Cheap Agile: Don’t Hire a Team of Coaches

Coaching the coachesMultiple times a week it happens: No matter how many ways I try to make clear I am only a solo or lead transformation coach, people contact me about joining the team of Agile coaches in some large company.

The worst part is not the waste of my time. The worst part is the waste of those companies’ money. There is no rational reason for them to be hiring teams of Agile coaches. Based on a pretty impressive mound of evidence, they should be training existing personnel to fill the roles.

The concept is called “train the trainer” (TT) and has been around for at least 50 years. Underlying TT is the fact that true experts on a given topic are relatively few and usually expensive.[1] To spread their knowledge further and faster, you hire one expert trainer, who in turn trains internal trainers from among your existing people. In the case of Agile coaching, this instantly saves you the costs of recruiting, onboarding, acculturating for months, and paying all those new coaches (not to mention staffing agency overhead if they’re contractors).

There is another critical benefit to the TT route: It reduces resistance to change, by eliminating the sense that change is being driven by outsiders who “don’t understand our business.” When a large group of insiders have been tasked by upper managers to implement Agile, that argument disappears.

Some of the biggest resistors to Agile are project managers, for the perfectly valid reason that Agile is, in part, project management done by self-organizing teams—not by project managers! By the same token, line-level and middle managers often resist because self-organization means fewer managers are needed. Out of sheer self-preservation, many managers and PMs (and PMOs) understandably try to block adoption, sometimes while claiming to support it.

I explained in an earlier post why they should not become Scrum Masters. What if, however, you trained many of them to be Agile coaches? In addition to providing some formal training on how to train, the expert Agile Transformation Coach can roll out Agile in phases through managers and/or PMs. That is, the coach trains the manager on the next Agile technique to implement, and the manager trains his/her teams.

If you are worried about the “gossip game” aspects—information getting distorted as it goes down the chain—know that a properly constructed Agile transformation project is going to have objective measures both for the transformation and for Agile team operations. This makes it easy for the “head coach” to spot and address problems. A good change leader will also maintain strong two-way communications with all levels in the company, and drop in on teams to do spot-checks and provide direct coaching.

A subset of the newly trained coaches can shift into permanent cross-team facilitator roles, like Full Stack Scrum’s “Agile Release Manager” or SAFe’s “Release Train Engineer.” Just as the Scrum Master is supposed to coach the team, these people, serving as the equivalent to the SM at the cross-team level, are the coaches at that level. (Given these roles, and the fact SMs coach the teams, the job title “Team-Level Agile Coach” should not exist.) Any former manager/PM who doesn’t find another position inside the firm will at least have a new, market-hot skill, and proof they’ve actually done the work, which is necessary for a valid certification like the PMI-Agile Certified Practitioner. There are other roles former managers can take in lean organizations, though.

But, you may rightly ask, is there any evidence the TT method works? I’m so glad you asked! As one set of researchers said after reviewing the science (citations removed):

“The train-the-trainer model has been found to be an effective method of information dissemination that allows for more people to be trained at a lower cost, as well as creating a better fit for program implementation because those trained are familiar with the settings in which they are implementing the program. Using a train-the-trainer model also has the added benefit of harnessing relationships that already exist… Furthermore, a previous study showed that program content that was delivered by ‘non-experts’ was actually more effective because non-experts tend to use concrete language, whereas experts are more likely to present material using abstract language and examples that are harder to understand”.[2]

Nonetheless, I went through the literature on TT myself for concrete evidence. The earliest use of the term I found was in 1968, and software vendors were training internal client trainers as early as 1993.[3] Training magazine carried a listing of TT workshops on a variety of business topics in 1994.[4] A 2007 article in the British Human Resources magazine claimed, “a massive 80% of UK training is actually delivered by subject-matter experts rather than professional trainers…”[5]

I looked for studies showing that TT students-turned-trainers actually changed people’s behaviors. I found virtually no research in businesses, but a lot of it in other fields. Militaries have used the model to train noncommissioned officers to teach their soldiers ethics[6] and detection of improvised explosive devices.[7] A sergeant who went through the latter “said the training has made the overall mission safer. It has given the soldiers a new view on current IED activity in their sector.”

