An Executive's Pocket Guide to Organisational AI

The Rise of SaaS

There was a time before cloud computing and software as a service (SaaS) when organisations installed and ran software locally. Significantly more companies had customized software that they had built themselves, or had customised for them.

With the advent of the internet, large software companies moved to the cloud. "Is it Cloud"? became the first question that was asked when evaluating new software. Now the first question is, "does it have AI?"

For the software vendor, there was an advantage to this. It removed the complexities of developing and supporting applications on the desktop, it eliminated the complexity of adding an e-commerce site to an in-house database. For the organisation, it also removed the cost and expense of managing servers, backups, remote access and technical support in house. The web browser became the default user interface for software, with all its limitations.

Since then, the web browser as user interface has been the baked in assumption for the subsequent 25 years. Web as the default, with a downloadable application if available, effectively a skin over a web-based experience.

And one of the side effects of that assumption is that organisational silos and operational complexity have arguably increased rather than been eliminated. And businesses have accepted that. They've accepted the challenge of having a separate ERP to a CRM. Of running marketing and e-commerce as separate solutions. And what we've seen more and more recently is the continued fragmentation of an organisation's tech stack. There might be "an app for that", but each of these software solutions, each of these applications is another licensing component. Multiply that by the number of users in your organisation and it starts to add up.

And the frustrating thing is because these applications are cloud, because you're locked into a vendor solution, the ability to design and deploy something that's simple but totally aligned to what you need it to do is always one small step, one licensing tier out of reach.

This is common thread in conversations we're having with many organisational leaders - how can I rationalise my tech stack, reduce my licensing spend and increase productivity? They sense that their current architecture is allowing productivity to silently leach out of their organisations. Information in disparate systems. Spreadsheets papering over the cracks. Automations grappling with the limitations of each vendor's solution.

The Return of the Desktop

Now AI is happening, and it's gone mainstream. And one of the most interesting things is that it is happening on the desktop. While many traditional SaaS companies appear to be struggling to gain traction with in-built, bolt-on AI (we suspect some companies are having their Kodak moment right about now), Claude, Chat GPT and Gemini have arguably unlocked the desktop at an individual level.

Without jumping from application to application, you can just get things done. Can't find that email? Just ask. Need to fix some data in a CSV file? Just ask. Need to build a presentation? Just ask. Goodbye Powerpoint, we have Claude deck envy. And all from a single chat-based user interface that feels comfortable.

So the potential is there, and this is where companies are seeing the first benefits from AI.

We're seeing genuinely transformational solutions designs becoming possible across applications, partially due to the genuinely new capabilities that both AI and automation now make possible, but also partially due to the appetite for employees to leverage AI to build solutions to problems directly. And this is largely happening on the desktop.

Let's look past the obvious organisational questions this raises, when Olivia vibe-codes a solution to solve her own personal productivity problem and then shares it with Liam. Let's instead welcome Gen Z to the workforce. And is the solution they've cooked up "cloud"? Probably not, the entry-level apps that AI will devise are typically on the desktop.

The human brain is an amazing thing. It looks for the most efficient solution to a problem. So the paradigm shift we're seeing in companies is not necessarily due only to the capabilities of the technology, it's also a result of employees having a willingness to use the new AI tools to find grass-roots solutions to problems that question or ignore allegiance to a 25-year-old software marketing-mantra-turned-paradigm.

Where to Start?

Most of the conversations we have with businesses begin the same way.

"We know we need to do something with AI. We just don't know where to start."

AI presents enormous opportunities, but not every opportunity is worth pursuing. The challenge isn't implementing more technology - it's identifying where it will create the greatest impact for your business, while respecting the context of your existing processes and systems.

We recommend starting by looking across the business and identifying areas where where AI can create meaningful impact and/or add capacity to your team.

Rather than aim for a "big bang" design that could take months to finalise and potentially years to implement, we recommend identifying areas where there are quantifiable benefits.

For some companies, it's reclaiming hours every week through automation.

For others, it's giving their team AI tools that improve quality and consistency.

Sometimes it's connecting systems that have never spoken to each other.

Sometimes it's redesigning a process that's frustrated people for years.

Identify the opportunities worth pursuing and prioritise them based on business impact, implementation effort and expected return on investment.

Initially, look for projects that can be implemented within a reasonably short time frame and achieve a measurable outcome.

This allows you to develop organisational competency progressively, to build organisational muscle in using AI.

A final thought

AI will continue to change. The models will improve, today's tools will be replaced by better ones, the way customers and employees use technology will evolve.

The more enduring opportunity is organisational.

Businesses have spent years accumulating software, systems, data and processes. AI gives us an opportunity to reconsider how all of those things work together - and, perhaps more importantly, how people work with them.

The organisations that benefit most from AI won't necessarily be those that make the biggest investment or adopt the most technology. They will be the ones that learn how to make their organisational knowledge accessible, connect it to the work being done, and progressively build the capability to use it.

That doesn't require having the final architecture worked out today.

Start somewhere useful. Build something that makes a measurable difference. Learn from it. Then build the next thing.

Organisational AI isn't a project with a finish line. It's a capability.