Attention to Detail - Consolidated Advice and Recommendations

I personally struggle with Attention to Detail, and I figured I wasn't the only one. I searched through a handful of posts and comments to consolidate the most helpful comments that other people on this forum have posted.
If you guys have any other advice or insights and would like to add to this post, by all means, please help out and share what worked/didn't work for you. 
________________________________________________________________________Pepperoni Toni

Print out decks; inspect documents in excel, ppt, and word for comments, external links, etc.; get up and walk around for 1 minute to review your deliverable for a third time to get a fresh set of eyes on your work; if you're leveraging old decks/documents, CTRL + F search the old company name and other old references to make sure the doc is cleaned up, insert many checks into your models (not just a net income or balance sheet check, I usually do checks for each major category (current assets, current liabilities, etc.) to make sure I'm copying client financials in correctly. Those are a few that I do, but I'm sure there are many more out there.
 

Asatar

• Look over every number and think at a high level "does this make sense? Does it match the story?"
• When you're putting through a markup, highlight every comment after you've made the change. This will help you stop missing things.
• As you say, print off.

I personally like to do a couple of readthroughs of something I produce, looking at various things each time. For example:
- Pass 1 = formatting errors (footnotes, page numbers, double spaces, colours, alignment etc) (easy win)
- Pass 2 = read the words and see if they make sense (typos, grammar, readability etc) (easy win)
- Pass 3 = check all the numbers / and charts and make sure everything looks right conceptually
- Pass 4 = do the numbers and the text link up together
 

TechCoverageBanker

Yeah, I got yelled at by my VP / assoc on the first bakeoff I did until my attn to detail improved.

Some practical things you can do...
- Always print your books and review by hand backwards (last page to first page) with a red pen
- Make sure numbers tie across pages by pulling out the specific pages and highlighting them as you confirm they tie
- Whenever you're called out for missing something, write it down on a post it and leave it somewhere you'll see every day
- Use Factset's trace precedents tool to make sure numbers are coming from the right place
- Build a "check" into your model (subtract assets from liabilities - adj. format to show three dec. places)

alpha_q

I think others will have a more fulsome perspective, but attention to detail to me is something that some innately have (they can go through extensive pages of work product quickly, not review thoroughly, and still come away error free because they are quite diligent). I'm not in this bucket ;P

As with any other skill, attention to detail is something that most people gain and build with time in my view. It's really important to build the foundation first before experimenting with things like speed or faster cadence. When you are a first year, the expectation will be that you produce error free work, and then get faster at producing this error free work as you continue to get the necessary reps. Everyone has their preferences and best practices. For me, It was always doing an initial scan on the screen to check for immediate errors, printing out my work, and then taking a highlighter to note any changes/additions made so as to ensure that any comments are properly being reflected. If there are any minor nits that I may have missed I'll make edit and then print one more time before sending. The other key for me was a bit more implicit, which was taking a step back and understanding the big picture of what I'm doing. If you're just plugging away numbers into a model, looking up precedents, or turning through comments for a deck, you might end up missing a few things vs. if you had a degree of understanding for why you were doing it. It's much easier to find mistakes if your understanding was that a multiple of [ ]x was well above norms. It also allows you to anticipate the next steps in the process (this is more personally for me from an M&A perspective, but applies broadly as well), which reflects well on you.

It's a bit of a cumbersome process when you start, particularly because most are not accustomed to looking through their work to that degree of rigour. Speed might get sacrificed initially, but it's far more preferable for a new 1st year to have good product if it took them a bit longer vs. fast error filled work product. I imagine that you will very quickly ramp up the speed just given the volume of work that you will be given, and the early principles that you instilled will be very helpful in getting to a point where you are both proficient and with a keen eye for detail. I'm sure you'll be just fine :)

OTOH, there does come a point in my mind where it would be helpful to prescribe to 80/20, particularly when v48 of the document isn't meaningfully different from v08. But you didn't hear this from me :)