By far the most TT studies came out of medicine and public health. After faculty members at an alternative medicine college were taught evidence-based practice, they set goals for teaching it to their students, and six months later 80% reported those were achieved.[8] Nursing teams trained on teamwork by other nurses who had gone through a TT program reduced their nursing mistakes.[9] A systematic review of TT programs to improve resuscitation skills had strong positive results.[10] When Chinese nurses were trained to teach patients to manage coronary artery disease, their patients improved on a range of health behaviors like diet, exercise, smoking and drinking far better than did a control group.[11]

In a direct comparison to training by experts, three groups of mental health clinicians were taught to use “motivational interviewing” (MI). Their skills were judged 12 weeks later, using role play or tapes of client sessions. The group trained by experts was rated the most adequate at MI (59%), but those trained by expert-trained peers were almost as good (53%), versus only 18% for people who had learned via self-study.[12] Similar successes were reported for different TT programs for reducing family violence, eating disorders (training students to train other students), and sex abuse.[13]

So don’t waste your money and cause resistance to Agile by hiring a team of outsiders as coaches. One will do, and your current people will do the rest.


Please share this post at the bottom of the page.

[1] Sarah L Morgan et al., “The ELIXIR-EXCELERATE Train-the-Trainer Pilot Programme: Empower Researchers to Deliver High-Quality Training,” F1000Research 6 (August 24, 2017): 1557,

[2] Christine Weingarten et al., “Examining the Utility of a Train-the-Trainer Model for Dissemination of Sexual Violence Prevention in Schools,” Child Abuse & Neglect 80 (June 1, 2018): 70–79,

[3] “Roundtable: Train the Trainers,” The Internal Auditor; Altamonte Springs 50, no. 2 (April 1993): 64.

[4] “Train the Trainer Workshops,” Training; Minneapolis 31, no. 2 (February 1994): 81.

[5] Jacky Hyams, “Train the Trainer,” Human Resources; London, May 2007, 56.

[6] Eva Wortel and Jolanda Bosch, “Strengthening Moral Competence: A ‘train the Trainer’ Course on Military Ethics,” Journal of Military Ethics 10, no. 1 (March 2011): 17–35,

[7] “Train the Trainer,” US Fed News Service, Including US State News; Washington, D.C., August 27, 2009,

[8] Elizabeth S. Allen et al., “A Train the Trainer Model for Integrating Evidence-Based Medicine into a Complementary and Alternative Medicine Training Program,” EXPLORE 7, no. 2 (March 1, 2011): 88–93,

[9] Peggy Ward-Smith, “Abstracts,” Urologic Nursing 34, no. 2 (April 3, 2014): 100–102,

[10] Chance R. Anderson and Breena R. Taira, “The Train the Trainer Model for the Propagation of Resuscitation Knowledge in Limited Resource Settings: A Systematic Review,” Resuscitation 127 (June 1, 2018): 1–7,

[11] Zhiyun Shen, Changying Jiang, and Liqun Chen, “Evaluation of a Train-the-Trainer Program for Stable Coronary Artery Disease Management in Community Settings: A Pilot Study,” Patient Education and Counseling 101, no. 2 (February 1, 2018): 256–65,

[12] Steve Martino et al., “Teaching Community Program Clinicians Motivational Interviewing Using Expert and Train-the-Trainer Strategies,” Addiction 106, no. 2 (February 2011): 428–41,

[13] Agnes Y. Lai et al., “An Evaluation of a Train-the-Trainer Workshop for Social Service Workers to Develop Community-Based Family Interventions,” Frontiers in Public Health 5 (2017),; Weingarten et al., “Examining the Utility of a Train-the-Trainer Model for Dissemination of Sexual Violence Prevention in Schools”; Lisa Smith Kilpela et al., “Reducing Eating Disorder Risk Factors: A Controlled Investigation of a Blended Task-Shifting/Train-the-Trainer Approach to Dissemination and Implementation,” Behaviour Research and Therapy 63 (December 1, 2014): 70–82,

Share this post: