Want AI efficiency? It won't just happen..

In the last twelve months AI is the one subject EVERYONE is talking about in relation to business. You would be forgiven for thinking that every organisation has already delivered a full suite of automated processes from top to bottom. 80-90% of posts on LinkedIn is talking about it and the number CV's citing AI expertise seems to double by the day.

Whilst there are a host of tools now available to businesses to automate processes using agentic AI, adoption is not where you might think given the pace of development across the tech industry. My experience working with a number of medium and large organisations during 2026 indicates that fear, apathy and a lack of appetite for the perceived investment costs are slowing down development of AI strategies in a lot of organisations. In truth many companies have simply provided colleagues with access to co-pilot in the hope of becoming more efficient as an organisation. Spoiler alert - it's not going to happen.

Firstly - I hate to be blunt but unless colleagues find something genuinely valuable to do with the time they have saved from using whatever AI tool they now have access to then they are just being paid for doing less........ whilst the tab for token usage still has to be paid. In a growth business this isn't a problem as you can expand without needing to increase headcount, but for most organisations in the current climate that isn't the case, so I ask - do you know how busy your team were were pre AI? Do you know how much time has been saved? And are you confident that data protection is being upheld and 'AI slop' avoided? Something to think about.......

Secondly - AI use by individual colleagues may make the 'super users' very efficient but it can lead to fractured processes across teams. Efficiency is all about standardisation of processes and elimination of waste. Businesses need to maintain ownership and control of AI enabled processes. Ideally all process changes should be planned, tested and then rolled out in a structured approach to embed usage. Unless everyone works in the new efficient way - you fail to reap the benefits. This is basic change management but the word on the street says it isn't always the way things are being done.....

Finally (because I don't want to overload you on a blog post), there is the holy grail of Agentic AI and full process automation. I had a client this year state an ambition of training AI chatbots to replace 50% of the customer service team. They lost the appetite when it became apparent that they had to either a) Implement only on new customers on a new ledger or b) spend 6 months cleaning up the data on the existing ledger. AI agents can follow simple standard operating processes very well, but if you haven't been running your business on SOP's and your data is all over the place (be honest now!) then those agents are going to have a problem.....

To conclude - I am a in awe of what AI can bring to organisations, but like any major business transformation there will be a lot of disappointments and disasters. My point here is to highlight that unlocking the value of AI requires you to understand how it can be used right across the business to support strategy whilst maintaining your core values. There may well be low hanging fruit that can be implemented quickly whilst full AI automation will need more thought to get right. Whatever you do though - do it with intention. Understand where you are today and make sure the benefits are defined and ultimately delivered in a structured programme of work. Without it you may find the token spend leaves you in a worse position than you started!