Strategy & competitive advantage
Video · 5 min read
The short answer
Watch: If every company has access to the same AI models, where does competitive advantage come from?
Watch the explanation, then explore the practical implications below.
Start with the reason customers choose you
The question is not simply whether competitors can use the same technology. It is whether your business can use it to deliver a better result. A faster proposal has limited value if it misunderstands the requirement, makes an unrealistic promise or creates more work for the customer.
Identify the qualities customers actually value in your service: understanding an unusual requirement, giving dependable advice, responding clearly or delivering what was agreed. Then examine how the workflow produces those qualities and where it currently falls short.
Make company knowledge usable in the work
Useful context includes approved specifications, past decisions, customer requirements and the constraints your team works within. Having those records somewhere in the business is different from making the right records available at the point of use.
Give shared information an owner and a clear scope. Distinguish current guidance from a one-off exception, preserve access permissions and show where important facts came from. More material is not necessarily better; the system needs relevant evidence that people can inspect and challenge.
Preserve the judgement that changes the result
An experienced colleague may recognise that a customer’s requested solution will not address the underlying problem. They may know when to ask a further question, when a commitment is unrealistic or when a standard response would damage a relationship.
Map those moments explicitly. AI can prepare evidence or a draft, while the appropriate person resolves the issue. This does not require someone to inspect every action. It requires the workflow to recognise the situations where human involvement adds value and route them with enough context.
An example: responding to an unusual enquiry
Consider a business receiving a request that resembles a standard job but includes an ambiguous delivery requirement. A fast automated response could quote the normal service and miss the exception. A more useful workflow would assemble the requirement, identify the ambiguity and ask an experienced person to resolve it before a commitment is made.
After review, capture the correction and its reason. If it applies more widely, have the responsible person approve an update to the guidance. Keep customer-specific information within its proper boundaries. This is an illustrative workflow, not a claim that every enquiry needs manual approval.
Turn repeated corrections into organisational learning
If the same expert fixes the same problem repeatedly, examine why the process keeps recreating it. The brief may be incomplete, the source information may be hard to find or an approved rule may never have reached the next person doing the work.
Create a deliberate route from correction to reviewed guidance. Test subsequent outputs to see whether the change helps. Do not silently treat every chat message as a new rule, and do not assume that storing a correction guarantees that future work will use it correctly.
Measure advantage through customer and business outcomes
Choose measures linked to the reason for the change. These might include response time, first-time completeness, review effort, delivery reliability or the frequency of avoidable corrections. Track the cost of review and maintenance as well as time saved in preparation.
At CSM, we would start with one workflow and the people who understand what makes it valuable. We would identify the knowledge, decisions and handovers that support the customer outcome, then test where AI improves that process. The objective is a company capability that can be maintained and improved, rather than a collection of isolated shortcuts.
Questions for your leadership team
Why do customers choose us?
Identify the specific qualities of the service that matter to customers and the evidence supporting that view.
Where does expertise change the outcome?
Use recent examples to locate the decisions or conversations that should remain visible in the workflow.
Which knowledge should the team be able to reuse?
Separate approved company guidance from confidential customer details and individual exceptions.
How do corrections improve the next piece of work?
Name the person who reviews changes and check whether future outputs use them appropriately.
What result would demonstrate an advantage?
Measure customer outcomes and completed work alongside preparation time, review and rework.
Full video transcript
The full video transcript is included below for reference.
I think this is where the way you organise the work starts to matter a lot. If everyone can access similar models, having access on its own doesn’t tell us very much about why a customer should choose you.
You might use AI to prepare a proposal, answer an enquiry or analyse a set of requirements. Your competitor might use it for the same things. The difference starts to appear in what the system knows about your business, what people contribute and what happens after it produces an answer.
Take an experienced person who knows that a customer always asks for one thing but usually needs something slightly different. That knowledge can change the whole conversation. If you automate the process without giving that person a place in it, you may make the response faster while making it less useful.
That’s why I think the idea of bringing people in at the right points is important. You don’t need a person touching every step to make the process valuable. You need to understand where their judgement changes the outcome.
Sometimes that’s because the work is unusual. Sometimes it’s because there’s a relationship involved. Sometimes someone needs to say, “We could do this, but it isn’t how we want to treat a customer.”
And some of that judgement should become shared knowledge. If your best person keeps correcting the same thing, the process should learn from those corrections. Otherwise, you’re paying for the same expertise repeatedly without improving how the company works.
So the advantage is going to depend on the combination. The AI, the people, the information, the process and the culture around it.
I’d be careful about assuming that maximum automation means maximum advantage. If you automate away the parts where your people make a meaningful difference, you may also automate away some of the reasons customers value you.
Put these questions to work in your business
Bring a real decision or workflow to a conversation with CSM. We can explore the evidence, responsibilities and practical steps needed to make AI useful for your team.