MANAGERS ARE GROWN-UPS TOO: TELL THEM WHY AI NEEDS IM
- Jonathan Stuckey
- 1 hour ago
- 9 min read
audience: Information management advisors, AI consultants, Project managers
author: Jonathan Stuckey
Every one wants the shiny remote-control car for Christmas: The one that moves faster, does tricks and is fun-to-use. AI is the business version of the car: it finds the answers, reduces the grind and helps people make better decisions - way faster! Yet so many organisations are trying to launch the shiny AI-buggy down a driveway full of potholes, loose gravel and yesterday’s rubbish.
Yet talking with Information and Knowledge management advisors all I hear is lots of talk about permissions, versions, retention and metadata as though they are technical chores that the business ought to fund out of duty. They are not. They are the conditions that determine whether AI becomes a useful business capability or an impressively fast way to reach the same old confusion and total your RV.
Stop Treating Information Management Like a Punishment
The business does not need another warning. It needs a reason to care.
Unfortunately IM has an uncomfortable communications problem. For years, information professionals have approached business leaders with warnings: storage costs are rising; permissions are inconsistent; version control is unreliable; audit trails are incomplete; records are scattered; content is duplicated; access needs tightening; difficult to manager retention...
None of these statements are wrong, they are all legitimate concerns. They're just not helpful in building a business case for investment - not when everything is hygiene.
These phrases are the language of information managers speaking to themselves. No one else talks like this - sorry, they just don't. To a business executive trying to improve customer outcomes, reduce operating cost, make decisions faster or wanting to deploy AI meaningfully, they sound like a request to fund housekeeping. Necessary perhaps, but not urgent, and certainly not exciting.
So, when we put in AI Foundations proposals (read: IM clean-up) and it fails to secure support, we blame a lack of executive understanding. That's unfair, and arguably our own fault - and yes, I have been found guilty-as-charged of this on occasion too.
The problem is not that business leaders, managers and everyday users are incapable of understanding information management or don't have "Good behaviours". The problem is that we have failed to explain the relationship between:
their everyday behaviour,
the state of their information environment and
the outcomes they expect the business to achieve.
We have described the potholes in the way, but not explained why the business wants to drive a car down a nice and smooth road in the first-place. hint: more speed, less pile-ups.
Treat Leadership like adults. Tell them WHY
It’s worth remembering that adults teach children about consequences through familiar stories.
At first, Father Christmas brings presents for good behaviour, and the tooth fairy leaves money for lost teeth. Over time, the story changes. Children learn that presents and money don’t arrive by magic. They’re part of a transaction shaped by choices, expectations and behaviour: if you want the good stuff, you have to play your part.
Business users need the same honesty about AI, and the information it depends on.
For decades organisations have acted as though their information will somehow sort itself out. People save files wherever convenient, invent their own labels, duplicate documents, share too broadly, retain everything indefinitely and rely on memory to find what matters... and when AI arrives the expectation is that it will make sense of everything... automagically.
That’s the Father Christmas version of AI.
This fiction assumes technology will appear overnight, understand every file, resolve every contradiction, identify the authoritative version, respect every business rule and produce reliable answers from inconsistently managed information.
It won’t. AI doesn't have any comprehension of putting stuff in a file called "superseded", or that version of file you've copied to your desktop isn't as authoritative or up-to-date like the approved one in the Resources library.
AI is extraordinarily capable, but it isn’t a fairy godmother for unmanaged content. It works with the environment it’s given. Feed it outdated, duplicated, poorly secured or badly described information, and it will work harder, perform less reliably and potentially expose material that should never have been so accessible.
That isn’t an AI failure. It’s your information environment doing exactly what it has been allowed to do for years—only faster.
The real issue is not versioning
Without shared language, meaningful metadata, sensible structure and clear ownership, organisations have made finding and reusing knowledge far harder than it needs to be:
A poorly managed version history is not the business problem. It is evidence of a business problem.
