Acknowledgements, &c.

Thanks to the many folks who helped with this along the way, including my wife, Brooke, John Capecelatro, Simon Goetz, and everyone else who Ive forgotten to mention here.

The fonts used in this book are NotoSans and Merriweather, both available through Google. NotoSans is available under an Apache License, version 2.0. Merriweather is available under the SIL Open Font License, version 1.1.

All content is (c) Kyle Bragger.

2 of 32

About Kyle & This Book

This tiny book is part one of an experiment Ive wanted to do for a long time now. Its written entirely in a Q&A format. Its questions were solicited from people across the Interwebs over the course of a few weeks, and range from questions about Forrst, a community of developers and designers I started in 2009, community building in general, product design, and more.

As for me, I'm a self-taught developer and product designer. Over the years, I've built, grown, and sold products related to communities, design, development, human behavior, and more. My most successful experiment so far has been Forrst, which I ran for three years before we were acquired in 2012. It now lives on at Zurb. I also started a job board for short-term development and design jobs called Tinyproj which I grew to 10,000 members. Currently, I'm working on Exposure at Elepath. I also made an iPhone app called Thinglist.

Enjoy!

Kyle

3 of 32

For communities based around content discovery, how do you personalize recommendations and extract the most relevant content without falling prey to the silo effect that leads to each user only seeing a narrow slice of the overall ecosystem?
Sacha Greif, @SachaGreif

To begin, Id identify what about the content being shared should matter. Is it the quality of the writing itself? The number of views, or likes, or shares? By enumerating what defines a quality post, you can then start to build systems to surface those. You can also surface content based on the viewing users interest graph, e.g. what kinds of tags theyve used over time. In my experience, the weight of each of these signals is something youll want to tune regularly as the expectations and scale of the community evolve.

Forrst has asymmetrical following (meaning User A can follow User B, but not vice versa), but we never really isolated content beyond a few basic filters: search terms, tags, and type of post (screenshot, code, etc.) In that sense, we didnt really leverage that following model to its full potential. Doing it over, I would have let users tailor their experience based on what they cared about Rubyists might only want to ever see Ruby posts from users who identify explicitly as developers, for instance.

Its possible to surface great content to users by a) understanding what matters about the content and weighting that higher, and b) letting users have explicit control over what they see. The best silo is one where a users seeing content they expected to see all along, plus content they didnt expect, but is pleasantly surprising and satisfying nonetheless. The ideal place to be is to have your content ranking algorithm that combines gut-feel signals based on your experiences/insight as the community steward with user-controllable levers, e.g. who they follow, what tags they use, and so forth.

4 of 32

Did you ever doubt a project youve (recently) shipped?

Fabe Schultz, @fschultz_

All the time! Its easy to doubt new ideas before theyve really had a chance to exist in the world. The important thing is to not let that deter you from actually sitting down and getting something up and running.

I had my doubts about Forrst, about Tinyproj, and about most other things Ive built over the years. A lot of what drives me to create a product is that I really want it to exist for myself. The irony here is that in that sense, its almost guaranteed to do well ... at least for me! But, I think self-doubt tends to happen once I start to consider things like will anyone else care about this? and is this good enough?

Since doing Forrst, Ive built a number of prototypes that were complete flops, but Ive also built a couple things that show promise outside of being useful to just me. I think as time has elapsed, Ive learned that whether or not I doubt somethings potential is actually irrelevant, so long as its user is elated to use it. (That user is usually me, at least for what Ive built in the recent past, but could be your friend, neighbor, dad, bus driver, etc.)

5 of 32

Do you think the invite-only model is still the best way to keep the content quality high (and to create buzz) on an early platform?
Miles Fitzgerald, @milesfitzgerald

In theory, if quality users invite their peers, then by association its likely their peers will also be of quality. As it turned out, this worked really well for about the first ten thousand users. After we hit a five digit user count, the sheer abundance of invites available* made it really hard to ensure that only people serious about contributing to the community were getting in.

