DEV Community: Ana Ulin 😻

Adding Structured Book Data to My Blog Posts

Ana Ulin 😻 — Tue, 18 Aug 2020 21:45:28 +0000

I’ve been slowly working on taking more control over my own data, extracting it from corporate silos and keeping my own copies of it. One piece of this effort is making myself less dependent on Goodreads, which I’ve used for some years to track my reading and get inspiration from my friends’ reading. I enjoy the social and discovery aspects of Goodreads, but I don’t love that it is owned by Amazon and only gives you partial access to downloading your data. So I’ve been getting more consistent in posting my reading notes as blog posts to this site, and linking from my Goodreads reviews to those posts.

In order to keep collecting structured data about my reading , I’ve started adding it as metadata to the frontmatter of my reading notes entries. In the frontmatter of each post, I include a book section which contains the metadata I’m interested in (title, author, dates read, ratings, etc). This serves as a simple database for my reading, and I can autogenerate some of the content of reading posts from this data.

For example, the frontmatter for my recent notes on Are Prisons Obsolete? looks like this (in toml format):

+++
title = "Book Notes: Are Prisons Obsolete?"
date = 2020-08-15T15:26:53.454775
tags = ["books"]

[book]
title = "Are Prisons Obsolete?"
author = "Angela Davis"
url = "https://bookshop.org/a/6402/9781583225813"
start = 2020-08-11
end = 2020-08-14
rating = "5"
image = "/img/book-are-prisons-obsolete.jpg"
+++

In the Hugo template for a post, I use this metadata to generate a little card with the info for the book. I have a separate book.html partial template that I use for this, which gets invoked both by the standalone post template and the RSS feed template, if there is a book element in an entry. A simplified version of this Hugo template looks like:

{{- $title := index .Params.book "title" -}}
{{- $author := index .Params.book "author" -}}
{{- $rating := index .Params.book "rating" -}}
{{- $image := index .Params.book "image" -}}
{{- $url := index .Params.book "url" -}}
{{- $start := index .Params.book "start" -}}
{{- $end := index .Params.book "end" -}}

<div class="book-card">
  <a href="{{ $url }}"><img src="{{ .Site.BaseURL }}{{ $image }}"/></a>
  <br/>
  <a href="{{ $url }}">{{ $title }}</a> by {{ $author }}
  <br/>
  Read {{ $start.Format "Jan 2, 2006" }} - {{ $end.Format "Jan 2, 2006" }}
  <br/>
  {{ "⭐" | strings.Repeat $rating }}
</div>

I like how simple this is to understand and to maintain, and I’m excited about the possibilities it opens up. Now that I have reading data in a programmatically-accessible format, I can use it to compute reading stats, or to automatically build a standalone page listing books read. My next step will likely be importing all of my Goodreads data (or as much as I can get) into this format.

Adding Webmentions to My Static Hugo Site

Ana Ulin 😻 — Wed, 22 Jul 2020 19:10:18 +0000

As another step in bringing indieweb ideas into my site, I have added support for webmentions. If you want to mention a post on this site, you can send me a webmention via this endpoint.

A webmention is a kind of linkback. It is similar to pingbacks or trackbacks, but with a more modern sensibility that wants to extend @-mentions out of silos and to the open web.

The idea is straightforward: any URL (the source of the mention) should be able to mention any other URL (the target of the mention). The target gets to choose what to do with the mention: they can ignore it, publish it on their own site, use it to increment a counter, etc.

In order to receive webmentions you provide an HTTP endpoint that listens for mentions. After you’re able to receive mentions, you need some way to process them. This workflow could be fully manual (e.g. you read your endpoint’s logs, then write and mail thank you cards to the sources of the mentions), but automating this with some code is probably less work in the long run.

After some research, I opted for what seemed the easiest path:

Use webmention.io to get a free, no-hassle webmention endpoint for my domain.
Use webmentions.js to display the mentions received by the webmention.io endpoint on my posts.

Getting a webmention endpoint and advertising it

This site is staticly generated (I use Hugo), so there is no “running server” to which I could easily add an endpoint. I could add a CGI script to my webserver, and use that as my endpoint, but that seemed like more work than I wanted to do for this experiment.

Instead, I decided to use Aaron Parecki‘s free webmention.io service. You sign up with your domain, and voila: webmention.io gives you a personalized endpoint that friends and trolls can send mentions to.

