Tomas Vemola

How to ask for help (in writing)

Being someone who is easy to help is one of the greatest productivity hacks I know. You get more done. Your colleagues see you in a better light. And they are more likely to help you next time!

But most people still regularly commit the hi sin. Yuck.

Bad. Bad, bad, bad. Bad for you and the receiver. Because essentially, when you decide to ask for help, you're blocked until someone helps you.

Being blocked needs to be seen as a worse outcome than being wrong; it is the most useless of all failure modes.

boz.

If your goal is to get on with your work as fast as possible, I believe you should ask for help in this format instead (explanation following shortly):

You can scale these sections up or down depending on the size of your problem and how expensive it is to help you (coworker helping you on a task vs. getting more budget for a cross-team project).

Ask for help informed

Before asking for help, you should make a serious attempt at solving your problem yourself and documenting it. You'll find that getting help is mostly about asking in the right way.

Stop asking for help before even attempting to solve your problem yourself.

Put the bottom line up front (BLUF)

Judging by the first one or two lines of your message, the recipient needs to be able to tell what you're asking for, how important it is and what you expect from them.

That's basically Bottom Line Up Front. It's the last thing you'll write. Because only after you've written the rest of the message, can you properly summarize it into a crips BLUF.

Follow this format:

While working on [task] as part of project [project name], I [the problem you ran into]. I recommend [your recommendation]. Could you [action] by [deadline]? Not doing this will result in [consequences].

And a concrete example.

While work on the email campaign for Client X, I noticed that some of our assets in the email template are outdated. I think we should update them. Could you take a look at the template by Saturday to make sure I didn't mistake this for something else? Not doing this will mean we send out the old assets, not a big deal IMO.

Notice that for simple issues (like this example), writing a clear BLUF is quite enough to get help. However, remember that arriving at a clear BLUF is a feature of going through the rest of the sections.

Diagnosis (describe the problem)

For meatier issues, add more context. Your goal is to establish enough common ground with the recipient that they clearly know what problem you're facing. The better idea they have of the problem, the more likely it is that they will 1) respond at all, 2) respond in a way that helps you.

Try to be as specific as you can. You should describe the problem in detail, including easy access to surrounding context. Link to issue descriptions, lines of code, documents, anything that brings your helper closer to where you are.

In particular, pinpoint the specific step/part of the problem where you got stuck.

The best way I've found to find the "stuck" point is to note down the problem as it lands in my head and then gradually go into more detail. For example,

  1. This design sucks.
  2. The color scheme in this design sucks.
  3. The combination of red and purple in this design sucks.
  4. The combination of red and purple in the logos of our partner companies in the top left corner sucks.
  5. The combination of red and purple in the logos of our partner companies in the top left corner draws too much attention from the main message of the design.

or

  1. This code is garbage.
  2. This function is garbage.
  3. This function is incomprehensible.
  4. This function's parameters are incomprehensible.
  5. This function's parameters were written before we came up with the technical glossary which makes them incomprehensible to newcomers like me.

With a detailed description of the problem, your next step is to describe how you attempted to fix the problem, and couldn't.

In example 1

My first thought was to change the color scheme, but then I checked the brand guidelines of our partners. They say that the only permitted colors are red and purple. So then I tried to change the size and positioning, but I don't think that was an improvement.

In example 2

I tried updating the parameter naming to our new convention, but then I realized the new terms do not fit the way the function calculates the output. Since there are no tests and I am not familiar with this part of the codebase, I'm not comfortable changing the code.

Again, don't forget to link to any designs, documents, lines of code, spreadsheets or models. Add whatever you used to reason about the problem, to provide as much context as required to whoever is helping you.

Hmm, have you just solved your own problem?

At this point, with the problem clearly described, you might realize you can actually move ahead on your own - also known as rubber-ducking. If you have enough confidence to move ahead, but would like a bit more piece of mind, phrase your message as "Hey, I got stuck here and chose this solution. Is that ok?". Basically asking for approval rather than for help.

I find that often it's not that the problem I'm facing insolvable - it's more that I haven't organized what I know and don't know about the problem. By writing it out, I gain more clarity and often identify the next steps to make progress.

But sometimes you're just stuck. In that case, move on to Alternatives.

Alternatives (things you've though of, but can't execute)

Your job is not finished by formulating a problem and concluding it is still a problem. To get unstuck as soon as possible, you should write out alternatives you have considered for solving the problem.