So, yes: I think invite-only systems are an incredible way to drive growth, keep quality high, and to create buzz for early stage products. For Forrst, this invitation model was the reason we were able to grow to 50,000 users without ever spending more than about $500 on marketing in three years. But, at scale, while invites will indeed continue to help grow the size of the userbase, there tends to be a marked drop in the overall quality of the invitees. (Ive also found that Forrst invites tended to act as a form of currency, with bearers touting them for an ego boost.)

* Users got invites based on reputation, which was computed every hour or so by a robust algorithm we designed to surface users contributing to the community positively.

6 of 32

How did you incorporate design and development together to build Forrst, and when it was necessary to prioritize one above the other?
Aluisio Azevedo, @aluisio_azevedo

As someone who fancies themselves relatively competent in both design and engineering, it became quite an interesting challenge to walk the line between both roles. The engineering-minded side of me generally wants to get the product working as quickly as possible, since I believe theres nothing that can replace actually using the thing. But, the design-minded side of me typically wants to approach things design first; that is, actually applying a bit of finish to the product before really digging into the internals.

These days, its hard to disagree that design is more important now than it ever has been (and when I say design, I mean more than just aesthetic; Im talking about designing for emotion, for user experience, designing the business, how it will be marketed, talked about, etc.) So, over time it became increasingly more likely that wed end up doing a full set of early designs before writing even a single line of code. Often, wed design in the browser based on a robust style guide our lead designer, Keith, whipped up. This helped us make better decisions and iterate to an ideal version of a feature more quickly.

Ultimately, marrying design and development is a balancing act; theyre both vital to create a great product, but Id argue that careful consideration of design is something worth acting on before any substantial engineering is underway.

7 of 32

What technology (programming languages, frameworks, etc.) did you use for your MVP, and why?
Rasmus Kjr, @rkjaer

The very first version of Forrst was written in Ruby, using Rails 2 and Postgres. I deployed it to Heroku, which at the time (late 2009), was a young but promising platform. This setup worked well for the first few months of Forrsts existence, but I later rewrote the site in PHP 5 using CodeIgniter and migrating to MySQL for the datastore. While Id been using Ruby and Rails for some time, most of my experience actually deploying and scaling production sites was using a LAMP stack. So, I decided to do the rewrite while Forrst was still an extremely simple product.

For what its worth, these days Im back to Ruby, Rails, Postgres, and Heroku for my prototyping (and scaling) needs. The LAMP stack has served me well, but post-Forrst I needed a break from PHP, decided Id really missed writing Ruby, so I took the time to gain more competency in deploying and scaling Rails applications.

Ultimately though, the tools you use to prototype and build products are just that: tools. Building great software can be achieved with nearly any toolkit, so long as youre focused on that, and not just the latest, greatest, shiniest new thing out there.

8 of 32

What did you find the hardest thing about marketing Forrst? In particular, leading up to the launch of the community. Did you have any apprehension that it wouldn't succeed or did you have a fair idea of how you thought things would progress?
Paul Jaques, @pauljaques

The funny thing is that I actually had no intention of building a community; it happened completely by accident. When I first started working on Forrst, it was much more like Tumblr or Delicious for code and design. As Id later figure out, the seemingly simple combination of like-minded people sharing specific types of content, and being able to interact with that content and with each other was the catalyst for community.

We launched officially in May 2010, so by that time it was indeed clear that there was a strong community forming. Looking back on my time building Forrst, the thing that stands out most is how afraid I was of ruining the community. I definitely tried to temper this fear with a healthy dose of reality (and perhaps some bravado), but it was the one thing that was constantly in the back of my mind. Its incredibly hard to gain the trust of a community, but all too easy to lose it.

In the end, my worst fears thankfully never came true, but one thing I took away from our launch and subsequent growth is that fear can actually be a positive thing. It made me hyper-aware of each new feature we released, each change in policy we proposed, because the community was and is paramount. That fear helped to keep us in check; everything we did had to be for the good of the community, or else it wasnt happening.