After signing up for a webmention.io endpoint, I added a link to it from my site’s HTML header, so that anyone looking to send me a mention can discover which endpoint to use:

<!-- in HTML header -->
<link rel="webmention" href="https://webmention.io/anaulin.org/webmention" />

Displaying mentions with webmention.js

Continuing with the theme of “what is the easiest possible thing I could do”, I chose to use fluffy‘s excellent webmentions.js library to show on each post the mentions it has received.

This is a neat piece of drop-in JavaScript that knows how to query webmention.io to retrieve mentions for the current URL, and then displays them on the page. It all happens in a visitor’s browser when they load a post.

To integrate webmentions.js:

Include the library’s JavaScript file somewhere in your site.

<!-- in HTML header -->
<script src="/js/webmention.min.js" async></script>

Add <div id="webmentions"></div> wherever in your HTML you want webmentions to display.

You can add some custom CSS rules for this div, which is where webmentions.js will be adding the actual webmention content. The library’s GitHub page has good documentation on additional options.

How does it feel?

This was surprising easily to set up. The hardest part was researching what was available and what would best fit my needs. I wasn’t able to find a great “Webmention 101” guide, and instead had to spend some time sifting through the IndieWeb wiki and other websites, until I was able to form a rough idea of my options and what would work with minimal investment on my part.

The biggest disadvantage of this setup is that I don’t own any of my mentions data. If webmentions.io disappeared in a puff of smoke, I would lose access to all the mentions this site has received, in addition to no longer having a webmention endpoint. Right now this doesn’t concern me.

If over time I end up receiving a healthy amount of interesting/amusing/valuable mentions, then I’ll consider automating backing up the mentions from webmention.io. It wouldn’t be too hard to add a step to the site publishing process where I download the mentions from the server, and then store them in some structured way with each post (e.g. in the front matter).

Using Your Site As Your Login

Ana Ulin 😻 — Sat, 18 Jul 2020 17:35:10 +0000

The most broadly useful technology I've encountered in the Indieweb world is the ability to use your personal site as your login on other sites.

An accessible way to do this is RelMeAuth. The idea is beautifully simple. A service that wants to authenticate you can look at your website, read any rel="me" links you've added to it, and use those to verify your identity.

For example, if you have a link to your email address, someone trying to authenticate you can send a one-time password to that email. You can then use that password on the sender's site to verify that you are in fact the person who owns that email. Similarly, if you have a link to your GitHub account, one can use GitHub's OAuth to verify that you are who you claim to be.

If you have a website, and want to be able to use it to log into other sites using RelMeAuth, you need to do two things:

Add a link on your site to whatever profiles you want to use for identifying yourself, with a rel="me" attribute on that link. For example, to use my GitHub account I'd do:

<a href="https://https://github.com/anaulin" rel="me">anaulin</a>

Make sure that those profiles link back to your site, also with a rel="me" attribute. (For the case of an email, you can't "link back", so in that case a one-time challenge method is used to verify ownership.)

A related, more complex idea is the IndieAuth protocol. IndieAuth extends the OAuth protocol by providing a way to discover authorization endpoints programmatically. In this way, a site doesn't have to implement a zillion "log in with X" flows, but instead can have one universal "log in with your own endpoint" flow.

Supporting IndieAuth requires you to run your own authentication endpoint, and so requires a bigger investment than RelMeAuth (the Indieweb Wordpress plugin seems to make this easy for Wordpress users). There are two services, IndieAuth.com and IndieLogin.com, which can help you get started supporting IndieAuth. IndieAuth.com gives you an IndieAuth authentication endpoint that you can advertise on your site, and IndieLogin.com lets you add sign-in to your site that supports IndieAuth. (IndieAuth.com implements its authentication by in turn relying on RelMeAuth, further muddying the waters of this IndieWhatever and SomethingAuth word soup.)

I enjoy how simple RelMeAuth is to implement on my site, and I appreciate that it saves me from having to create yet another username and password for indieweb services. Since it supports email, I can be free from having to rely exclusively on third parties to verify my identity. I like it.

Updated July 25 2020: Clarified distinction between RelMeAuth and IndieAuth.

Adding Syndication Urls to My Posts

Ana Ulin 😻 — Tue, 30 Jun 2020 19:31:16 +0000

Inspired by this weekend’s IndieWebCamp West, I added support to record and display syndication urls for my posts on this site.

The indieweb¹ community encourages posting your content to your own site, and then liberally sharing either copies or links elsewhere on the web. The idea is that you own your content, and you can also give it wider distribution via social media or news aggregrators. They call this POSSE (Publish on your Own Site, Syndicate Elsewhere). Once you share your content “off-site”, it is helpful to visitors if you indicate on the original post where else tt can be found.

