Your browser does not support JavaScript!
  • LinkedIn
  • YouTube
  • RSS

Why do 47% of Agile Transformations Fail? 

Chart describing the success and failure rate of Agile Projects


A good time was had by all at Scrum Day India 2020. My talk on “Why 47% of Agile Transformations Fail!” was well received. After many requests for the slides I am posting them here.

I recently read that 47% of Agile Transformations fail and since 77% of Agile Transformations are Scrum Transformations, most of these failures are down to bad Scrum implementation.

Questioning these numbers motivated me to dig deeper. Prof. Kotter in his book Accelerate says he has never seen a long-term successful Agile transformation using Waterfall leadership. At a recent Agile conference with over 200 people attending my session, I did a brief survey. I asked how many agile transformations had Waterfall leadership. The response was over 95% of those attending. So it seems the failure rate is somewhere between 34% and 95%.

I decided to dig even deeper and looked into the Chaos Report data from Jim Johnson, CEO of the Standish Group. Jim has been collecting data for decades on over 500,000 IT projects and we have tested his data by querying over 1,000 Certified Scrum@Scale Practitioners on their success rates which average lower than the Standish Group Data. So Standish Group data is known to be conservative as to failure rates.

 Why Scrum Works - Decision Latency

Jim Johnson and I had coffee last summer to review why he thinks Scrum works. It is down to how long it takes to make a decision when a decision needs to be made. He calls it Decision Latency. This is the average time it takes to make a decision, which he is tracking in thousands of companies. When Decision Latency is less than an hour, the success rate is 68% independent of process. If the average decision time is over 5 hours, the success rate falls to 18%. Scrum helps reduce decision time by pushing development decisions to the team and all priority decisions to the Product Owner radically shortening decision time to an hour or less. In Scrum@Scale implementations, an Executive Action Team is meeting daily to make higher-level decisions. An oil and gas company cut well drilling average time from about a month to six days or less by fast executive decisions.

 So I asked Jim to give me data on Decision Latency for his agile teams broken down into finer detail. Here it is for 10,000 Agile Projects - Email titled: 10,000 Agile Projects, from Jim Johnson, Standish Group, 28 Jul 2020.

The Standish Group has a systematic way of collecting data on how long it takes to make a decision on average for companies in their database. Jim says he has seen six teams wait for six week for executives to make a decision. My venture firm would calculate this loss as $150K x 6 x 6 = $5.4M and would discuss what management changes to make at the next board meeting.


So the best data we have shows 41.62% success and 46.92% late, over budget, with unhappy customers, and 11.46% failing completely and delivering nothing. So I see 47% failed Agile Transformations as actually, a conservative number.

A follow up on failed transformations by the MIT Sloan Business Review (MIT Sloan Management Review 20 Jan 2020).showed that 67% of failures are terminal, i.e. the companies go bankrupt or are acquired (This was a note in their Research Insights on 20 January 2020 that is no longer accessible on the web). At the same time, the most successful and valuable trillion-dollar companies, Microsoft, Amazon, and Apple all use Scrum and they have shown different paths to Scrum@Scale to achieve phenomenal success. Why do so many companies go out of business because they cannot “get” Agile when there are such obvious examples of radical success?

Implementing Scrum properly is certainly one of the primary reasons. Well trained high-performing Scrum teams make decisions quickly and with confidence. Often, training and coaching, particularly for leadership, is just not effective. Enterprise Agility for Dummies points out 10 reasons for failure all of them directly related to poor leadership.

Effective Agile Leadership Means Executive Participation

Agile Enterprises, or those that aspire to be Agile need a scaling framework that focuses on reducing decision time at all levels of the organization. This requires more than executive sponsorship, it requires executive participation as Agile leaders. They need an Executive Action Team and an Executive Metascrum as provided in Scrum@Scale. Without an Executive Metascrum prioritizing all initiatives, 30% of the projects have no significant business value and should be terminated. So many, if not most companies are wasting 30% of their budget before they do anything useful. That is a good indication of failure to achieve business agility.

The Standish Group data shows consistently over the last 20 years for over 500,000 projects that 64% of features delivered to customers are never or rarely used. So without a good Product Owner organization, of the 70% of projects that might be useful, we waste 64% of our staff on things the customer won’t use. At OpenView Venture Partners we call these junk stories. If we eliminate the junk we get investments like Data Dog which has exploded during COVID-19 and gave us a 53300% return. 

For the few features being built, effective Scrum teams are critical. The best measure of team performance is Process Efficiency and a typical Scrum team has an average process efficiency of less than 15%. The definition of Lean is over 25% and this can be achieved in one sprint by a team focused on eliminating waste.

 So, with roughly 1/3 of companies trying to practice Agile actually failing and going out of business it is clear that there needs to be a transformation of Scrum training and coaching. With the COVID disruption over 3600 companies filed for bankruptcy in the first half of 2020, many of them household names, some of them trying to get agile.

We need to go back to the grandfathers of Scrum, Takeuchi and Nonaka, and examine how they defined Scrum Project Management in their 1986 Harvard Business Review Paper, “The New New Product Development Game.” They were looking at autonomous, cross-functional teams with transcendent goals in Lean hardware manufacturing companies. They described the intense collaboration and continuous improvement practice of these teams and called them Scrum after the scrummage formation in Rugby.

Scrum at its root is Lean and about eliminating waste. The latest book on “The Secret Behind the Success of Toyota” shows that lean manufacturing generates only 5% of Toyota profits. The other 95% is generated by the Shusas, the Chief Engineers, who deliver twice the value at half the cost. In today's world Tesla has shown that Lean plus Agile is worth more than twice as much as Lean by itself. See Brandon's EV Financial Stats.