9 of 32

How did you manage the demands of friends, family, and a full-time job while building and growing Forrst?
Mike Wales, @walesmd

First, it was important to understand what my priorities actually were, then to make my best effort to never waver from them. For me, I always tried to put my family and friends first; ideas and projects come and go, but its simply not worth losing the love and respect of friends and family. As anyone who does a startup will tell you, this isnt always the easiest of tasks. It certainly helps to have an understanding partner, and Im very fortunate to have one. I think, if you truly listen to the inner you, its fairly easy to know when youre approaching that dreaded point of man, I really havent been spending enough time with X.

During the first four months of Forrsts life, it was a side project for me while I worked full-time as lead developer of another small web property. This meant that I only had nights and weekends to work on Forrst. This actually worked out well, since it forced me to be really efficient about building features. As my time was very limited, I became hyper-focused when it was Forrst-building time; I tried to set small, realistic goals and to hit them without fail. This also served to get me in a solid mindset once I did start working on Forrst full-time, as Id already conditioned my brain to work this way.

10 of 32

What is your process for vetting ideas for whatever your next venture will be? Do you take them through some formulaic approach? For example, vetting the market, idea on paper, with friends, then building some MVP, look at the metrics, kill whats not working, etc. Or, do you just key into an idea and know a good one when it comes?
Andrew Baisley, @abaisley

Lately, Ive been writing a name I think works, plus a quick pitch or a few notes about an ideal instance of the product. Then I walk away from it for a while. That might be a few days, or a week, or a month. The excitement of a fresh idea often clouds my judgement about whether its worth pursuing, even as a prototype.

Ill quickly sketch the major screens of the app, just enough to help clarify structure. Once I feel like I understand the product well enough, I aim to get a working version running within a week or so.

With the product in an experimental phase, Ill bring in a few friends or colleagues, depending on who the target user is. I typically dont define metrics here its just too early. But, I am looking for people to get it, that is, to make some kind of connection with the product. You can tell when something just isnt resonating with people. From there, its just a matter of relentlessly iterating and polishing the product into something great and ready for (more) public consumption.

One thing I cant stress enough is that you absolutely must know who your user is. It could be you, it could be your neighbor, or your wife, or your team, or your plumber, but you need to know them by name. At this early stage, generic demographics do not a user make.

11 of 32

Once you had the idea for Forrst in your mind, how did you proceed? For instance, did you start by defining a clear user experience or did you begin on the data level?
Dan Sackett, @dansackett

My original vision for Forrst was to build a lightweight web app for cataloging useful bits of code, interesting design patterns, and so forth; anything that might be inspiring to me later as an engineer and product designer. I really wanted it as a user (which I think is super important), but nothing quite existed that worked the way I wanted it to. I briefly tried Tumblr, but thats a pretty general-purpose platform and ended up being too much effort to use.

To begin with, I started sketching main screens of the app (post list, new post, post page with comments) to flesh out the user experience, then moved to code pretty quickly. It was really important to have something real to play with, so I could validate whether the idea I was thinking about was even useful (fortunately, it was!) I knew the data model could change if it had to; that didnt have to be perfect right off the bat.

12 of 32

How did you solve the chicken & egg problem of communities? Did you have to artificially seed the community to begin with, or did it grow organically?
Zac Halbert, @zachalbert

The first users were all friends of mine. A lot of them were people who worked out of the office I was sharing at the time. The next batch were friends of theirs; my gut instinct was that theyd also be high quality folks. The first big batch of strangers, so to speak, was an influx of users from the Hacker News community, then from some early press we got; in total, Id say the first 3,000 or so users came from one of these batches, and this all transpired within the first four months or so.

The funny part is, I really didnt set out to build a community; thats something that I dont think can be forced. To me, the magic formula is really just providing a better place to share, discuss, or create things that matter to a group of like-minded people, and let things evolve organically. In Forrsts case, the original product (which was most decidedly not a community product) had a lot of utility, and had like-minded people using it, and from there the community grew and blossomed. Knowing what I know now, its very much a combination of things you can control a great product, bringing people together and things you simply cant, like human nature, and that magical spark that tends to happen when all of those pieces are in place.