I don’t provide a way to comment on my site, but I do share some posts on social media and in other places that are more conducive to conversation. I thought it would be nice to point visitors at those other places where you can find the content and maybe even interact with me.

To capture the URLs, I decided to use a new field in my posts’ frontmatter, named syndication_urls and shaped as a list of strings. Each string in the list represents an alternate URL at which the post is syndicated. The metadata for an example post with syndication URLs looks like:

+++
title = "Pomodoro Technique as Self-Awareness Tool"
date = 2018-11-01T16:57:52-07:00
tags = ["working better"]
syndication_urls = ["https://dev.to/anaulin/pomodoro-technique-as-self-awareness-tool-2n0k", "https://twitter.com/anaulin/status/1058147565647134721?s=20"]
+++

[...post content...]

In the templates that generate the HTML for a post, I display these links like so:

{{ if .Params.syndication_urls }}
  <div class="syndication">
      <i class="fas fa-link"></i>
      This post was also syndicated to:
      {{- range $index, $url := .Params.syndication_urls }}
        {{- $parsed_url := urls.Parse $url -}}
        {{- if $index }}, {{- end }}
        <a class="u-syndication" href="{{ $url }}" rel="syndication">{{ $parsed_url.Host }}</a>
      {{- end }}
    </small>
  </div>
{{ end }}

It’s somewhat ugly, as Hugo‘s Golang-based templates tend to be. But it is simple and it works! As a bonus, it includes the requisite microformats to indicate that these are syndication urls.

Once the post has been rendered, the links looks something like this:

I’m pretty happy with it.

An unexpected bonus is that now I will have a structured way to collect this data, which might come in handy in the future. A future enhancement I’d like to add is replacing the plain text link with nice icons for some well known sites, like Mastodon or Twitter.

I don’t love the term indieweb, but I’ll use it for lack of a better one. To me, indieweb is just, well, web. The ideas of owning your own data and making your own site (and having fun with it!) existed well before someone invented this term. The prefix indie, in particular, feels like trying too hard to say “hey, i’m the small guy, i’m the good guy”. I guess folks wanted to be very clear that this is not Facebook or Medium or some other Big Tech Vortex of Doom, dammit. Fair enough. You can read more about this indieweb business on the IndieWeb wiki. Naming quibbles aside, there are good ideas in there. ↩︎

This post was originally published on https://anaulin.org/blog.

How To Write Better Job Descriptions

Ana Ulin 😻 — Tue, 20 Nov 2018 16:31:02 +0000

A poorly-written job description will not only lose you good candidates, but can end up with you hiring someone with mistaken expectations about the role.And yet, it seems we often forget even the basics of decent writing when we are working on a new job description.

Conversely, I’m aware of certain companies and their job postings only because they were so thoughtfully put together that others shared the descriptions with me as shining examples of the craft. Examples that come to mind include this job post from the Recurse Center, where they explain in detail and plain English what the job entails, the pros and cons of it, what the interviewing process will look like, and even details like their usual work hours. Similarly, this job post from Splice, while much briefer, also does a great job of giving you a sense for the company, the job, and the interview process.

Below are some suggestions for writing more effective job posts. This advice applies mainly to hiring managers who have control over the job descriptions for their open roles. If you’re at a larger company, where a larger machine posts these things outside of your control, you can still add your own summary when you re-share the job posting with your own networks, or when you have your sourcing team reach out to candidates.

Be the actual human you are

We say in management that employees leave managers, not companies. Conversely, candidates will choose a certain job to a large degree because of the prospective manager. Let this manager — yourself, if you’re the hiring manager — shine through in the job post.

A common pitfall is that, in trying to be “professional”, we use jargon and start sounding like a drone. Are you trying to hire someone who wants to be around corporate drones? If so, go ahead. Otherwise, try to use language that is precise, but also language that feels natural to you.

Think of how you would talk to someone in person. Would you say “do you have executive-level experience” to someone in an interview? Or would you say something more precise, such as “do you have experience managing directors” or “do you have experience setting a long-term strategy for a large team?”.

Speak to your intended audience

This implies that you have to think about your audience first. Hopefully you’ve done this already, as part of deciding which qualities are you looking for in this role.

So think about who do you want to appeal to. Picture your ideal candidate. What are their interests, passions, curiosities? What might be their concerns?

Try to address those, proactively, in your job posting. If you don’t know how to start, keep it simple: “As a , you will get to do X, Y and Z exciting things. But don’t worry, we don’t expect you to do A nor B. Occasionally, you might be called to help with P or Q.”