An overshared site is not the business problem. It is evidence that the organisation has not made clear decisions about ownership, accountability and appropriate access.
A document that cannot be found is not merely a search problem - your staff recreate work, make decisions using incomplete evidence, delay a customer response or fail to reuse knowledge that the organisation has already paid to create.
Similarly, different people may describe and store the same business document in entirely different ways: Janet may look for a “customer onboarding pack”; Bill may call it a “new-client implementation guide”. Both are looking for the same thing. Neither is necessarily wrong.
The consequences of only talking about "practices" are that we miss the tangible impacts IM has, and issues it will help solve:
Staff spend time recreating material that already exists because they cannot find, trust or access the original.
Teams make inconsistent decisions because they work from different versions of policies, pricing, procedures or customer information.
Subject-matter experts become bottlenecks because critical knowledge sits in personal folders, inboxes or individual memory.
AI tools generate weaker answers because they are asked to navigate a cluttered landscape of conflicting and unreliable sources.
Net result: the organisation incurs avoidable cost, risk and delay while believing it has a "technology problem". The versioning, retention and permission issues matter, but they matter because of what they prevent the business from doing.
A better business case: build the road before buying the car
As a child hoping for a remote-control car at Christmas, and images all the fun of powering-around up and down the drive-way - telling them:
“You can't use your new car because the driveway has potholes, the gravel is uneven and the garden is full of rubbish,”
is technically accurate but its hardly inspiring.
A better conversation is:
“If you want the remote-control car to go fast and be fun, we need a clear place to run it. Lets fill the potholes, rake the gravel and remove the rubbish. Otherwise, the car will get stuck, break or spend its time going nowhere!”
The work has not changed. The reason for doing it has. Information management needs to be presented in the same way.
The business does not invest in information governance because version control is aesthetically pleasing. It invests because reliable, accessible and appropriately governed information allows people and AI to work faster, make better decisions and deliver more consistent outcomes ...but governance is a drag. Its boring.
If the road is uneven, full of rubbish and broken-up by holes, a faster vehicle merely reaches the problems faster. Knowing where the pot-holes are, and what needs clearing away is just the price for wheel-spinning happiness.
Fill in the IM landscape pot-holes and level the AI road ahead
To get the attention, and support you need for AI we need to frame the action in the real-context for the business:
Information management activity | Technical description | Business outcome |
Establishing clear ownership | Defining accountable owners for sites, content and data | Someone can make decisions about what is authoritative, current and worth keeping |
Improving document structure and descriptions | Applying consistent naming, metadata and classification | People and AI can find the right information without relying on individual memory |
Removing duplicate and obsolete content | Reducing redundant, trivial and outdated information | Staff make decisions from more reliable sources, and AI has less noise to interpret |
Fix permissions and sharing practices | Applying appropriate access controls and regular review | People gain access to what they need while sensitive material is less likely to be exposed |
Manage records and retention | Applying lifecycle rules to business content | The organisation retains evidence when needed and reduces unnecessary information burden |
Create authoritative sources | Identifying trusted locations for key policies, procedures and knowledge | Teams stop debating which version is correct and can act with greater confidence |
AI makes the need clearer, not less important
AI has made the consequences of unmanaged information more visible.
Before AI, poor information management often produced slow, localised pain. A staff member spent twenty minutes looking for a document. A project team rebuilt a template. A manager asked three people which policy was current. These frustrations were costly, but they were spread across the organisation and therefore easy to tolerate. AI changes the scale, and the visibility of the issues.
An AI assistant can help people find knowledge, summarise material, draft responses, identify patterns, support decisions and automate routine work. That promise is real. The quality of those outcomes though depends on the quality, accessibility and authority of the information available to it.
Put plainly: AI can accelerate a good information environment. It can also accelerate confusion.
An organisation cannot claim the benefits of AI while treating the information underpinning it as an ungoverned by-product of day-to-day work. The two are inseparable.
