My company Waivern is betting on this shift entirely.

Bottom line: SaaS companies that don't become flexible and open platforms will be thoroughly eaten alive.

In today's competitive business world, building strong customer relationships has never been more important. Customer Relationship Management (CRM) systems play a crucial role in organising customer data, managing communications, and streamlining sales pipelines. With the integration of Artificial Intelligence (AI), these systems have advanced far beyond their traditional capabilities. AI-driven CRMs now enable businesses to automate tasks, deliver personalised experiences, and predict customer behaviour with remarkable accuracy.

Now for the elegantly written long-form version!

I’ve been a product manager for more than 20 years. Several months ago, I took a leap and started my new company Waivern, focused on removing the paperwork pain points of regulation, while keeping governance outcomes robust. Since then, I’ve found a technical co-founder, found and hired a first engineer and we’ve jointly worked out our go-to-market strategy.

There’s one thing that’s missing. Finished products.

Are we going to be a consultancy shop instead? Nope, at least not primarily.

I can see many of you scratching your heads. Surely at least a SaaS product business model is just “table stakes” to get going in this space?

Maybe for some. But maybe not for Waivern. What has changed?

Manually coding working software products was (and remains) a tricky task. And demand for the skills to do it has outmatched supply for most of the last 40 years. But the demand is there, because the time/money saved and quality improvements in commerce and in life that software has enabled have been exponential. Just look around. Software has indeed, as Marc Andreesen predicted in 2011, eaten the world.

Driven by the similarity of many common organisational tasks and the scarcity of coding (and latterly, UX design and product) talent, the main arc over the last 20 years has been in “off the shelf” software products, both delivered on desktops, via web-based services and mobile apps. All these initial tasks were intensely generic and commonplace: accounting, payments, document writing, presentation slides, communications, marketing/sales, distribution and customer support. Almost all the giants of tech in our modern world continually iterate to provide their best version of those products, with ever increasing breadth of features and with those features made ever more accessible through UX design. Microsoft. Stripe. Google. Adobe. Meta. Salesforce. Amazon. Intercom. All variations on this theme.

Larger organisations, “enterprises” in sales parlance, also wanted and were able to afford deep customisation of these products. Those came at eye-wateringly high cost, both up-front and in ongoing maintenance. But the ability to customise products to the exact needs of global businesses spawned an ecosystem of consultants and advisors to help meet their needs.

Having bitten off and digested the big “generic” tasks and the “high value” customed enterprise software markets, software products then went granular, building offerings to attempt to streamline the shared needs that individuals in their professional life might have.

Need to get business cards printed? Need to get a conference organised? Need to buy new widgets from suppliers around the world? “There should be a site or an app for that!” became the battle cry of tech entrepreneurs from the days of the first internet bubble in 1998. That has continued right up to today.

This world of “off the shelf” software applications for deep or broad niches is where most of today’s product managers, designers and engineers have cut their teeth. So long as the product broadly meets the need, the subscription price is acceptable and if (this is a big “if”) the market is large enough, then there has often been enough value in “niche-but-standardised” product offerings to justify the purchase for the customer and to provide a return for the software provider. This is the pot of gold at the end of the venture-funded or bootstrapped rainbow that SaaS vendors pursue.

What has suddenly changed? The emergence of high-quality AI-driven coding tools.

Tools that non-technical product managers like me can use to create working software or adapt existing code. Other tools are emergent to assist with UX design for non-designers. And the foundational LLMs will write a reasonable Product Requirements Document. Even if all these AI-driven outputs are imperfect today, it doesn’t matter; over time, they’re going to massively improve. In so doing, they’ll the release latent demand for the business gains that deeply specialised software tools can create, even if those opportunities are obvious only to few people.

But what does it mean for significantly more complex organisational problems where the service is supposed to accelerate an otherwise manual process spanning several teams with different perspectives, all in a potentially unique and complex organisation?

There are many examples of software like this. It could be software for media content creation, traversing writing, editing, print/online formatting, as well as print/digital advertising teams. It could be software for logistics, combining sales team data with production, packaging and distribution team workflows. Or, in the case of my company, software that helps remove the friction, delay and inaccuracy in compliance workflows across product, engineering, marketing, compliance and legal teams. Suddenly, companies where previously only “off the shelf”, expensively customised or rigid SaaS offerings were the only choices, now will have a cost-effective ability to build solutions that work exactly as they would ideally want.

Will organisations do it all themselves? Some, like Satya Nadella of Microsoft, see AI agents eventually replacing SaaS products. We think his concept is a strong candidate for the endgame, but specialist areas (like the legal contracts space, like privacy and AI compliance or product standards conformance) are domains where the wrong decision can cost billions, can indeed cost lives. Handing those decision rights to an agent will be, well, ultra-high risk for most companies.

This is why “products” in the specialist spaces are not going to just go away. But they will have to evolve, becoming modular and open to AI-accelerated adaptation. This will have huge implications both for product teams and for the business models of SaaS companies. I’ll come back to that again in a future post.

I and my team’s core thinking on how this impacts digital product offerings is unpacked here.

Rigid productisation = no longer viable: The age where organisations, even small ones, were prepared to accept rigidly defined products that can’t be adapted, customised and reconfigured by the customer or their delivery/advisory partners – rapidly, easily and cost-effectively - is coming to an end. With the cost of coding falling exponentially, the “viable” market size from the customer’s perspective may end up being one. One engaged user that wants his product “just so”, now willing to pay a newly low price. AI coding tools opened this door.

Specialist product frameworks, rather than “off the shelf” and rigid products, are the future: many business challenges are – even in their simplest possible state - bigger than any one user’s needs. Software products to solve these challenges will need to be modular, open, flexible and scalable. And business is a cross-team-exercise: all those aspects need to be able available to their ecosystem of end-users, engineers, designers, customers, decision-makers, consultants and lawyers. Especially in compliance tasks that traverse the organisation.

Closed and non-transparent software = unacceptably high risk: organisations, even small ones, are increasingly aware of security risks in the supply chain of software vendors. They’ve seen peers go down, paying ransoms to release data or paying through the nose to consultants to rebuild a virus-free version of their tech stack in a hurry. Now, they want full transparency on how software works, who built it and where it takes their data. Successful hackers in the closed software ecosystem have, unfortunately, opened this door.

Humans + bespoke tools + specialist judgement in big decisions = good: specialist expertise and human judgement still counts and won’t be replicated by machines, simply because humans will not trust “black boxes” when significant decisions are being taken. This doesn’t necessarily mean that AI can’t be involved, but the explainability and evidential trail needs to be crystal clear, accurate and easily verifiable. The almost infinite variability of circumstances and contexts in the real world is what is keeping this door open.

So, what can you take away from all this?

The summary of the outcome is that SaaS, as a business model, is entering into a period of flux. Flux in terms of the attributes of the products it needs to deliver. Flux in how it needs to engage with customer and partner organisations. Flux in terms of the organisational configuration and revenue/cost models that a SaaS business will need to adopt for sustainable success in the future. Many SaaS business will find themselves architected for a future that no longer exists. The ones that thrive will be built for constant adaptation.

This is a philosophy that we’ve taken on board from day one. Waivern will definitely have products, but our products and our organisation around them will be different. Our products, our frameworks and our offering to the market will be anchored on these tenets and empowering to every participant in the compliance and compliance tools ecosystem; customers, partners and potentially even for competitors.