Be detailed and transparent

The more concrete you can make the job, the more compelling the posting will be to the right candidate. Paint a picture, and make it as textured as you can. Be open about the parts that are hazier, and about the parts that might change.

For example, don’t just say “you’ll have to help with recruiting”, say “our business is undergoing strong growth and we want your help expanding our team by X % in the next Y months”.

Some might object to making such strong, concrete statements in a job posting. There are some very sensitive goals (e.g. concrete revenue numbers) that you might not be able to share publicly, but those tend to be rare. Again, it helps here to think about what questions would most candidates have, and address those in your posting.

Be upfront about what the application and interviewing process will be like: how many interviews will they be, and of what kind; what response times may the candidate expect, etcetera.

Place the job role in the broader context of your company and industry

Realize that a prospective candidate might come across your job description knowing nothing about your company or your industry.

Be sure to talk about what your company does, what stage is it at. Mention any salient details about culture and work style. Again, you are trying to give the candidate a clear sense of what working will be like at your company.

Use inclusive language and encourage applicants of underrepresented groups

It’s 2018 and you care about building a diverse team. Use welcoming and inclusive language — you can use tools like textio and joblint to help you catch any glaring jargon, male-coded language and more.

Avoid listing requirements that aren’t actually required. If it is a nice-to-have, say so, or maybe omit it altogether. Remember that candidates from under-indexed groups will tend to only apply if they feel that they fulfill 100% of the requirements, so the more unnecessary requirements you add, the more you’re skewing your applicant pool towards those that veer towards the overconfident.

If you are interested in under-represented candidates applying, go ahead and say so directly in your job post. You can use the stock phrase “Women, minorities, people with disabilities, and veterans are encouraged to apply”, or maybe rephrase that in a way that sounds more “you”.

Remember that you’re selling this job to the candidate

Make the job description lively, compelling, welcoming, intriguing, warm, fun.

As you know, dishonest or sleazy salespeople don’t do well, so avoid falling into over-selling or obscuring the truth. You’ll be building a long-term relationship with whomever gets the job, and you won’t want to start with a foundation of bent truths.

This post was originally published on https://anaulin.org/blog.

What Are We Saying When We Say 'Professional'?

Ana Ulin 😻 — Mon, 19 Nov 2018 03:41:03 +0000

I hear people referring to “professionalism” — “he is being unprofessional”, or “she always acts professionally” — and yet I don’t think we really know what we’re talking about when we use that term. I say “we” because I’m guilty of this too. Cat knows that I’ve used “you’re being unprofessional” as shorthand for “you did something I didn’t like”, but I am trying to do better.

The idea of “professionalism” is used in the workplace¹ to keep us in line, to make sure we don’t color outside the lines. But when we say “act professionally”, we are seldom pointing at anything that has to do with the actual way that person performs their job — their profession. For example, when we say about a software engineer that she is not “being professional”, we never mean by that that her code lacks documentation or is unmaintainable or that she doesn’t collaborate well with others. Typically we use the term to mean that the subject was somehow irresponsible or unreliable, perhaps too emotional or “irrational”². We use it as a shorthand to mean “not as you should be”, and are too lazy to define exactly what are we trying to say, what is it that we want to change.

“Professional” means different things in different environments, and we never bother to clarify in which ways is it different here vs there, we just expect employees to “pick up on it”. Working at Google, at first I marveled at the wonderful freedom from the conventional trappings of professionalism: you don’t need to wear business clothes, nobody cares about your appearance and you are free to do good work and be judged solely on its merits. But gradually I came to realize that it wasn’t quite true — it’s not that at Google there isn’t a norm of “professionalism” that you have to conform to, it’s just that it is a different uniform (and there are different uniforms for employees in different functions — what’s ok in sales is not ok in engineering, and vice versa). So maybe there is a little bit more leeway in what is acceptably “professional” at Google, but there are still boundaries. You are welcome to be weird, but only as long as you are the right kind of weird — you definitely don’t want to be an engineer wearing a blazer to work, or appearing too feminine, or admitting that you vote Republican.

One piece of this idea of “professional” is this image of the very reasonable, perfectly logical person. There is an ideal here of the perfectly rational white man. It means that excessive emotion is not welcome, except in the places where it is required: you should be passionate about the mission, and the company itself, of course. But be sure to not appear too frustrated or upset, or you’ll be deemed childish or lacking grit.