This does not mean every document needs perfect metadata, every site needs a committee or every user needs to become an information manager. That would be governance theatre, and businesses have enough theatre already.
It means designing practical habits and controls around the information that matters most:
Identify the information that drives key decisions, services, customer interactions and regulated obligations.
Make it clear where the authoritative version lives and who is responsible for maintaining it.
Make information easy to find using language that business users recognise, not only language that system administrators recognise.
Apply access rules that support productive work while protecting sensitive content.
Remove or archive information that is no longer useful, reliable or required.
Build these expectations into everyday processes rather than relying on periodic clean-up campaigns.
Stop selling fear
A business case based solely on negative consequences has a limited shelf life:
“Do this or storage costs will increase” may obtain a short-term response,
“Do this or audit will find a problem” may create urgency,
“Do this or AI will expose something embarrassing” may gain attention
...but fear rarely produces sign-off and money for project because you can see the benefit from sustained behavioural change. This just frames information management as a tax - a burden or a punishment for people who have been getting on with their jobs.
A better way connects the work to an ambition the business already holds. That ambition might be faster customer service, improved delivery quality, reduced rework, better operational resilience, lower risk, more reliable reporting, successful AI adoption or the ability to scale without adding equivalent administrative effort.
Dont' clean up information because IM says so. The message is:
Improve the conditions that will let the business get the outcomes it wants.
That is an adult conversation. It respects leaders’ ability to weigh trade-offs. It gives users a practical reason to change habits. It makes information professionals accountable for explaining value, not merely listing defects.
The responsibility sits with IM professionals too...
Business users should not be expected to understand the machinery of information management before they are allowed to benefit from it.
They do not need a lecture about retention labels, sensitivity labels, metadata schemas, inheritance breaks or storage consumption. Those are tools, not outcomes. They matter, but they should sit behind a clear explanation of what the organisation is trying to achieve.
Information professionals have to translate doom, gloom and boredom into the real-reasons for doing it...
Instead of “We need to rationalise SharePoint permissions,” say, “The users need access to the information to serve customers - without creating unnecessary exposure of confidential material.”
or saying, “We need better metadata,” say, “People need to find an approved answer in minutes rather than recreating work or relying on whoever happens to remember where it was saved.”
and rather than “We need to reduce content duplication,” say, “We need one trusted version of the information that drives decisions, customer commitments and operational activity.”
The technical work remains important. In fact, it becomes more important when linked directly to business value, but it should always be an enabler, not the destination.
Clear the path ahead, then accelerate
Every organisation wants the remote-control car. It wants the speed, the capability and the enjoyment of seeing work happen more easily.
AI can help deliver that. It can reduce routine effort, make knowledge more accessible, support better decisions and allow people to focus on work that needs human judgement - but it needs a path clear enough to travel.
Don't frighten executives and chivvy them along like a parent, or present governance as an administrative punishment to a child. The answer is to explain the bargain clearly - to an adult. The business can either;
improve the information environment, and the business can move faster with greater confidence using AI, or
continue with unmanaged behaviour, and the business will keep paying for the consequences in rework, delays, unreliable outcomes and decisions and increasing risks of security and privacy issues.
If your organisation leaves the information roads potholes unfilled, the permissions unresolved, the rubbish uncleared and the content scattered, the business shouldn't be surprised when AI struggles to deliver a smooth ride.
That is not an information management message. Its a business message.
Acknowledgement
Generative AI has been used in this article in: clean-up and review of article copy (spelling, grammar, tone), the creation of the article images, and QA review prior to publishing. Tools used: Wix Photo Studio AI Image Creator tool, and Perplexity.ai Sonar 2 model. The primary article topic and initial copy, ideas and visual imagery, key discussion points, frustrations and similes were created by the author based on personal experience and interactions with customers and business professionals. Any errors or issues with the content in this article are entirely the author's responsibility.
About the author: Jonathan Stuckey
Comments