{"id":1258,"date":"2026-11-11T14:30:00","date_gmt":"2026-11-11T14:30:00","guid":{"rendered":"https:\/\/cotocus.org\/blog\/?p=1258"},"modified":"2026-11-11T14:30:00","modified_gmt":"2026-11-11T14:30:00","slug":"stalled-skills-diagnosis","status":"publish","type":"post","link":"https:\/\/cotocus.org\/blog\/stalled-skills-diagnosis\/","title":{"rendered":"Why Team Skills Stall, and a Drill-Based Fix"},"content":{"rendered":"<p>A symptom that many engineering managers recognise: the team has been running production systems for years, yet certain tasks still take as long as they did in the first month. Incident triage drags. Pipeline fixes get passed to the same two people. New joiners take ages to become useful in the same few areas. The instinctive remedy is more training time, but the diagnosis is usually different. The hours exist; they simply are not aimed. This article walks from symptom to cause to remedy, and argues that a routine of <strong>deliberate practice<\/strong>, meaning short drills on one weakness with fast feedback, addresses the cause better than extra volume does.<\/p>\n<h2>The symptom: experience that does not compound<\/h2>\n<p>Start by describing what you observe, in concrete terms. Useful descriptions sound like &#8220;diagnosing a failed deployment takes our median engineer about forty minutes&#8221; rather than &#8220;we need to be faster&#8221;. Collect three or four such statements from the last quarter. If the same weakness appears in several people, it is a team habit, not an individual shortfall.<\/p>\n<p>Next, check what has already been tried. Typical prior attempts include a vendor course, a lunch-and-learn, a reading list and a &#8220;just get more reps&#8221; instruction. If those produced no durable change, you can rule out lack of exposure and lack of information as the primary causes. That narrows the field.<\/p>\n<h2>Diagnosis: four possible causes<\/h2>\n<p>When competence stalls despite exposure, four causes account for most cases.<\/p>\n<ol>\n<li><strong>Repetition on autopilot.<\/strong> The task is performed often, but never with attention on the weak step, so errors become habitual.<\/li>\n<li><strong>Targets that are too broad.<\/strong> &#8220;Improve incident response&#8221; cannot be rehearsed; &#8220;read the first five minutes of alerts and decide a severity&#8221; can.<\/li>\n<li><strong>Missing or soft feedback.<\/strong> Nobody checks the approach, only the outcome, so a lucky success hides a bad method.<\/li>\n<li><strong>Changes made in bulk.<\/strong> After a failure, the team revises process, tooling and roles at once and cannot tell what helped.<\/li>\n<\/ol>\n<p>You can test which one applies with a quick audit. Ask three engineers to narrate how they handle the weak task. If their narrations differ widely, feedback is missing. If they are identical but slow, autopilot is likely. If they cannot describe a specific step they find hard, the target is too broad.<\/p>\n<h2>The remedy: a drill loop<\/h2>\n<p>The treatment mirrors the causes. Replace autopilot with attention, broad goals with narrow ones, soft feedback with explicit checks and bulk change with single adjustments. Concretely, a drill has five parts.<\/p>\n<ul>\n<li>A target written as one sentence, naming a weak step.<\/li>\n<li>A sandbox or replay that lets you repeat the step without production risk.<\/li>\n<li>A time box of ten to twenty minutes.<\/li>\n<li>A check that can fail: a pass mark, an automated test or a reviewer with one question.<\/li>\n<li>A single change before the next attempt, followed by a two-line note.<\/li>\n<\/ul>\n<p>When picking targets, it helps to look at a reasoned list of capabilities rather than invent your own. A piece on <a href=\"https:\/\/cotocus.org\/blog\/essential-aiops-skills-every-modern-it-professional-needs\/\">the abilities modern IT operations staff are expected to build<\/a> can function as a menu from which you choose one capability and carve out a drillable fragment.<\/p>\n<h2>Evidence to collect, and how to read it<\/h2>\n<p>A diagnosis is only useful if the remedy can be evaluated. Before the first drill, record a baseline: the time to complete the target task, the number of review comments it attracts and the confidence of the person rating themselves from one to five. After four weeks, collect the same three figures.<\/p>\n<p>Interpret them with caution. A fall in completion time with unchanged review comments suggests speed gained without quality loss, which is good. A fall in time with more review comments suggests rushing, and the drill&#8217;s pass mark needs tightening. No change in any measure after four weeks usually means the target was still too broad, or the feedback source was too lenient. Fix those before concluding that drills do not work.<\/p>\n<p>Be careful not to over-claim. A single team&#8217;s before-and-after is weak evidence, since other things change over a month. Treat the result as a signal for a next iteration, not a proof.