Reach more users

[ErikBjare]

ActivityWatch is getting pretty good (at least that's what our current users think) and it's getting to the point where the impact/utility it has is limited by the number of people using it.

It would be nice to get it into the hands of more people, not the least because more users mean more people potentially donating. We have metrics on users/downloads/pageviews (see #83) as well as survey data from current users (see #291), so we're well prepared to evaluate which efforts give results.

Questions:

• How do we reach more people going forward?
• Why do some people who find ActivityWatch don't start using it?
  • Likely related to most requested features yet to be implemented?
  • Is installing ActivityWatch too complicated? (currently the case on macOS, see macOS Catalina complains repeatedly about an unverified developer #334)
  • Are we not designing for their use case?
• How do people find us now?
  • I have data on this from website analytics (a lot of people come from AlternativeTo or through Google searches)
• How have efforts to reach users in the past performed?
  • Toggl has a blogpost on how putting effort into social media presence has limited impact (can't find it now though)

We've done some work on this in the past (#38), this issue is intended to track the general continuation of that work.

Accepting suggestions for how to reach out to a wider audience 🙂

[xylix]

Probably somehow ensuring our solutions are actually found when people search for a time tracker would also be a step towards making possible users find us.

For example, currently (11.03.2020) if I search for "time tracker" on google play there's hundreds of apps but can't really find activitywatch anywhere.

Actually even searching on desktop for "time tracker" activitywatch can't be found very easily in the results. We could maybe try some low effort SEO at least, to make sure common search phrases for people who might be looking for activitywatch actually find it.

[snovak]

LOVE the project by the way. Just pitching in some ideas here.

Influencers! Reach out to a few "self help" and/or "high-performance" and/or "startup" gurus, like Tim Ferriss. If you can get just one to plug you, you've just reached a TON of people in a meaningful way.

Videos! "How do you use Activity Watch in your routines?" For example, I work remotely for a big company, ~90K people, and many others are remote too (especially now with coronavirus). We have to log our hours every day (at end of the week), and split up our time into categories Until I found AW, IT WAS SO TEDIOUS!!!

Integrations. Oh, I don't know. Let's just say there was an integration to something like GitHub, Slack, or Basecamp, or some other productivity tool. Like, for Github, maybe it would be useful to highlight tasks that actively contributed to a bug fix or new feature and it would spit out a string, or something, that you could drop in the commit. Uhhh, I think that's a feature request.. I think I'm getting off track. But, what I think I'm saying is, if AW works with some of these niche tools, you'll very quickly gain many of their users too. Or, like I mentioned above, crappy time logging systems NEED this. But, can't be mandatory. And they can't control the data because then it's creepy, which is why open source is great solution!

Hope something here helps!

[xylix]

Hm yeah great point about the influencers. Personally I follow a few time management / productivity persons on YouTube (for example Matt D'avella and Nathaniel Drew) and asking for a change to pitch activitywatch on some such people's podcast could be a great way to get through to more people interested in tracking time etc.

Thanks for the input :). It looks pretty useful.

[alexriabtsev]

Hey, team!
Kudos for great app! I used to use 0.8.4 and updated today to the latest. The difference is huge!

Regarding promo - there are three main ways: people (influencers + referral), ads/money (Google + FB networks), content (blog/email/telegram etc.).

BTW, you can have Slack for community (it used to be great Trello community in Slack).

Feel free to ping if you need further thoughts!

[samuelsimoes]

Hi, I love your effort to do an app like this open-source, thank you for all the work.

I think if you work more in your "call to action" in your home page with a highlighted one-line command to install it through package managers like apt and macOS brew would help you a lot to get some more users because today you need to download a zip (from a small button, btw), extract it and read the "get started" to understand how to start the app, which hurt your "call to action" IMHO by my experience.

Nonetheless, installing through repositories will probably help you to not lose users that may face issues with missing dependencies.

[johan-bjareholt]

I think if you work more in your "call to action" in your home page with a highlighted one-line command to install it through package managers like apt and macOS brew

@samuelsimoes We don't want to recommend installing from the command-line by default, all users are not comfortable with that. We do however already have support for brew and we are working on building a .deb
#10

I do however agree that the installation process is not obvious. Personally I especially want it to be easier to make aw-qt autostart on macOS and Linux (currently only easy on Windows).

[fmiqbal]

I have listed AW on ArchWiki List of Applications, if its fine by you, https://wiki.archlinux.org/index.php/list_of_applications#Time_trackers

[Oymate]

I found this app through itsfoss.com, so maybe some articles from sites?

[ErikBjare]

@Oymate We've tried getting people like itsfoss.com to write about us, as you can read about in #38 🙂

[jackpoc]

I found you searching "open source activity tracker" for which you're ranking in p1- congrats 🙂

I happen to work in SEO and agree with above that adding content to your site would help. So that would mean writing content to match certain keywords people search for. Maybe specific to activity trackers, specific features, or targeting users with broader content relating to productivity.

[dumprop]

Add iOS platform

[ErikBjare]

@dumprop Unfortunately that's practically impossible due to limitations in iOS and App Store policies. See this tweet I wrote the other day: https://twitter.com/ErikBjare/status/1360667982775599107

[dumprop]

@dumprop Unfortunately that's practically impossible due to limitations in iOS and App Store policies. See this tweet I wrote the other day: https://twitter.com/ErikBjare/status/1360667982775599107

But AFAIK rescuetime can get total use of device and unlocking count (unfortunately without individual apps), so it better than nothing
I hope apple will review that (after court?)
Also it can distributed without app store (not preferred way, but rescuetime have app there)

[jackpoc]

Following up on SEO again. I think a good direction for you is to position yourself as the free, privacy-focused alternative to RescueTime.