In the classic management book Peopleware, DeMarco and Lister tell the story of a company where upper management sent a memo to employees forbidding the popping of corn in the breakroom microwave. “Popcorn is not professional”, the memo said. DeMarco and Lister go on to posit that “unprofessional” is a term used by weak managers to characterize “surprising and threatening” behavior, anything that is “different from average”. It seems to me there is something here about control from management, about not wanting to consider whatever is happening on its own merits (or lack thereof). The shorthand of the conventional or ordinary — “average” — as a substitute for thinking.

Tema Okun’s paper White Supremacy Culture, in which she describes common characteristics of white supremacy that show up in organizations, touches on several aspects that are closely related to this idea of “professionalism”. Okun says, “The characteristics […] are damaging because they are used as norms and standards without being proactively named or chosen by the group. “ The sections on “Objectivity”, “Only One Right Way” and “Paternalism” feel very relevant. The paper is filled with a treasure trove of “antidotes” to these characteristics. As a taste, here are Okun’s proposed antidotes to the “Objectivity” pattern:

realize that everybody has a world view and that everybody’s world view affects the way they understand things; realize this means you too; push yourself to sit with discomfort when people are expressing themselves in ways which are not familiar to you; assume that everybody has a valid point and your job is to understand what that point is; recognize that we can know things emotionally and intuitively in ways that we may not be able to explain “rationally;” understand that often “rational” thinking is actually an emotional response couched in logic

So next time you want to use the word “professional”, stop and ask yourself what is it that you really mean. Find a more accurate word for what you want to convey. Then, if this is still something you want to express, go ahead and use that more precise, less wishy-washy word.

This post appeared originally on https://anaulin.org/blog/.

The contemporary, American workplace, that is. YMMV. ^[return]
I can’t help but put quotes around “irrational”, because I don’t think we really know what we mean by that word, either. But that’s a whole other conversation. ^[return]

Tactics to Keep Up Your Motivation

Ana Ulin 😻 — Sat, 10 Nov 2018 06:12:10 +0000

Last week, I decided to use November and its NaNoWriMo as an excuse to challenge myself to publish some writing every day. Today is day 9 and this is my 9th blog post this month. I can feel my energy and motivation starting to flag.

I’m reminded of this article I read a few days back (and which, sadly, I can’t find right now) , by a NaNoWriMo veteran, where she talked about understanding and being prepared for what she called “the rhythm” of the challenge. We start energized, but about 10 days in or so things start to look worse — we’re tired and maybe even really disliking our writing. I am at the point where I can feel the doubts creeping in: “is this any good?” “what’s the point of doing this?”.

But I’m trying to push through, so I thought I would share some of the tactics I’m using to keep myself chugging along. These are some of the weapons in my inventory, perhaps some of them will work for you.

Note to self: +10 motivation and +5 energy

When this all started, I knew that maintaining momentum would be hard. So I wrote down a note for myself with detailing why I had decided to take up the challenge. Now, when I start thinking I should give up, I go and read my note and see if those motivations still hold. Turns out, they do. (It is interesting to note that some days some of my motives resonate more than others, and on the next day it will be the other way around. So it helps to have more than one reason.)

A buddy or two: +20 accountability and +10 energy

I shared my intention to write every day with some of my friends, and with my partner. Their encouragement and affirmation is very helpful. Yesterday I found myself saying “I haven’t written yet, I think maybe I’ll skip today”, but someone reminded me that it is about keeping momentum, not about producing something perfect. And that was enough to get me to write just one day more.

A timer with a short interval: +15 energy

I’ve been setting myself a 25-minute timer, calling it a “writing sprint” and typing away without editing. I’m willing to commit to 25 minutes to most things; 25 minutes doesn’t feel so hard. Once I’ve started, the energy and the words start flowing. Like now. (I wrote about my experiments working with a timer here.)

Light at the end of the tunnel: +20 motivation

I’m leaving on a trip tomorrow evening. I’ll be away for most of the week at a retreat center. I tell myself that, if it doesn’t feel right, I don’t need to be publishing anything while I’m on retreat. The retreat itself is about introspective writing, so I know I’ll be writing anyway (just not necessarily publishing). I can decide once I’m back if I want to pick up the challenge again or not. It helps to know that it is just a couple more days, that this daily challenge is not forever.

Community: +15 energy and +20 joy

One unexpected thing that’s been really helpful is the engagement of the dev.to community (where I’ve been cross-posting my writing). My original intent was to reach a broader audience (as far as I know, only my brother-in-law reads my blog; thanks Abe!). But the response on dev.to has been very supportive. Just having one or two thoughtful comments on a post make me feel like it is worthwhile to do this. (It is also, frankly, a little scary! Who am I to be having Opinions in Public? I might be wrong!) The responses have also helped me keep the conversation going, thinking and writing about more topics.

