Mediocrity is becoming a leadership choice
I think we have reached the point where accepting mediocre work in sales, marketing, business development and organizational learning should make leaders uncomfortable.
Not because artificial intelligence makes everybody brilliant.
It does not.
But because so much of what used to prevent us from getting better has become easier, faster and less expensive to overcome.
We can investigate a market, challenge an assumption, prepare a customer conversation, compare approaches, build a prototype, improve a text, create a learning program or examine a process in a fraction of the time many of those activities once required.
The technology will not make the right decisions for us. But it has removed a remarkable number of excuses for not doing the work needed to make better ones.
Being less capable than a competitor is not embarrassing. Being satisfied with an avoidably weaker version of your own organization is becoming difficult to defend.
The bar has moved, whether management has noticed or not
For years, a great deal of ordinary commercial work was constrained by available hours.
Proper market research took time. Testing several positioning options took time. Preparing an account plan took time. Training every salesperson individually took time. Turning a promising idea into something a customer could actually see took time and money.
Those constraints have not disappeared. But for many knowledge-intensive tasks, they have been reduced dramatically.
A sales manager can now ask for a structured examination of a target market, test the reasoning against sources, identify gaps in the evidence and turn the verified findings into a more useful plan for the team.
A marketing leader can develop and compare several messages, examine them from a buyer's perspective, prepare variations for different audiences and establish a process for learning from responses.
A business developer can examine adjacent markets, map competing alternatives, formulate hypotheses and prepare experiments before committing substantial resources.
A leader responsible for competence development can turn scattered knowledge into practical exercises, feedback routines and accessible support for people doing the work.
None of this guarantees an excellent result.
But it changes what we should regard as an acceptable starting point.
When the cost of preparation falls, the standard of preparation should rise.
Look at what we still accept
I find it increasingly strange how often organizations continue to accept work that could be made substantially better with a modest amount of disciplined effort.
Salespeople approach companies without a convincing reason why those companies should speak to them.
Marketing teams publish material that looks professional but does not answer a real buyer question.
Business-development meetings produce interesting ideas but no testable hypotheses, owners, decisions or next steps.
Managers repeat familiar training programs without establishing whether behavior changed afterward.
Product teams discuss what customers might want for months while inexpensive prototypes and direct feedback could settle important questions much earlier.
AI cannot solve all these problems. In some cases, it can make them worse by producing polished material that disguises poor thinking.
But that is not an argument for continuing as before.
It is an argument for raising our standards of evidence, judgment and execution.
A plausible customer list is not a qualified pipeline. A beautiful campaign is not a changed buying decision. A polished strategy is not an implemented strategy. A training session is not improved performance.
We should no longer confuse the ease of producing an output with the quality of the outcome.
The best version of ourselves is a practical management requirement
I am not arguing that every company can become the best company in its industry.
That would be an empty promise.
Different organizations have different resources, markets, constraints and ambitions. Some work also depends on judgment, relationships, practical experience, trust and accountability in ways that AI cannot replace.
I am arguing something both more modest and more demanding.
Every organization should be expected to work systematically toward the best version of what it can realistically become.
That means asking, repeatedly:
- Are we making decisions on the best evidence we can reasonably obtain?
- Do our salespeople know whom to approach, why the timing matters and why we have a credible Right to Win?
- Does our marketing help buyers understand a problem and make a decision, or does it merely fill a publishing calendar?
- Are we improving our services based on observed needs and customer feedback?
- Do we use technology to remove unnecessary work and strengthen human judgment?
- Can we show that the changes we make improve a meaningful result?
These questions were important before AI.
The difference is that many of the activities needed to answer them have become low-hanging fruit.
It is no longer reasonable to treat every basic improvement as a major transformation project.
This is a leadership issue, not an AI-adoption contest
The wrong management question is, "How many people are using AI?"
The better question is, "Why are we still accepting results we now have a realistic way to improve?"
A company can have an AI subscription for every employee and remain mediocre. It can generate more presentations, more posts, more leads and more reports without creating more value.
It can even become less effective if the volume of plausible but unverified work increases.
Leadership means setting the standard, choosing the problems worth solving, providing safe and lawful ways to use the tools, ensuring that people develop their competence and checking whether the work actually improves.
It also means protecting time for learning and making it acceptable to expose weak assumptions.
You cannot demand better results while rewarding only the familiar routines that produced the old ones.
And you cannot call it transformation when the only measurable result is an increase in content production.
The failure is not that people have not mastered every new tool. The failure is when management knows improvement is possible and does not organize the work required to pursue it.
Excellence is not a prompt. It is a system.
This is where I think a lot of the excitement about AI misses the point.
The advantage does not come from asking a model to "make this world-class."
It comes from combining better tools with a better operating method:
1. Define the outcome worth improving. 2. Establish what is actually happening today. 3. Separate verified facts from plausible assumptions. 4. Use AI to explore alternatives, prepare work and reduce friction. 5. Let people make accountable decisions and test them in reality. 6. Measure what changed, keep what worked and repeat.
In our commercialization work, this distinction matters constantly.
We can produce more leads than ever. The difficult part remains establishing which companies may genuinely need something, whether we can help them and why they might choose us.
We can create excellent-looking analyses. The difficult part remains making sure the evidence is sound and that decisions improve because of it.
We can build systems quickly. The difficult part remains proving that someone benefits enough to use or pay for them.
The better the tools become, the less impressed I am by output alone.
I want to see what changed in the market, the customer relationship, the employee's capability or the company's results.
What changed
My own threshold for what counts as good preparation has moved.
Work that once required a specialist team or a lengthy external assignment can sometimes be investigated, structured and tested by a small team with capable AI assistance. That does not eliminate the need for expertise. It changes how expertise can be applied.
I am becoming less patient with the argument that there is simply no time to understand the customer, improve the message, test the process or learn from what happened.
Sometimes the constraint is real. Sometimes information is unavailable, the risk is high or the organization lacks essential capacity.
But increasingly, the missing ingredient is not access to technology.
It is management's willingness to change what it expects and how work gets done.
What I think matters
AI has made continuous improvement more accessible. That makes complacency more expensive — and harder to excuse.
The ambition is not to be flawless, to automate everything or to claim superiority over competitors.
It is to refuse avoidable mediocrity.
A leader should be able to point to the important work of the organization and explain how the team is becoming better at it, what evidence supports that assessment and what it will improve next.
If the answer is only that people are busy, that the material looks good or that "this is how we have always done it," I think we have a leadership problem.
The embarrassing part is not being behind today.
It is choosing to stay there when a credible path to improvement is within reach.
What remains uncertain
We do not have a universal measure of how much AI can improve every occupation or organization. Gains vary with the task, the quality of information, available systems, employee competence and the consequences of error.
Faster production can also increase mistakes, weak decisions, security risks and unnecessary activity. Better tools can reduce the cost of doing the wrong thing.
Some improvements require investment, organizational trust, regulation-aware implementation or changes that are neither quick nor cheap. And being the best possible version of a company cannot be reduced to a single benchmark score.
That is why my argument is about a management obligation to investigate, test and improve — not a claim that AI automatically creates excellence.
We still have to establish whether a change works.
But uncertainty about the size of the gain is not a good reason to avoid looking for it.