13 of 32

Did you experience any sort of psychological hurdles (lack of selfconfidence, waning motivation, yada-yada), and if so, what helped you work through them?
Chase Oros, @chasetastic

I definitely did. Trying to grow a business isnt easy business, and I think communities are especially challenging to turn into something self-sustaining. For as many high points as there were running Forrst, there must have been at least an equal number of times where I was ready to throw in the towel.

In the very beginning, maybe two weeks into building a prototype of Forrst, I started having pangs of this isnt very interesting and actually stopped working on it for a little while. Now having built and launched a number of projects, and in talking with countless friends and colleagues who are also makers, I can say with confidence that thats a pretty normal feeling to have, and it doesnt go away over time. Its natural to have bouts of self-doubt doubt that what youre doing matters and I certainly did.

The other big killer for me was burnout: waking up and not having an ounce of motivation to get any meaningful work done, and wanting to literally just give up and walk away. The community backlash against trying to move Forrst closer to our goal of building the best place to help developers and designers advance their craft, and so on. It was pretty dark for a while there.

What helped me get through it, though, was a combination of a few things: my (now) wifes endless support, my friends and family, and some great investors & advisors. Having a solid therapist doesnt hurt either (especially living in New York City). Regular exercise (back then, running, and now, crossfit) is an absolute must. Seriously. It will literally change your life and get you seeing the big picture again.

14 of 32

What advice do would you offer to the recent influx of younger talent in the web design and developer communities?
Conor Haining, @conhaining

I think there are a few things to think about as someone young and wanting to break into the industry. First of all: this industry can be amazingly welcoming, but also equally horrible. It really helps to have a thick skin (easier said than done, especially as a teenager).

Its important to never give up, either. If you truly care about building things, about programming, about design, you have to be persistent. Youre going to make a lot of stuff that seems like it sucks. This is normal! It takes years (some, myself included, would argue that its more like a lifetime) of hard work, practice, and persistence to gain some of the useful skills youll need to ship great products. It doesnt happen overnight, and it certainly wont happen if you arent persistent.

Its also important to develop your own voice. Dont accept things the way they are just because thats the common wisdom or fad of the day. Evolve your thinking. Dont fall into tropes. Practice, practice, practice. Build on others work and make it your own. Remember that design isnt just what we see, its much more nuanced than that. Focus on solving problems in an honest and thoughtful way. Dont get caught up in making what this or that tech journalist says as canon. Form your own opinions.

Most of all, have fun. Youve got one life to live, so make the best of it.

15 of 32

Was there any point during the process of making version one of Forrst that you lost confidence in either your abilities, team, or in the idea itself? If this was the case, how were you able to rebound to move forward?
Harry Lewis, @imharrylewis

About two weeks in, I started having pangs of this actually isnt very interesting I think I took a couple days off of working on Forrst and started thinking about throwing in the towel. Of course, just two weeks and an unlaunched prototype dont exactly provide enough data to know whether it was in fact worth building (or not). Fortunately, taking a breather helped me regain the focus I needed to wrap up the prototype over the next week or two and start inviting some friends in. Ive learned, though, that this self-doubt is actually quite a common reaction to have, whether two weeks, two months, or even two years in to building something.

There are a myriad of external forces working to chip away at your confidence. Maybe its a negative review about your product, maybe its an especially angry user, or unexpected downtime over the weekend, or exciting new feature that flops. Whatever the case, there are plenty of times when the fun seems to wear off and things feel like a slog. The thing I simply cant stress enough is this: thats totally normal. Stuff happens. You cant please everybody. (I wish Id known this sooner!) Its really important to keep sight of the big picture: why are you building what youre building? What problem are you solving, and for whom? Are they happy? Are you happy? Thats all that really matters in the end.

Focus on building honest, thoughtful products that make your users day a little better (and that user could be you, too!) and the rest tends to fall in to place. Once I realized that, everything changed.