Originally published on https://anaulin.org/blog/.

Things Nobody Told Me About Being a Software Engineer

Ana Ulin 😻 — Fri, 09 Nov 2018 06:16:39 +0000

My dad started teaching me Basic when I was about 7 years old, and since about that age I knew that I wanted to build software for a living. It was obviously a fun and useful thing to do, and something that I could do well. But there is a big — and I mean Grand Canyon-scale BIG — difference between “programming” and working as a professional software engineer.

Here are, in no particular order, some things that nobody told me about working in tech as a software engineer:

That it is normal to have more lines of tests than production code.
That having no choice but to use vi early in my career would mean that I can never leave those keybindings behind.
That we spend a ridiculous amount of time setting up builds, releases and general tooling. Sometimes as much or more time that you spend coding. (And don’t even get me started about operations, on-call and all the other pieces that turn “programming” into “a product in hands of users”.)
That my gender or my age or my ethnicity or my sexual orientation or my weight or my clothes might (will!) have an impact on the perceived quality of the software I build. (Or, in other words, that this is not really a meritocracy, and doing a good job is not nearly enough.)
That writing code is only a small part of what goes into shipping production software.
That so much of my work life would be subordinate to the choices of barely-clued-in VCs, and overwhelmed first-time founders with no industry experience.
That everything old is continuously made new again — from MySpace to Facebook, from IRC to Slack. PHP is cool again, and functional programming is trending.
That having decent people skills makes my technical skills suspect, in the eyes of some.
That learning to manage my own energy and mindset is the most important skill through gnarly debugging and hairy code reviews alike.
That you can have more than a 100% base salary difference doing the same job, depending on if you work at a small startup or one of the large companies.
That not all software engineering roles have the same cachet in the eyes of the local culture. (Alternatively: that good frontend and test automation specialists are grossly undervalued.)
That the idea that working at a small startup will make you rich is outdated. Barely anyone IPOs these days, and you can save up more easily if you work for a bigger, stable company.
That CSS is the most complex modern programming language.
That internships should always be paid at market rate.
That spunky online book vendor would become synonymous with distributed infrastructure. That the new Altavista competitor would become a less bad Microsoft Office. That the new version of My Space would help damage democracies the world over.
That, 15-ish years in, my main professional contributions could be summarized as: nudging folks to write better and/or more tests, and getting people to talk to each other.
That the best technology doesn’t necessarily win. In fact, it often doesn’t.
That I would be this suspicious when the tests pass on first try, and have to invert my assertions to force a test failure and convince myself that things are working as intended. (Aside: Why is this not a standard feature of testing frameworks? I want some way to re-run tests flipping some of the assertions, to make sure they are testing what I think they are.)
That appropriate use of emojis would become such an important communication skill.
That it is all people, all the way down. They are (mostly) lovely, and not nearly as hard-shelled as turtles.

What is something that surprised you about working as a software engineer?

Originally published on https://anaulin.org/blog/.

Why Do We Have “Performance Evaluations”, Anyway?

Ana Ulin 😻 — Wed, 07 Nov 2018 18:57:02 +0000

In the follow-up to my recent posts about performance management and judgement vs metrics, someone commented:

And in fact, the whole idea of routine performance evaluations is suspect here. Why is performance being evaluated? Do we mistrust our management staff? Was there a problem? Was there a huge success? Or are we doing it because we've always done it or business magazines say we should do it? What's wrong with defaulting to paying people market value for the work they do, then dealing with exceptionally good or bad performances as they occur?

These questions really sunk their hook in me. As an engineering manager, formal performance evaluations are my least favorite part of the job (my second least favorite is interviewing, which is sort of like a formal performance evaluation, but for someone that you’ve only known for an hour). Formal evaluations feel very different from one of the pieces that I love best: supporting someone to grow in the ways that fit them best. You’d think that encouraging employee growth and helping your team level up is what performance evaluations are about, but it doesn’t feel that way — the formal process tends to focus more on the broader organization than on the individual and their needs (or even their team’s needs), and comes with big overheads.

One big piece of formal performance evaluations that companies care about is that they be done as systematically and homogeneously as possible across the organization. The idea is to try to be as objective as possible. The process of getting all of these evaluations discussed and honed and equalized takes a fair amount of work — going as far as long meetings where managers debate evaluation scores down to the decimals, as in Google’s “calibration” meetings. The resulting scores feed into things like compensation formulas or promotions, but have little bearing on what really matters: improving how we work.

So, could we do without this process? Maybe.

