A new analyst joins the team on Monday. HR pairs her with a senior colleague for three days of shadowing. She sits next to him, watches him work, listens to a few calls, and reads over his shoulder. On Thursday she is back at her own desk, feeling slightly more lost than before. Nobody told her what to notice. Nobody gave the senior analyst a framework for what to show. She watched a week of work and learned almost nothing.
This is the most common job shadowing failure, and it happens in companies that genuinely believe they are investing in development. Job shadowing works. The problem is that most programs are unstructured observation dressed up as training, and unstructured observation teaches very little.
This post is the honest version: what job shadowing is, when it develops real skill, when it becomes theater, how remote work changed the design, and what separates a good program from a bad one.
What is job shadowing?
Job shadowing is a development method where one employee observes another performing their work to learn the role, the reasoning behind decisions, or the tacit skills that never appear in a job description. The observer is the shadower. The observed is the host.
It sits between two other common methods. Training is structured and content-led. Mentoring is relationship-led and stretches across months. Shadowing is contextual: you watch real work happen and try to learn from it. That is also its weakness. Without design, watching is passive, and passive watching rarely turns into capability.
Common examples you have probably seen:
- A new sales rep sits with an experienced colleague for a week and listens to customer calls.
- A junior customer service hire shadows a veteran during live calls to learn how the team handles difficult conversations.
- A developer moving into a tech lead role shadows another lead across planning, review, and one to one meetings.
- A manager considering a move into people operations shadows the HR business partner for two weeks.
Each of these can produce real learning. Each can also produce a week of wasted calendars. The difference lies almost entirely in the design.
If you want a wider view of training methods this sits inside, the top 13 methods of employee training post is a useful map.
When job shadowing actually works
Shadowing works when three conditions hold together. Remove one and the program starts to wobble. Remove two and you have theater.
The work being observed is reasoning-heavy, not task-heavy. Watching someone enter data into a spreadsheet is pointless. Watching a senior account manager decide which objection to answer first in a stalled deal is gold. Shadowing earns its keep in roles where judgment is the skill. Judgment is almost impossible to teach through documents, so watching it in action is the point.
The host explains their reasoning out loud. Great shadow hosts narrate. They say things like, “I am going to open with a question about their renewal because I want to see if the procurement conversation has started yet.” That one sentence teaches more than an hour of silent observation. A host who cannot or will not narrate is a poor host, even if they are a strong performer.
The third condition is defined observation targets. Before the shadow starts, someone needs to answer: what should this person be able to explain at the end of this week that they cannot explain today? Without that answer, the shadower defaults to copying mannerisms and missing the substance.
When these three hold, a three day shadow can compress months of trial and error into a week. When they do not, even a month of sitting next to someone produces little beyond familiarity.
When it becomes theater
Theater is the polite word for programs that exist on paper, get credit in the onboarding deck, and produce almost no learning. A few patterns to recognize:
The host treats the shadow as an interruption. The shadower sits quietly, asks nothing, gets no context, and leaves with a vague sense that the host is busy and probably good at their job.
The work is routine. If the host spends their days doing the same thing you could learn from a wiki, shadowing them is the slowest possible form of reading the wiki.
There is no debrief. The shadow ends on Friday and everyone moves on. Without a conversation that turns observations into lessons, the shadower is left to assemble meaning on their own, and most people do a poor job of it.
HR measures completion, not learning. If the only metric is “shadow completed yes or no,” the program drifts toward the cheapest version of itself within two cycles.
The shadower is expected to be invisible. Good shadows involve questions during natural breaks, small tasks that let the shadower try something, and a host who pulls them into the reasoning.
If any of these patterns describe your program, the fix is not to cancel the shadow. It is to redesign the host expectations, observation targets, and debriefs.