Unless you're 100% clueless as to what could be the solution, identify a couple of not-too-shabby solutions that would solve the problem. This is different from the different solutions you've tried before - the Alternatives here should be actions that would solve the problem. But either you can't try them out (don't have permissions), or you're not sure how (unknown part of the codebase).

As with the previous sections, focus on being as concrete as possible. E.g. in case of a bad looking design, the alternatives are not "fix the design" and "not fix the design". A good set of Alternatives demonstrates you spent a bit of time in rigorous thought, considered the most suiting paths to solve this problem and described them in detail.

Going back to our example of red and purple logos

Diagnosis The combination of red and purple in the logos of our partner companies in the top left corner draws too much attention from the main message of the design.

Attempts to fix My first thought was to change the color scheme, but then I checked the brand guidelines of our partners. They say that the only permitted colors are red and purple. So then I tried to change the size and positioning, but I don't think that was an improvement.

Alternatives I see three alternatives. 1) You might have a quick fix? 2) Ask Joe for help - but I know he is crunching on project X. 3) Break the brand guidelines and recolor the logos to white.

Notice that you might discover new solutions to the problem you haven't tried yet. Go and try them! If they work, bravo, you have solved your own problem.

Recommendations (explicit best guess)

Next, choose the most promising Alternative - the one you'd make happen if you could, and offer it as your recommendation.

Coming up with a recommendation is absolutely crucial, because it forces you to have an opinion about the problem, to commit. Later, you can compare your recommendation to the actual solution. You can judge how far off your were. The person helping you will often have an easier time - if you recommendation is sensible, they will say "ok" and perhaps guide you how to make it happen. That's much more ergonomic for them than having to come up with a solution themselves.

The learning happens when you resolve the conflict between your recommended solution and the outcome. Without solidifying your guess into an explicit, the outcome has nothing to push up against, and you will just rationalize your way into believing you would pick this option as well. Worst case, you get a false sense of confidence.

Our example, including a recommendation:

Description The combination of red and purple in the logos of our partner companies in the top left corner draws too much attention from the main message of the design.

Attempts to fix My first thought was to change the color scheme, but then I checked the brand guidelines of our partners. They say that the only permitted colors are red and purple. So then I tried to change the size and positioning, but I don't think that was an improvement.

Alternatives I see three alternatives. 1) You might have a quick fix? 2) Ask Joe for help - but I know he is crunching on project X. 3) Break the brand guidelines and recolor the logos to white.

Recommendation I would go with interrupting Joe, unless you have a good idea. Breaking the brand guidelines seems like big trouble, would only do it as last resort.

Consequences (you gotta help, or else...)

To nudge the recipient toward action, include the consequences of doing nothing. Tasks not being done, deadlines slipping, processes not being followed, whatever it is. Everyone has their priorities and helping you is sometimes pretty low on the list. Putting a little danger in your message helps tip the scales in your favor.

Ideally, this consequence is important to the recipient as well. Imagine you're in their shoes. What do they care about? If possible, frame your problem toward their incentives (e.g. success of a joint project).

For example:

Recommendation I would go with interrupting Joe, unless you have a good idea. Breaking the brand guidelines seems like big trouble, would only do it as last resort. Consequence If the design is not finalized by Friday, we will have to push the go-live on social media for project X.

Action

Last but not least, ask for a specific action from your recipient. Often, this might be something like

"Please check my recommendation above, and let me know if you would do the same."

or

"When you are free, can we look at this together?"

Stay focused focused on how you can save your recipient some thinking. Asking for a specific thing to be done is a great way to receive a quick reply.

Notes on recipients

My main objective is to get stuff done. If I'm stuck waiting for help, I'm not getting stuff done. If you also care about getting stuff done, find someone who can get you unstuck in a timely fashion. Obviously, the recipient needs to have the capabilities to help you. You should know your audience! Don't ask the payroll manager to review your graphic design. On top, you will have a much easier time when yours and the recipients' incentives match up. After you pick the recipient, review your message.

The template

Putting all of this together, here's a template you can use as a starting point. Add and remove sections according to the hairiness of your problem. Good luck!

Hi [recipient's name],

While working on [task] as part of project [project name], I [the problem you ran into]. I recommend [your recommendation]. Could you [action] by [deadline]? If this is not handled, [consequences].

Specifically, [detailed problem description].

I have tried

As next steps, I see the following options

I recommend going with [your recommendation], because [reasoning for this recommendation].

If we are not able to handle this, [consequences of inaction].

Could you [action]?

Further reading

#communication #longform