A young manager comes to you. The release is slipping. A good engineer has gone quiet in
stand-ups. Within sixty seconds you have found the problem, told him the fix, and moved on
to the next thing. That felt like mentoring. It was troubleshooting.
We in IT are trained to close tickets fast. Apply that same speed to a human being and you
build dependence, not capability. He comes back next week with the same kind of problem.
And the week after. Nothing has been transferred.
The First fix is you, not him
There is an old idea called ‘sthita pragya’ or a settled mind. The person who hears bad news
as information, not as an attack. This sounds like philosophy. It is not. Picture yourself in that one-on-one at 6 pm, after a P1 outage, a vendor escalation, and a board deck that came back covered in red. Whatever you are carrying, your mentee reads it in the first thirty seconds. He then tells you only the safe things. You go home believing the conversation went well.
Three things spoil us. Raga — it should happen my way. Krodh — why did it not happen my
way. And fear — what if I fail. Most conflict inside a delivery team is one person’s attachment colliding with another person’s anger.
Here is a simple test. How many days pass between a problem starting in your team and
you hearing about it? Write the number down. That number is your real standing as a
mentor. Not your engagement score.
Why This Matters More in IT Than we Admit
Every large ERP or platform rollout has the same post-go-live story. The system is live.
Training is done. And quietly, in the corners, the old Excel trackers are still running. Some
people use the new system, some ignore it, nobody says so out loud.
That is never a software problem. It is a behaviour problem that no amount of user training
will touch, because behaviour does not change through a classroom. It changes when a senior person notices one specific thing you did, says so plainly, and comes back to check.
Which is exactly what most of us have never been taught to do.
Write What you saw, not What you Concluded
“He lacks ownership.” That is a verdict. Nobody can work on a verdict.
“In Tuesday’s release call, when QA raised the regression, he said the sprint was already
closed and did not reopen the ticket.” That is an observation. That, he can work on.
Verdict, or observation? Most of what we write in appraisals is verdict wearing observation’s
clothes. Learn to feel the difference and half the job is done.
Decide What Counts
You tell your architect: spend more time with the junior developers.
Does a corridor chat count? Does “any blockers?” in stand-up count? Does a code review
with three comments count?
Until you answer that, you have handed him a feeling, not a habit. Write it down — counts
when, does not count. That single line is the difference between measuring something and
merely hoping for it.
And give him three habits. Not ten. Ten is what we produce when we have not decided.
Do not Judge at day 30
Your team’s reported defects jump from 2 to 15 in one month. Bad news? Usually it is very good news. At the start, a low number means silence, not health. People begin reporting what they used to absorb quietly. The count is supposed to rise first, and
fall later, once causes are actually fixed.
Behaviour becomes visible somewhere between day 30 and day 60. Numbers move
between day 60 and day 90. Declaring failure at day 30 is how good mentoring gets killed
early — and how bad mentoring gets away with it.
Krishna did not Give an Order
Arjuna is the classic high performer in collapse. Skillfully intact, conviction gone.
Krishna does not command him. He spends the entire conversation building a framework —
the nature of action, of duty, of consequence — so that Arjuna can reach his own decision.
Then, at the very end, he says: think it over fully, yathecchasi tatha kuru — now do as you
wish.
He hands the choice back.
That is the hardest act for a senior technology leader. We are the escalation point. We are
the final ok. Giving the decision back feels like abdication. It is the opposite — it is the
only way judgement gets built in anyone else.
What to do on Monday
Pick one person. Write down three things you actually saw — with date and place, no
adjectives. Choose two. Give it 120 days, not one sprint. Find one peer who will call you
every Friday and ask two questions: did the conversation happen, and what did you notice
that you would have missed a month ago.
A plan without a partner is only a wish. In ninety days you will not remember a single line of
this article. You will remember whether somebody called you on a Friday.