16 of 32

What is your definition of design?

Steven Chung

I think the thing that comes to mind for most people when they hear the word design is aesthetic. This isnt an entirely incorrect definition, though; to the contrary, one aspect of design is indeed how something looks. But to me, design is much, much more than that.

When I think about designing a product, Im thinking about everything from how Ill talk about it, what its called, how it works, how a user will feel using it, how Ill market it, the business behind it, and yes, how it looks. To me, every piece of a product should be well-designed. Aesthetics are important, but should be informed by how the product works, not the other way around.

Well-designed products should be thoughtfully designed in all of these areas. To focus on aesthetics first, or worse - to the exclusion of everything else will only cripple your ability to ship honest software.

17 of 32

How do you deal with criticism, specifically from well-known, or talented/respected peers?
Tyler van der Hoeven, @Jeepsing

Having a thick skin and being able to take criticism (and act upon it in a healthy way) is something that I think requires a lot of practice. I specifically say practice, because its something that takes experience, time, and repetition to do well.

Building a product whose membership grows beyond a small group of friends usually means youre going to have a lot of feedback coming in, both good and bad. When Forrst started to hit that point, I was initially really overwhelmed. A lot of the feedback was positive, but, as youd expect, it also got a fair amount of criticism; everything from this is just a ripoff of X to folks thinking Forrst was pointless or poorly designed, or that the community was somehow inferior. (Even the occasional piece of hate mail.)

The fortunate thing about hearing criticism from well-known industry folks is that 95% of the time its actually really thoughtful, constructive, and usually pretty easy to take away something actionable from it. But, I think thats in no small part due to people of that caliber simply having a lot of experience taking criticism from others. Ill take constructive feedback from folks time and time again over praise, though, because ultimately thats what will help me build a better product and grow a healthier community.

18 of 32

Any good strategies on how to avoid spam accounts and content?

Magnus Hjelm, @magnushjelm

We implemented a few different measures at Forrst, and in fact, Id say 30-40% of our feature development revolved around building tools to make our moderation efforts more effective.

The first (and most successful method) was simply keeping the site invitationonly. This meant that wed be able to avoid automated, mass signups. We did see a handful of manually created spam accounts from time to time, but our community had a sharp eye and usually reported these to us fairly quickly. (Side note: make it really easy to get in touch with your staff!)

We also implemented a bozo box feature, which let us selectively hide users and posts from the community without the offending user realizing what was going on. That way, they continued to do whatever it was they were doing anyway, but the rest of the community didnt ever see anything. Wed investigate any posts or users flagged more than a certain amount (and were notified automatically when we needed to take action.)

We also prevented users from abusing the ability to flag content, turning off email notifications if a user mass-followed tons of other users (follow spam), and automatically flag content containing any kind of commonly used spam content or tactics, like BBCode or tons of URLs.

Above all, we made it easy for any member of the community to report stuff that looked suspicious, and we made sure we had effective processes in place for dealing with it.

19 of 32

Do you think theres room for more communities like Metafilter that people have to pay a small fee to join? Had you thought of charging for Forrst?
Jed Sundwall, @jedsundwall

Im pretty fascinated by the idea, actually. I think Forrst became something quite special because it was invitation-only, which in essence is just an artificial limiter on growth. Charging for access is another thing thatd probably achieve the same effect (and in fact, we actually considered this multiple times.)

I think the idea that you have to pony up to participate also acts as a natural filter for keeping people who arent very serious it out. I often wonder if such communities tend to have overall higher engagement rates (even if fewer users), simply because a small fee is a good litmus test for interest.

On the other hand, if a community like Reddit had charged from the beginning, would it have been successful to nearly the same degree that it has been? What about Twitter? I think the takeaway here is that it all depends on your goals for the community, and that its an interesting enough idea to give serious consideration.

