This is the third of three posts on staff work. The first and second cover doing the work and preparing the package.
This one is about something I didn’t make explicit enough earlier: staff work isn’t just an officer problem. It is everyone’s problem because decisions depend on it.
By staff work, I mean the thinking, analysis and communication that help somebody understand an issue and make a decision. I’m not talking about the broader universe of supporting a boss, eg settling meals, travel and accommodation.
The officer may produce the first draft, but everybody who comments, rewrites or clears it becomes part of the work. We can’t clear a bad paper and complain later as though we had nothing to do with it.
Both sides have lazy shortcuts. Officers can click “accept all” without studying the changes. Managers can rewrite the whole thing because it is faster, sometimes without showing any tracked changes. Or we leave broad comments such as “sharpen this” and expect the officer to reconstruct the full thought process in our heads.
The paper may still get cleared. But nobody learnt anything!
To borrow the title of Josh Seiden’s book, it should be outcomes over output. The corrected paper is the output. The outcome should be a better decision, and hopefully, an officer who can do the next piece of work better.
To be clear, managers and officers need to pitch in. An officer wouldn’t respect a superior who can’t give useful guidance, can’t teach, can’t learn. The same applies to officers who keep receiving feedback but show no interest in learning from it.
A. For officers: comments are part of the work
Red comments can teach a lot, if you look
A document covered in comments, tracked changes and strikethroughs can tell you a lot about what good looks like. You can see what was removed, which point was moved forward, where a claim was narrowed and how the recommendation became clearer.
Some officers study the changes and apply the lesson to the next piece of work. Others click “accept all” (and maybe stop tracking) and move on. In the past that’s what I sometimes did… the next submission then comes back with similar problems.
So yes, red comments can teach a lot. They just can’t teach someone who doesn’t care enough to look at them.
Look at the context, not only the words
When someone comments on your work, consider:
- What is the context of the question?
- Does the person know something I don’t?
- Is there a wider issue beyond the exact sentence?
- Could the problem be with the evidence, phrasing, clarity or amount of detail?
- Am I about to become very extra and infer an entire hidden agenda?
Sometimes the words contain the whole question. Sometimes they point towards a concern the person hasn’t fully spelt out.
“Why did we not grant this?”
Suppose your boss asks:
Why did we not grant this?
The immediate answer may be:
The request didn’t meet criterion X.
Ya, technically correct. It also begs more questions.
Are we applying the criterion too tightly? Have we granted similar cases before? Is the refusal defensible? Does the person suffer a significant disadvantage? Did anybody screw up earlier in the process?
I would probably check the precedent and implications, then reply along these lines:
We didn’t grant it because the request fell outside criterion X. I checked the comparable cases. We granted one earlier exception because Y applied, which is different from this case. The refusal is consistent with our usual position, although it creates disadvantage Z. I didn’t find a processing error. If the concern is that criterion X has become too tight, we should review that separately.
Sometimes the boss genuinely wanted only the surface reason. That’s fine. You should still consider the surrounding issues before deciding how much to say.
A small comment can expose a bigger problem
A question about one statistic may mean the wider claim has no real evidence. “Can we say this?” may mean the sentence is too absolute, or that the reader knows of an exception you missed.
The useful change may sit somewhere other than the sentence that attracted the comment. Maybe the example is poor. Maybe the paper has become too convoluted. Important evidence may be buried in an annex when it belongs in the main text.
Don’t treat every comment like a ticket where your job is to fill in one field and send it back.
Reply with the answer and the edit
Where useful, answer the question and update the work at the same time.
For example:
The original phrasing was because of X. I agree with your comment on Y and have added the relevant cases from the last decade in paragraph 6. The overall recommendation remains.
This is more useful than:
Noted, will check.
It also avoids another round where the reviewer has to ask you to update the paper after you have already found the answer.
The suggested words are not always the fix
Bosses sometimes type a proposed sentence because they are thinking aloud. Their wording may solve the concern, or it may create another problem elsewhere.
You usually know the detailed context better because you spent days on the work and they read it between meetings. Understand what they care about, then make the best amendment.
A clear instruction remains a clear instruction. Every comment isn’t necessarily dictation.
Clarify when your interpretation would change the recommendation, create a lot of additional work or send the paper in a completely different direction. For smaller choices, use your judgement and move on.
Don’t go completely overboard
The think-ahead habit has a downside.
Sometimes the boss just wants a few more stats or is a bit jumpy. There may be nothing fundamentally wrong with the paper, so there’s no need to reopen everything and return with something twice as long.
Tweak what needs changing, stay calm and clarify before running off to do tons of work.