We could delegate to individual managers monitoring for exceptionally good or bad performance, as the comment above suggests. One possible pitfall is lack of consistency across managers, which in turn could lead to unequal treatment of people on different teams. We would want to have some level of transparency and checks-and-balances, a way for managers to share their assessments with each other and align on a shared standard. Probably we’d want to document this, so we would write down what we mean by “exceptionally good” and “exceptionally bad”, and share that with everyone on the team. Suddenly it seems we have something that looks a lot like a nascent performance review system — perhaps a kinder one, with no decimal point scores, but still something that looks eerily familiar.

In my ideal world, we would work in a way where we can set compensation based on criteria as objective as possible, including market rates and some kind of measure of expertise (and a way of adjusting these over time). It would be amazing if we could trust managers to do a good job of helping their teammates grow, to recognize exceptionally good performance when they see it, and to correct poor performance as soon as it appears. This does involve, first and foremost, trust, and a caring management team that is working hard to be as unbiased as possible. Unfortunately, most executive-level leaders I’ve worked with in tech seem to have trouble getting to this level of trust towards their managers (as I’ve recently written about here).

Some younger companies are experimenting with these ideas. Buffer, for example, says they no longer do performance reviews. Others are playing with tweaking the format and frequency of performance evaluations. However, I have yet to see a company with more than a couple of hundred employees that has figured out a way to get rid of some sort of periodic systematic performance assessment. I am hopeful, though, that over time more companies will experiment in this space and come up with better ways to handle employee performance and growth.

There Is No Substitute For Good Judgement, Especially if You Want to Be Data Driven

Ana Ulin 😻 — Tue, 06 Nov 2018 18:10:45 +0000

My recent entry on measuring a software engineer’s performance led to an interesting comment: but if you rely only on judgement, isn’t that a recipe for bias?

It’s a really good question. We want to be as objective as possible in our evaluations. We want to make sure that we work against our prejudices, assumptions and biases. But is finding a metric or statistic a recipe against those pitfalls? I don’t think so.

To use a metric to come up with a decision or evaluation, we still need to apply our own judgement to make sure that this measure is accurate and objective, and does not introduce its own bias. And then you have to apply your own judgement to make sure you are computing and applying this metric correctly.

It would be wonderful if we could determine everything by fully objective data-driven measures, but the reality is that even interpreting numbers still requires judgement.

We can see examples of this everywhere. In the social sciences, studies are often re-interpreted from a new perspective, with the same data leading to different conclusions — see, for example, the new analysis of the Stanford Marshmallow Experiment , or Gina Perry’s re-examination of the Milgram Experiment, both of which cast doubt on the original findings. Perhaps the most startling examples of the influence of judgement on interpreting numbers is what we see in politics: the same statistic is often used to support opposing stances, the conclusions highly dependent on the worldview of the person analyzing the data.

I would argue that good judgement is even more important if you’re trying to use a seemingly-objective metric. When you're using numbers, it is easy to lull yourself and your audience into a false sense of security (“look, the data proves this!”) and come to the wrong conclusion.

So yeah, as scary as it might sound for those who insist that they can be perfectly objective and data-driven, there is no substitute for good judgement. If we want to avoid bias, there is no replacement for introspection, asking for and listening to feedback, and keeping an open mind. There is no magic metric that will save us from the hard work.

The Demands of 'Executive Presence' in the Workplace are Exhausting

Ana Ulin 😻 — Tue, 06 Nov 2018 03:39:26 +0000

Yesterday I attended “Executive Presence for Women”, a Stanford workshop ran by Allison Kluger. It was good, and I would recommend it to any woman that wants to grow in her career and is looking for ideas to hone how she is perceived in the workplace.

Allison gave good tips not only on how we present ourselves (choice of words and tone, physical appearance, body language, etc), but also on preparation, not asking for permission and other tactics to help us get heard and recognized as leaders. These ideas are helpful even if you are not interested in being a leadership position — they can help you be more effective overall.

The main caveat is that you’ll have to decide for yourself which of the tactics discussed will work for you (as always, no substitute for good judgement!) — for example, some of the tips around dress code and appearance seemed to me a little antiquated and not a good fit for my role as a technical leader in a tech company.

Despite all the inspiring and interesting content, by the time the workshop ended at 4 PM I was tired beyond words. It wasn’t just the dense content delivered in a short amount of time, nor was it simply the effects of my lingering cold. Spending six hours listening to all the things that we should do in order to be heard in American corporate settings is simply exhausting.