20 of 32

Can you talk about a time you removed or disabled a feature, knowing some or most of the community would be disappointed? How did you approach this and what was the outcome?
laura gluhanich, @lauraglu

Definitely. The most notable is probably a thing called Discussions. It was basically a real-time chatroom for members. The version we initially released didnt have any way to create rooms, or do anything except literally send messages to anyone else who was in the room with you. Looking back, its clear the feature just doesnt feel like it belongs with the rest of the product. We killed it with little fanfare after only about a week or so. I didnt feel it warranted an email blast, since so few users actually cared about the feature to begin with. Its death was fairly painless and well-received.

There were a few lessons from this. One was to stay focused! It sounds easy, but its just as easy to fall victim to the wouldnt this be cool? game. Discussions wasnt something that users were actually asking for to a large degree. It was a fine feature from an implementation perspective, but it didnt solve a problem. In that sense, its no surprise that it was a flop! We built something that was more or less based on conjecture. Maybe people will use this if we build it, we thought. Thats something that, unfortunately, tends to not work as expected a lot of the time. Its important to build with a real user or users in mind, not just generic demographics.

Dont be afraid to kill stuff. Take it slow, work on a single thing at a time, and use a combination of gut instinct and data to inform your decisions. Sometimes, though, things just dont work as planned, and thats perfectly okay. Its easy to make excuses to leave things in, like this took so long to build or maybe people will use this if..., but ultimately, the product will only suffer if you are afraid to be ruthless about pruning stuff that just isnt working.

21 of 32

How did it feel as the smaller service in comparison to Dribbble? Did it affect your design, development, or business decisions?
Joshua Sortino, @joshuasortino

I have huge respect for what Dan, Rich, and their team have built. This question is (unsurprisingly) one that has come up a lot in the past. I cant stress enough how important it is to know these two things: who your user is, and what youre building them. Sadly, a lot of products exist whose creators dont have an answer for one or both of these points. (Ive been guilty of this, too.)

I think its a poor use of time to be reactionary to other companies you may view as competitive. Similarities often exist between products in a market, especially a niche one; thats a given. But, if each feature you build is framed within the context of whos the user and what does this help them achieve?, you maximize that features potential to successfully elevate its users. But, when you start to let competing products inform your decisions, I think its really hard to answer this question with integrity. Youre no longer building with purpose, youre playing catch-up. Ultimately, the product and your users with suffer for it.

22 of 32

How did you decide on the minimum set of features to ship with Forrst? When did it feel right? Did you plan it or go by gut?
John McDowall, @MrMcDowall

I really just went by gut instinct. I think listening to your gut instincts is one of the most under-valued tools we have as product builders. Theres a lot of talk about making data-driven decisions, but I think a better term for this (that I definitely didnt come up with, but just really like) is making data-informed decisions instead. To me, that means combining gut feel with metrics, and making a well-informed decision based on those factors. Generally, I think a lot of great products are built by people acting on a hunch, while keeping in mind who theyre building for, and integrating meaningful data to help focus what theyre building.

Now, for a side project in its infancy, of course there wont actually be much data to go on, so in that sense, things are indeed much more gut-driven. For Forrsts MVP, I started simply by stating what I wanted such a product to let me do to catalog interesting code and design and started working from there. The resulting MVP could hardly be called a complicated product, but each feature it had was solid, complete, and gave me a good starting point to observe how I and other early users actually used it.

23 of 32

What is the best feedback a designer can receive?

Gyan Prayaga, @gprq

Have you considered ... ?

As product makers, quality, thoughtful, constructive feedback is one of the most invaluable things someone else can give us. In order to truly move our craft forward, we have to be open to receiving this feedback. That means putting ego aside and being open to criticism.

P.S. I think this answer truly applies to everyone who builds products, not just designers.

24 of 32

What is your background? And have you launched anything else than Forrst, successfully or not?
Jonas Kelstrup, @JKelstrup

Im a self-taught engineer who eventually found his way into product design, largely out of necessity. Around age twelve I found a copy of Visual Basic 3 and some basic programming tutorials around the web. Ive always been obsessed with building things whether Lego or model airplanes or a miniature trebuchet (yes, this really happened). I tend to geek out over the inner workings of something.