<\/p>\n<h2>Getting outside perspective when the loop saturates<\/h2>\n<p>Internal feedback has a ceiling because colleagues share your assumptions. When improvement flattens even though drills are well designed, the next gain often needs someone with a different vantage point. A coach who watches a ten-minute drill can usually name a habit the group has stopped seeing. If you need to find such a person, <a href=\"https:\/\/www.devopstrainer.in\/\">a trainer directory covering DevOps, SRE, DevSecOps, MLOps and AIOps<\/a> lets you look at instructors and coaches by domain and by the format you need, whether that is individual upskilling or a team workshop. Brief them on the specific drill and the baseline you recorded; vague briefs produce vague sessions.<\/p>\n<h2>Worked example: slow incident triage<\/h2>\n<p>Consider a group whose median triage takes forty minutes. Diagnosis: engineers differ in where they look first, a sign of missing feedback. Target: &#8220;choose the first three things to check in an unfamiliar alert&#8221;. Drill: replay a recorded alert in a sandbox, ten minutes, narrate aloud. Check: a short list agreed by the most experienced two engineers, scored by a peer. Change: one adjustment per round, such as checking recent changes before metrics.<\/p>\n<p>After four weeks the group&#8217;s replay time falls and narrations converge, which is the pattern you would expect if shared habits are forming. Real triage times will vary with other factors, but the drill gives the team a specific, observable lever to pull. For readers planning learning in adjacent fields, a staged plan like <a href=\"https:\/\/cotocus.org\/blog\/best-practices-for-learning-mlops-from-scratch-a-step-by-step-guide\/\">a step-by-step route into learning MLOps from nothing<\/a> shows how a broad subject breaks into small pieces that can each become a drill.<\/p>\n<h2>Objections you will hear, and responses<\/h2>\n<p>&#8220;We cannot spare the time.&#8221; Fifteen minutes twice a week is roughly the length of one status meeting; the cost is small compared with a task that consumes forty minutes every time it appears. &#8220;Our work is too varied to drill.&#8221; Varied work still contains recurring fragments, and fragments are what drills target. &#8220;Sandbox practice does not feel real.&#8221; True, which is why the drill should be followed within a week by applying the improved step to a live task and noting the result.<\/p>\n<p>A fourth objection comes from senior staff: &#8220;I already know this.&#8221; Often they do, in the sense of recognising the right answer. The drill tests whether they can produce it quickly, unprompted, under mild pressure. Recognition and production are different skills, and only the second shows up during an incident.<\/p>\n<h2>Residual risks<\/h2>\n<p>Three risks remain. Drills can become theatre if nobody is allowed to fail. A narrow focus can neglect judgment calls that matter in real incidents. And teams may drill only what is measurable, ignoring harder-to-score skills. Counter these with occasional open-ended exercises, rotating reviewers and an honest review each quarter of what the drills are not covering.<\/p>\n<h2>Summary of the diagnostic path<\/h2>\n<p>Symptom: skill does not compound with exposure. Probable cause: unfocused repetition, broad targets, weak feedback or bulk change. Remedy: short drills on one named weakness, with a check that can fail and one adjustment per round. Evidence: baseline and four-week comparison, read cautiously. Escalation: outside coach when internal feedback saturates. Run the loop for two cycles before judging it.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>A diagnostic walk from symptom to cause to remedy: why years of experience may not compound, and how short drills on one weakness can restart progress.<\/p>\n","protected":false},"author":4,"featured_media":1259,"comment_status":"","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[129],"tags":[],"class_list":["post-1258","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-devops"],"_links":{"self":[{"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/posts\/1258","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/comments?post=1258"}],"version-history":[{"count":1,"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/posts\/1258\/revisions"}],"predecessor-version":[{"id":1329,"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/posts\/1258\/revisions\/1329"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/media\/1259"}],"wp:attachment":[{"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/media?parent=1258"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/categories?post=1258"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cotocus.org\/blog\/wp-json\/wp\/v2\/tags?post=1258"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}