You gotta’ do you
In the ideal company, everyone works in perfect harmony to give customers the best possible experience and product. You are outperforming the competition, and your primary concern is how to remain modest during a period of outsized success. You may even be contemplating lavish holiday parties.
The people in your company move in an artistic flow, and the business ballet that emerges is elegant and assured. You operate in a way that makes you the obvious choice for customers, having made an integrated and consistent set of decisions that competitors struggle to imitate.
And then you wake up.
You do, in fact, have strong products and loyal customers. The people in your organisation are capable and conscientious, doing good work within real constraints to deliver a solid customer experience and keep the company running.
But the system as a whole does not feel like a ballet. It feels more like a scramble up a steep slope. You are generally moving upward, but missteps send rocks tumbling, and occasionally those rocks land on people who did nothing to deserve it.
You are not doing “you”.
A central issue is the set of constraints imposed by your software landscape. You have an ERP, a CRM, finance systems, workforce management, manufacturing systems, and HR platforms. The enterprise resembles a soup of semi-connected islands, each useful in isolation, none quite coherent as a whole.
This is still preferable to the alternative, which is having no systems at all, with departments hoarding knowledge in documents and spreadsheets.
But these systems demand conformity to their models. When your organisation does not operate the way they expect, workarounds proliferate. Vendors provide configuration options and external interfaces to automate around the edges, but the core functionality remains stubbornly rigid.
So, in the effort to adapt yourselves to the systems rather than the reverse, everyone builds workarounds.
Large organisations often attempt to force the fit. They may spend as much, or more, on configuration and customisation as on the software licenses themselves. Even then, manual processes persist, stitched awkwardly between automated steps.
This is why, as a customer, a transaction that follows a company’s default path feels smooth, while anything requiring special handling triggers phone calls, delays, and apologies. On the other side of the interaction, people are compensating for systems that do not bend easily.
Smaller companies experience the same problem differently. They export data, live in spreadsheets, and manually reconcile reality across tools that were never meant to work together.
Automation tells a similar story. Historically, automating higher-level business processes has been expensive. Small companies often avoid it altogether. Large companies invest heavily, gain workflow efficiencies, and then discover they have built systems that are rigid, slow to change, and fragile in the face of exceptions.
Despite all this automation, every organisation ends up with a layer of people above the systems. These people extract data, assemble spreadsheets, produce reports, and coordinate the real processes that actually run the business.
This is where generative AI begins to matter. These systems can interpret ambiguous inputs, work with incomplete data, and initiate appropriate workflows. They are more tolerant of exceptions. They can absorb a meaningful portion of office busywork, creating space for the human collaboration that truly drives outcomes. And they can be shaped quickly, using adaptable prompts rather than brittle code.
In short, they make it easier for you to be yourself, and to execute your strategy in a way that aligns with how your organisation actually works.
Which is precisely why generic AI will not be enough. You cannot simply buy it and expect it to fit. You will need to craft something that reflects your organisation, your constraints, and your ambitions, because every company is different, and should be.