You will occasionally answer too narrowly. On another day, you will become very kiasu and do too much. Experience is partly remembering which way you overdid it last time.
B. For middle managers: you are part of the work too
Rewriting everything is also a shortcut
When a draft is bad and the deadline is close, rewriting it yourself is often faster.
I’ve done this, maybe too much on hindsight! Sometimes the paper simply has to go out. You already know roughly what it should say, and explaining every problem may take longer than replacing the whole thing. (I used to make peace with myself by thinking maybe they’ll compare both pieces of work, but obviously, I was lying to myself :’()
The officer then sees the final answer without seeing how you got there. They may not realise that you changed the objective, found a better angle, removed pages of accurate but irrelevant background, or softened a claim that the evidence couldn’t support.
The next paper comes back in the original form.
There are times when you need to take over. Just don’t confuse rescuing the paper with developing the officer.
Broad comments can be lazy
Another shortcut is to leave comments such as:
What’s the angle?
Please sharpen.
Rethink.
Sometimes that’s enough. An experienced officer who understands the issue and knows how you think may only need a short steer.
At other times, the comment is too broad to be useful. The officer can’t tell whether the problem is the objective, recommendation, evidence, structure or wording. They rewrite blindly and return with a different version of the same problem.
Tbh, I’m guilty of this quite often. It takes ten seconds to type “sharpen the angle”. It takes much longer to explain why the current angle doesn’t work and what the paper is supposed to achieve.
You don’t always need to give the answer. You should at least explain the problem clearly enough for the officer to work on it.
Go through the thinking when needed
The useful lesson is often not the sentence you changed. It is how you arrived at the change.
What is the objective of the paper? Is the officer actually answering it? Which facts matter to the decision? How firm can the recommendation be? Where does it need more nuance?
Then comes the tedious part of rewriting and rearranging the paper so that it reflects that thinking.
Going through all this can take longer than producing your own version. You can’t do it for every draft. If the same problem keeps returning, though, another broad comment probably isn’t going to save time.
Sometimes edit it together
Depending on the officer, sit down and edit the work together in real time. I found it really effective when I was a young one.
Explain the choices as you make them. Why did you move the conclusion forward? Why was the technically correct answer still inadequate? Why did you remove half the background? What did you think the boss was actually asking?
The discussion also helps you understand how the officer approached the task.
Maybe they misunderstood the assignment. Maybe they couldn’t tell which facts mattered. They may have rushed, or thought compiling everybody’s input was what you wanted. Sometimes they simply didn’t care enough, which requires a different conversation from another lesson on staff work.
There’s no need to explain every comma. Spend the time on the judgement they can reuse later.
Let the officer see the important changes
Tracked changes can look horrible, but they show what moved and what disappeared.
If you need to produce a clean version urgently, keep the earlier version and go through the main differences afterwards. Comments can explain larger changes that aren’t obvious from the redlines.
Without that explanation, the officer may conclude only that you prefer a different writing style.
See whether anything transfers
The next piece of work tells you whether the lesson landed.
Did the officer put the point first without being reminded? Did they check the obvious precedent? Are they still compiling everybody’s input into one long paper?
A corrected document may look excellent because the manager effectively rewrote it. That doesn’t mean the officer can now produce the same quality independently.
If the same problem appears for the fourth time, leaving the same comment for the fifth time is probably not the answer. Try a live edit or give a clearer example. If the officer understands but doesn’t care, address that directly.
Match the guidance to the officer
Some officers learn well from written comments. Another may need to talk through the issue before drafting. Someone else understands the general principle but struggles with a politically sensitive case.
Newer officers usually need more direction. As their judgement improves, give them more room.
Don’t take the work back merely because the officer wrote it differently from how you would. Their version may still work.
Your steer can be wrong
Managers sometimes infer what a senior person wants and guide the team accordingly. Sometimes that inference is completely off.
I’ve done this. Tell the officer, own the mistake and correct the work. Don’t quietly rewrite history so that the officer appears to have misunderstood an instruction that came from you.
Managers can also misunderstand the ground or overreact to one comment from above. The officer may know something you don’t.
Being more senior doesn’t end the need to learn.
C. The decision depends on the work
A polished paper doesn’t prove that good staff work happened. The manager may simply have rewritten everything. The disagreement may also have been worded away instead of properly dealt with.
The officer owns the first draft, but not the entire responsibility for what eventually goes up. Everyone who reviews, changes or clears the paper affects the decision that follows.
Officers need to care enough to learn from comments. Managers need to know enough to guide them, while remaining willing to learn themselves.
The paper is the output. The quality of the decision, and whether the people doing the work improve, are the outcomes.
That is why staff work is everyone’s problem.