Later on, I discovered HTML and CSS, and from there, PHP. The idea that I could build an application that anyone in the world could use was astounding to me. Amazingly, people were even willing to pay me to build apps for them. I landed a front-end engineering position at a company in New York City. After that, I did some consulting, tried building a few of my own (terrible) ideas, and eventually started Forrst as a side project while at a company called Corkd.

Ive launched lots of experiments over the years, but Forrst is definitely the most successful thing that I built (so far). Second is probably either a thing called Tinyproj, which was a mailing list of short-term dev and design jobs, or Thinglist, which is an iPhone app that makes it dead simple to keep a want todo list. I ended up selling Tinyproj to a company called GroupTalent shortly before Forrst was acquired in 2012. As for things that sucked, well, thereve been dozens of those. But, as long as I can learn something from each one, its hard for me to consider it a failure.

25 of 32

As your community grows in size, how do you prevent a general slide towards mediocrity?
Luke Chamberlin, @lchamberlin

Something that didnt click until much later was that folks doing great work often just dont have the time to devote to participating in a community to the degree I would have hoped. Our goal at Forrst was to provide developers and designs of all levels with a place to get (and give) thoughtful, constructive feedback. The idea of writing paragraphs upon paragraphs of critique doesnt exactly thrill most folks who already have a full plate.

A lot of the engineering we did was building systems behind the scenes to elevate users who did take the time to provide such feedback, even if only sparingly. As Forrst grew, we refined these systems regularly each new influx of users meant new behaviors and therefore required us to constantly be thinking about what defined a great Forrst user.

For a niche community, Forrsts scale was certainly interesting once we crossed into the multiple thousands of users range. Each new user has no concept of what the community is, stands for, strives for, etc.; that has to be assimilated through participation, as well as how we (the staff) position features, through the copy we write, and the guidelines we put in place. While I think mediocrity is inevitable, by elevating the users and content which best represent the ethos of the community, it can thrive regardless.

26 of 32

How did you find Pasquale DSilva?

Hariawan Bangun Safutra, @haribangsa

Believe it or not, on Forrst! He was user #221, and joined when the site was only a few months old. Pasquale was incredibly generous with his insights and feedback into what Forrst could be, and we became fast friends. He eventually went on holiday to NYC, and we made a point to meet up. I was such a huge fan of his work that I hired him to help develop Forrsts identity, which ended up including a new logomark and some awesome illustrated characters for our homepage and error pages. Fast forward a few years, and were still great friends, and also have the pleasure of working together at a venture called Elepath.

Pasquale is only one of many, many great folks I met through Forrst. I actually ended up hiring our community manager, Mike (user #27), and our lead developer Zack (user #2,304) from there, not to mention countless other amazing makers with whom Ive developed friendships.

Community can be a powerful thing.

27 of 32

How did you get the first 1,000 people on the platform? Word of mouth? Marketing?
Cesare Rocchi, @_funkyboy

There are a few phases I went through to get the first thousand users. The first 50 or so were friends and colleagues of mine, who I invited when Forrst was just in its infancy. (Id built an invitation system into the site so that I could limit access while the product was still maturing.)

The next batch of users, perhaps another hundred or two, were invited by the existing userbase. My thinking was that liked-minded folks I respect are most likely going to associate with like-minded people they respect, and so on. So, I gave them all a handful of invites to send out, which many did. As expected, it worked, and Forrst ended up with a few hundred awesome people using it regularly.

From there, as I recall, the next phase was doing a Show HN on Hacker News, which ended up growing the waiting list to around a thousand people, who I invited in. My thinking here was that most people reading HN are the types of people Id want using Forrst.

Something worth mentioning is that early on, the email youd get when someone invited you was actually pretty long. It was kind of an open letter from me to the recipient, talking about why I wanted Forrst to exist, what they could expect from the product so far, and more importantly, what was expected of them. Id later learn that setting this ethos early was something that ended up being vital to Forrsts growth as a community.