Make sure to not look tired or frazzled. You probably want to wear some makeup, but not too much, you don’t want to seem to be overly made-up. Some heel is always nice, but nothing too sexy. Be sure to speak out for yourself, but try to not come across as too aggressive or bitchy. Maybe you want to defuse things with humor, but watch out for self-deprecating humor, as that might backfire. Have a body language that looks relaxed and dynamic, but not too “wired” or nervous. Try to use stronger, more powerful language, and avoid hedging or sound like you’re asking for permission. But make sure that what you choose feels authentic to you, too, or else folks will pick up on the inauthenticity. Show some warmth and passion, but not too much, you don’t want to seem like you are out of control or driven by your emotions. You want to seem strong, but also be human, not invulnerable.

Oof. Just writing the above paragraph was absolutely draining.

The corporate environment demands that we present ourselves as these perfect images of what our culture has deemed is “leader-like” — a shared ideal that we are all supposed to aspire to. These demands are placed on men, too, but women have it even harder, since much of the ideal is based on masculine traits, and there is still much sexism and plenty of double standards. It sucks.

At the same time, it seems hard to opt out, no matter that a lot of this is plain old bullshit. If you want to continue working, you’ll likely need to play by some of these rules. I’m grateful to be able to work in tech, where there is a little bit more flexibility on things like appearance, and where there are some (tiny, but growing) pockets where folks are trying to bring in more inclusive practices and less bias.

If you are as overwhelmed as I am by the list of dos and donts above (only a small portion of what was discussed in Allison’s workshop!), my suggestion is that you take it slowly, baby step by baby step. Think of these ideas as ways to tweak your effectiveness here and there, gradually and over time. Don’t try to adopt them all at once. Start with one thing that seems doable or that particularly resonates. Incorporate that into your daily work, see what results it gets you (or doesn’t — not everything will work for everyone), and then iterate. Our career will be several decades long, and you don’t have to get to perfect right this second. Think of this as a marathon.

In the meantime, I hope that we’ll continue building and nurturing work environments where different ways of being are equally valued. Where we all strive, despite our cultural expectations and biases, to see others fully. Where we can hear someone’s ideas and engage with them no matter what they’re wearing, the color of their skin or their gender.

Inability To Delegate As A Lack Of Trust And Self Awareness

Ana Ulin 😻 — Sat, 03 Nov 2018 17:11:33 +0000

The hardest job a startup CEO has is learning how to let go of stuff she used to need to be heavily involved in and trust that the team she hired knows what they’re doing.

— @ericajoy on Twitter, posted on 8:33 PM, 1 Nov 2018

A friend and former coworker texted me a link to Erica’s Tweet above with a comment about how this reminded him of a startup founder we both used to work for. I don’t know if trust and delegation is the hardest job of a startup CEO, but it sure is one of the hardest and most important things that leaders have to learn.

At first glance, the consequences of lack of delegation seem clear. On one hand, a leader that can’t delegate is not making space for her team to grow; this in turn means building a less capable team, where folks who are leveling up need to leave the team to find space to grow. On the other hand, a leader that can’t delegate is overburdening herself with work that could be adequately accomplished by others; as a consequence, she doesn’t have the bandwidth to do a good job of the really important things on her plate. So you end up with an overworked manager doing a mediocre job, with a team with low motivation and possibly high turn-over. Not exactly a recipe for excellence.

I think, though, that the real harm is done by the lack of trust implicit in an inability to delegate. When a leader can’t delegate, they are telegraphing: “I don’t think you can do a good enough job of this. I need to do it myself to make sure it’s done right”. Imagine someone saying this to your face. It’s unpleasant and hurtful. Perhaps it brings up some familiar self-doubt. Perhaps it makes you angry or causes you to shut down. Whatever your reaction, it is unlikely that your immediate response will be to get energized and motivated. And isn’t that what we want? A team where folks are taking ownership, where they are stepping up to solve problems and make things better?

I get it. Delegating can be scary. “What if they don’t do it right? I’ll be responsible for the bad outcome!”

But ask yourself: is this the main reason why I’m having trouble delegating? Might there be also a deeper reason, aside from this obvious one? Maybe it is a desire to feel needed, a fear of becoming less relevant because you’re “doing less”, or an unhealthy perfectionism that prevents you from seeing different ways of accomplishing a task as equality valid. This is where the hard work of increasing your self-awareness comes in. Work with a therapist, or a pastor, or a coach. Talk to others, journal, meditate or pray. Find the tools that work for you, and learn to look deeper inside yourself. Don’t be afraid of what you might find. We all have unflattering traits, things we want to improve or unlearn. What matters is what you do to grow past those areas that are holding you (and your team!) back. This is the true work of leadership.