Poorly designed vs well-designed shadowing
Here is the side by side. If your program reads like the left column, you have theater. If it reads like the right, you have a development method.
| Dimension | Poorly designed shadow | Well-designed shadow |
|---|---|---|
| Purpose | ”Get familiar with the role” | Three specific skills or decisions the shadower should be able to explain by the end |
| Host selection | Whoever is free that week | Someone who can narrate reasoning and wants to develop others |
| Host preparation | None. Host finds out Monday morning | Host gets a brief one week prior with the shadower’s background and goals |
| Observation structure | Sit and watch | Defined things to watch for each day, with space to ask questions |
| Shadower’s role | Silent observer | Active participant who asks, takes notes, and occasionally tries |
| Check-ins | None during the shadow | Short daily debriefs of 15 to 20 minutes |
| Closing | Shadow ends, everyone moves on | Formal debrief with the host, the shadower, and the shadower’s manager |
| Measurement | Completion rate | Shadower can explain decisions, apply them in their own work, and rate their own confidence shift |
| Follow-up | None | A 30 day check to see which lessons stuck in real work |
The left column costs nothing and produces nothing. The right column costs a few hours of design and a few hours of host time and can reliably cut months off time to productivity. The math is not close.
Two programs, same company
Same company. Same onboarding budget. Two shadows. Completely different outcomes.
The bad one. Rachel joins the customer success team. Her manager pairs her with David, a senior CSM, for three days. Nobody briefs David. He finds out Monday morning when Rachel appears at his desk, with three escalations already open. He tells her to “just watch and take notes,” and goes back to work.
Rachel sits behind him for three days, hears half of six calls, and cannot tell which of his activities are normal and which are exceptions. On Friday she has a page of notes about things David said. She cannot explain why he said any of them. Two weeks later her first hard conversation with a customer goes badly because the one thing she needed, hearing someone reason through a similar moment, never happened.
The good one. Next quarter, Olivia joins the same team. Her manager picks Peter as the host because Peter is not the most senior CSM but he is the best explainer. She sends Peter a one page brief a week early: Olivia’s background, the three skills this shadow should build, the decisions she should be able to explain by Friday. Peter blocks the days, cancels two non-essential meetings, and adds two debrief slots per day.
On day one, Peter walks Olivia through how he thinks about his book of customers before they watch a single call. On day two, he narrates two renewal calls in real time, then unpacks why he opened each one differently. On day three, Olivia sits in on a tough escalation. Afterward, Peter asks her to write down what she would have said and why, and they compare notes.
Olivia’s manager joins the Friday debrief. Olivia can explain three decisions she watched Peter make and the reasoning behind each. Thirty days later her first tough customer conversation goes well.
Same company. Same budget. Same role. One person wasted a week. The other compressed what usually takes two months into four days. The only difference was design.
How remote and hybrid work changed shadowing
Traditional shadowing relied on proximity. The shadower sat next to the host, overheard hallway conversations, watched body language, and picked up on things nobody articulated. Remote work removed most of that. Many HR teams responded by adding a shadower to Zoom calls as a silent attendee, which is not the same thing and does not produce the same results.
Remote shadowing needs a different design. A few things change:
Observation has to be explicit. On video, side conversations and visible reactions vanish. Remote shadows need written observation targets and a daily prompt like “what did the host do today that you would not have done?” to force reflection.
Narration becomes the whole show. A great remote host narrates constantly, either in the call or in a parallel chat with the shadower. “I am going to push back here because the customer’s silence usually means they are not the decision maker” is the kind of real-time sense-making that in-person shadows used to pick up by osmosis. Without narration, remote shadowing is almost useless.
Debriefs also get longer. When you are not walking to lunch together, the casual processing time disappears. Remote shadows need 20 to 30 minute scheduled debriefs instead of three minute hallway ones. Recording call segments and rewatching them together is often a stronger learning tool than in-person shadowing ever was.
Shadowers need more agency too. Remote shadowers lose attention faster than in-person ones. Good programs give them small tasks: drafting a reply, writing a meeting summary, preparing a question for the next call. Passive attendance produces passive learning, doubly so on video.