28 of 32

Did you have any personal or professional life mottos along the way that gave you hope?
Christopher Kennedy, @kennedysgarage

A big theme in running Forrst was to do everything for the benefit of the community. Being mindful of that, and framing every decision in that context made things much simpler; whether it was a new feature, or pricing or policy decision, Id ask myself is this healthy for the community? The majority of the time, it was pretty easy to answer with either a yes or no. Things that I couldnt answer yes to I generally didnt move forward with. Thats not to say that every single decision was or should be a binary one, but leading with that question helped filter features and decisions that would likely be a waste of time to pursue.

I also tried to remember to have a thick skin. This isnt always easy, but its important to realize that in any sufficiently large group of people, youll have folks who love you, hate you, or maybe just dont care. Its the ones who are intent on conveying how much they loathe the community (they somehow are still using regularly; go figure...) and hate you for existing, etc. But, heres the upside: its nothing personal! Not everyone is going to love what youve created. Some people are simply unhappy, and look for any chance they can to take that out on others. Just focus on the members of the community who are being positively impacted by what youre building, and ignore the naysayers. Not to mention, Id rather be building something that people loved or hated, vs. something they simply didnt care about one way or the other.

29 of 32

How many of your users were you personally in touch with?

Ben Lang, @benln

I tried to be very accessible to users. This resulted in me being inextricably bound to Forrst (not to mention some sweet fan art being created.) Early on, the invitation emails were actually pretty lengthy; I talked about my vision for Forrst, what I expected of users, and what wasnt quite working yet. As we grew, it naturally became more challenging to be able to personally respond to every single support and feedback email, but I did my best. We ended up implementing Uservoice to collect user feedback, which unfortunately got to the point where a lot of items where simply never responded to; in retrospect, I probably needed to hire a second support person in addition to Mike, our community manager.

From time to time, Id also send email updates to all users about new features, policy, and the like. I also blogged with some regularity about the same, though as we grew it also became harder to keep up with. When we released Forrst v3 in late 2010, we built a beta group of a few hundred users and they were simply amazing at finding bugs, giving great feedback, and generally helping us to refine the new product we were about to release. Similarly, I found that it was almost impossible to personally respond to each. This is something that really pained me, as Forrst was the first time in my professional life Id ever had to even consider ignoring someones note to me. But, I made sure to read each one, even if I couldnt respond.

Having an open line of communication with the community is super important; simply participating isnt enough; there are sure to be issues folks just dont feel comfortable discussing publicly. Be open to ideas, brilliant, insane, and everything in between.

30 of 32

What are some of the things you did to engage your users?
Teng Siong Ong, @siong1987

The earliest thing we did was run the Forrst App Contest. We encouraged members to team up and build something cool together. We gave folks a week to get something online, and after a week we chose a winner. We sent them some Forrst swag (one of our Survival Kits, which included a coffee mug, shirt, stickers, and wristband.) Just a handful of people participated, but it was really cool to see camaraderie between members, even so early on in the communitys life. (I still get notes from time to time from old users telling me they still use the mug daily.)

As far as site features, we rolled out a pretty standard notifications center; itd tell you about new content you probably would be interested in, such as comments on your posts (or posts you were following), new followers, and so on. We saw a permanent spike in daily engagement on the order of about 1.52x the day we launched.

Email was the other medium with which we tried to engage our users. Wed send out periodic updates about what was going on with the community, new features, featured content, and so on. Our open rate over the three years I ran the site was something like fifty or sixty percent, with click-thrus around 25%. I tried to only send newsletters sparingly, and for transactional emails, made it dead simple to opt-out on the spot, no login required. Its easy to overdo it with email, but I think that keeping things tasteful almost certainly helped keep users coming back.

31 of 32

Thank you.
Thanks so much for your purchase. I sincerely hope you came away from this with some new insight into building community products. If you have any feedback, questions, complaints, or otherwise, Im on Twitter at @kylebragger, and reachable via email at kyle@kylebragger.com.

Cheers, Kyle

32 of 32