Organizations are constantly making investments and improvements, large and small, to improve results.  

The bulk of these improvements take place within a function. Engineering used to introduce Agile to improve coding practices.  More recently, they introduce AI to get the next leg up on coding output.  Marketing lobbies to invest in Data Analytics to improve funnel management.  The Call Center augments their agents or replaces them all together with Agent-assisted AI bots.

It makes sense that the bulk of organization improvement efforts are generally "within function."  The functional leaders seek the investment and have a vested interest in making sure the changes are successfully introduced and adopted because they are often counting on those improvements to help them deliver the numbers they signed up for.

But periodically companies also engage in company-wide improvement efforts where they are trying to get the whole organization or at least large swaths of it to adopt a particular approach or framework or system that touches nearly everyone or perhaps to change behavior in a coordinated way, say, around customer focus or experimentation or problem-solving or servant-leadership.

Six Sigma, which I was deeply involved with, was certainly an example of this. 

AI is the latest, and though AI is being implemented differently across functions, many companies are asking all their leaders to figure out how AI can be leveraged to drive improvements.  It is touching everyone.

Efficacy of Corporate-Wide Improvement Efforts

I started wondering about the history and impact of these broad-based organization initiatives.

I decided to ask ChatGPT to list all the corporate-wide efforts it could find and to rank order them based on their impact on organizational outcomes.  

Major caveats about the question alone not to mention any answer an AI gives are are obviously and immediately needed. 

1. The question posed is fraught.  What does corporate-wide mean to AI?  My search queries pushed for efforts that touched almost every function in the company or at least multiple functions, such as the previously mentioned process improvement, leadership training, and culture initiatives.  While the AI might be able to find and published results, I doubt it could assess how "broad-based" an initiative was in each company.
2. Speaking of public results, an impact/ranking analysis like this is going to be limited to and biased by "published" results. There are plenty of companies that engage in broad-based improvement efforts and have no interest in reporting the results.  They may not want the publicity or they may think their successful efforts to improve outcomes for customers might be a source of competitive advantage.
3. And what results are being published?  Since this is not academic research and more likely business publications, it is likely going to be financial results that are being highlighted.  When companies engage in culture change to more align employee behavior with key corporate values or the external marketplace, how would we describe and rank order the value of that?
4. Then there is the whole question of teasing out the impact of any cross-company initiative from all the other efforts inside the company to improve performance.  Satya Nadella became CEO of Microsoft in 2014.  There is no question that he changed the culture of Microsoft. When he took over, the market cap was around $300 billion, and as of September 4, 2025, it stood at approximately $3.76 trillion.  How much of that cap change was due to the change in culture vs. product mix, acquisitions, business development, go-to-market changes, etc?  No one can say and thus any conclusions about ranking the impact of cross-company initiatives is suspect. 
5. Like the Microsoft case study linked to in #4, published results that are out there are case studies.  You won't find research with some kind of control group, even "yoked-controls" (as was done in the book Good to Great) where a company implementing a broad-based initiative is compared to a company of similar size from the same industry who did not take on the corporate wide initiative.  One has to be careful about drawing conclusions from one-offs.
6. I was sure my AI inquiry would be able to find connections between various types of organizational interventions and positive outcomes.  I am also sure it likely ignored implementation disasters at particular companies. Six Sigma ranks high in the AI summary that follows because it collectively saved companies that implemented it billions of dollars.  But the "B sides" are not hard to find.  Case in point: go find the Business Week article on the implementation of Six Sigma at 3M.
7. There could be real problems with the differences within and between the 20 initiatives that the AI settled on.  How do all the flavors of culture change or leadership training or DEI or Team Training get grouped?  Surely some are better than others.  Maybe there should be 40 flavors of corporate-wide interventions instead of the 20 listed here.
8. AIs do not tell you how they do their analyses.  Heck, the overlords who built them often don't know how they did their analyses.  With a refereed journal, procedures, analytical methods, and references are clearly laid out so that you can critically examine and even attempt to replicate the studies or so you at least know how to better hedge the conclusions.  AIs are a black box.
9. And, as any good psychometrician knows, you have to have reliability before you can have validity. Are we comfortable we are there with AI's?  I am not sure.  They still hallucinate.  Test-retests are often still recommended as are layered nuanced inquiry strategies. Trying different AIs is encouraged. This sounds more like YMMV...your mileage may vary...than a conclusion you should hang your hat on.

All these caveats make the conclusion so suspect, it's right to ask, "why bother?"

Because, as you will see in the wrap-up, with organizational change the questions are as important as the answers.

ChatGPT 4.0 Ranking of Broad-based Organization Improvement Efforts

Tier 1 – Proven, System-Wide High-Impact Efforts

Tier 2 – High Value, Context-Dependent Approaches

Tier 3 – Innovation & Role-Specific Interventions

Tier 4 – Niche, Experimental, or Complementary Approaches

 Other Ways of Grouping the Interventions

In addition to the impact tiers here, there are other ways to group the organizational improvement approaches.

Centralized, Tops-Down vs. Decentralized, Bottoms-up. Centralized, Tops-down approaches are directed, non-optional efforts. Examples would be Digital Core Systems initiatives like ERP implementations and the typical efforts to drive overall organizational outputs by pushing simultaneous process improvement across Product Development, Marketing, Sales in a "sum-of-the-parts" approach.

Other examples would include deciding to apply for the Baldrige Award or ensuring all leaders were trained in a certain methodology where their progress and results using the new methods were tracked.

Decentralized, bottoms-up improvement efforts are approaches where people throughout the organization are given the tools and it is up to them to decide how to leverage them. OKRs and Balanced Scorecards might be examples here depending on how closely scrutinized the implementation was.

But leadership training is the best known example of decentralized, bottoms-up improvement. Leaders are often given a range of tools from strategic thinking, values training, people development approaches, and "ways of being" ideas on how to "manage self" more effectively to improve results.

In most of the leadership training roll-outs, there is no real central project plan or central tracking of results, other than to try to "get everyone through it."  Leaders take what they learn and are empowered to figure out how to implement in their environment.

Well-defined vs. Less Well-defined.  Another dimension you could organize these 20 approaches by would be from well-defined to less well-defined, bordering on squishy.

You can disagree, but in my view applying for the Baldrige, implementing Agile and Lean methodologies, and ERP roll-outs are well-defined broad-based initiatives.

On the other hand, DEI, AI-augmented Knowledge and Learning Systems (I believe these will get more structured, but they are too nascent at this point), and Holacracy are less well-defined. 

Therefore, if you buy ChatGPTs ranking of impact presented here, then, as would be expected, more well-defined, tops-down initiatives clearly have more impact. 

And that, ladies and gentlemen, is the most obvious thing you will read for months. 

Broad-based Initiatives or Not?

Like a boxer's left jab sets up the straight right, you can just consider this initial inquiry as a set-up the more important follow-on questions.

In your own company or in organizations you work closely with:  

• Is the organization engaged in any corporate-wide improvement that is impacting or trying to impact how almost everyone in the company works or behaves?
• If yes, are any metrics being used to assess progress and impact? If not, should there be?
• If there is not an effort that would be described as organization-wide, do you think there should be such an effort?  What improvements do you think the organization would achieve from such an implementation?
• If you think they should be pursuing a broader-based initiative, why do you believe they currently are not?

In Part 2, I will examine some of the broad improvement efforts that surveyed HR leaders said they are considering in 2025.

Dennis Adsit, Ph.D. is an executive coach, organization consultant, and designer of The First 100 Days and Beyond, a consulting service that has helped hundreds of newly hired and promoted executives get great starts in challenging new jobs.