Remote shadowing can be as good as in-person, sometimes better, but only if it is designed for remote. A remote shadow that is just an in-person shadow with Zoom bolted on is worse than no shadow at all. If you’re onboarding remote hires more broadly, the buddy program at work guide covers how shadowing fits into the wider first month experience.
Benefits, when the design holds
When the design holds, shadowing produces a few specific results:
- Compressed time to productivity. New hires who shadow a good host reach baseline faster because they pick up the reasoning behind the work, not just the steps.
- Transfer of tacit knowledge. Every team has knowledge that lives in heads and has never been written down. Shadowing is the main way that knowledge moves.
- Lower early attrition. New hires leave early when the role turned out to be different from the one they were sold. A shadow in week one surfaces the gap before it becomes a resignation.
- Career exploration without role switching. Internal shadows let people test a role for three days instead of speculating about it for a year. This feeds into succession planning without any formal commitment.
- Development for the host. Hosting a shadow tests whether someone can explain their own work, which is a prerequisite for managing others. Good hosts often turn out to be your best emerging managers.
Run a weak program and none of these show up.
How to implement a program that is not theater
Most of the work is in the design, not the process. Five pieces matter.
Start with one question. What should a shadower be able to do after the shadow that they cannot do today? If you cannot answer, the program is not ready. “Learn about the role” is not an answer. “Explain how a tech lead decides which bugs to escalate, why, and to whom” is.
Pick hosts by ability to explain, not by seniority. The best hosts externalize their reasoning. Ask prospective hosts to walk you through a recent decision out loud. If they can explain it clearly, they will host well. If they freeze, no amount of briefing will fix that. Give your hosts a short brief on what good hosting looks like: narrate, welcome questions, debrief daily, expect participation rather than passive observation.
Each day needs two or three observation targets, written down in advance. For a sales shadow, day one might be opening moves in discovery, day two objection handling, day three a stalled deal. For a manager shadow, day one might be one to ones, day two prioritization, day three difficult feedback. Check in against the targets at each debrief.
Build in daily debriefs. Fifteen to twenty minutes per day. What did you see? What surprised you? What did you not understand? This is where observations turn into learning, and it is the step most programs skip.
On the last day, host, shadower, and the shadower’s manager meet for 30 minutes. The shadower explains what they learned in their own words. Thirty days later, a quick check: what stuck, what did not, did the shadow show up in real performance? Completion rate is not a measurement. Whether the shadower can explain the host’s reasoning and apply it to real work is.
If you want to go deeper on providing feedback and evaluation during programs like this, it applies directly here. Good shadowing is feedback-rich, not feedback-free.
Where shadowing fits in a wider picture
Shadowing is one method, not a program. It sits alongside training, coaching, mentoring, and job rotation. Use shadowing when the skill gap is about tacit judgment. Use structured training when it is about explicit knowledge. Use mentor matching when the gap is broader than a single role. Use job rotation when someone needs deeper experience across multiple functions.
Most companies do not need more shadowing. They need better shadowing, placed correctly inside a development plan that already has the other pieces.
Closing
Job shadowing is one of the cheapest development methods we have. It can also be one of the most wasteful. The difference is almost never the people and almost always the design.
If you run a shadowing program, pick one thing from this post and change it this quarter. Write observation targets. Pick hosts by ability to narrate. Add daily debriefs. Build the 30 day follow-up. Any one of these moves a mediocre program closer to a real one.
If you want to know where your team’s skill gaps actually are before you design any program, the coaching assessment and assessments for managers are free and take about fifteen minutes each. And if you want to see how Risely helps new managers and senior ICs build the skills that make them good hosts and good shadowers, try Merlin for free.
For broader context, the complete guide to managing your team is a good next read, and top skills for a new manager is the fastest way to get aligned on what “good” looks like before you start asking new managers to host shadows.
Shadowing is worth doing. It is just not worth doing badly.