One place to start is by creating a page called "Free RescueTime Alternative". People do search for "rescuetime alternative" for example.

Or you could do a write up like the one here and put yourselves at the top:

https://www.timedoctor.com/blog/rescuetime-alternatives/

They're ranking well for this and seem to be paying off for them.

Not sure you'd be able to outrank them for the target term (rescuetime alternative) but you'd easily get on page 1 of Google for it.

In my opinion you should still be positioning ActivityWatch in this way and even looking at developing similar features to RescueTime relating to increasing productivity and getting useful insights from the data.

[CrazyPython]

Hi,

I've gotten 20 different links to the frontpage of Hacker News, sometimes multiple from one organization, including 3 of my own. I can help get you to the frontpage.

Assuming your build works on macOS and Windows, if you don't object, I can submit for you.

Alternatively, to maximize exposure, make a version of your landing page https://activitywatch.net/ with the Screenshots front and center, above the features. Hacker News likes specifics, and the screenshot is the best way to explain what it does to those not familiar with automatic time-tracking software like RescueTime and Timing.

If you don't want to build that version of your site, we can take a chance and I'll submit to HN with a specially crafted title anyway.

[pcuci]

Here's a bad idea, as I see two large types of users:

• the folks who like buying pretty things that "just work", the key thing there being thing
• the DIY crowd
• maybe the productivity crowd, but if you're already actively searching for that content, you must've picked some low-tech hacks along the way and successfully implemented them, at which point AW would only be of marginal benefit

I came across this post recently, showing how to Build a Tiny Certificate Authority For Your Homelab

Imagine running aw-server-rust on that thing - we'd get security for free. Perhaps an alternative to: #35?

Sure, to pull off something like this would require a crowdfunding campaign like the Solo V2, for example.

I guess the general idea is to "productize" something that can be sold in real physical form.

[pcuci]

@ErikBjare - re:

Why do some people who find ActivityWatch don't start using it?
- Are we not designing for their use case?

I noticed that a couple months after getting my Fitbit (yey, surveillance capitalism!) I started to pay less attention to it, because I sort of grew a "6th sense" of what it's going to tell me. Doesn't mean I don't use (I wear it every day), it's just that I've calibrated myself to "just know", and I now can tell what a 150 bpm "cardio" heart rate feels like within 10 points accuracy, so I look at it much much less. At the end of the day, that's all I really needed to get my ass moving during covid. Might be a satisficers vs. optimizers thing - who really is AW targeting?

I think the key will be in making ActivityWatch become a wholistic passive tracker that you need not consult, because our awareness of how we spend our time in front of our devices has increased, so much so that we instinctively close off timewasting apps. Put another way, AW will have helped us build healthy internal triggers so we can snap out of tech misuse more easily and at will. I suppose I tend to agree, more than most, with Nir Eyal's Here’s My Review of The Social Dilemma: No, Social Media Is Not “Hijacking” Your Brain. Maybe that's AW's purpose: a transient crutch towards quantified self awareness. :-)

I don't know what the use case for newbies is, it might be that they see the initial onslaught of bad behavior and turn it off?

There's some evidence showing that missing a positive streak is demotivating for people who want to lose weight, so much so, that the entire experiment backfires and they put on more weight than when they started. It's possible ActivityWatch has some similar negative side effects (that'd be hard for us to detect without diligent usability tests). Nonetheless, how might we design with that in mind?

[ErikBjare]

I've gotten 20 different links to the frontpage of Hacker News, sometimes multiple from one organization, including 3 of my own. I can help get you to the frontpage.

Assuming your build works on macOS and Windows, if you don't object, I can submit for you.

@CrazyPython That's really nice of you! I posted ActivityWatch once but it didn't get any traction. Builds on Windows should be rock-solid, macOS builds have some issues but people on HN should be able to overcome them.

Alternatively, to maximize exposure, make a version of your landing page activitywatch.net with the Screenshots front and center, above the features. Hacker News likes specifics, and the screenshot is the best way to explain what it does to those not familiar with automatic time-tracking software like RescueTime and Timing.

Agreed! I've been meaning to do this for a while (the screenshots were too far down on mobile). I moved them to the top on small screens in ActivityWatch/activitywatch.github.io@fd785e7.

In summary: Go right ahead!

[ErikBjare]

Put another way, AW will have helped us build healthy internal triggers so we can snap out of tech misuse more easily and at will.

@pcuci I read this paper (thanks @jwiese) a while back where they built an app called TimeAware and compared two interventions:

1. You get positively reinforced when you're productive
2. You get negatively reinforced when you're distracted/wasting time

The result of the study was, perhaps surprisingly, that (2) was significantly better at nudging the user towards being more productive (this factors into the design of ActivityWatch/aw-webui#258, and other plans for future features).

Imagine running aw-server-rust on that thing - we'd get security for free. Perhaps an alternative to: #35?

Yeah that'd be cool, but I'm really not keen on making ActivityWatch something people expose to the network due to all the security implications (as mentioned in the new page in the docs: https://docs.activitywatch.net/en/latest/remote-server.html)

I guess the general idea is to "productize" something that can be sold in real physical form.

@jwiese talked about this too! He suggested selling a Raspberry Pi with ActivityWatch bundled, working as a remote server. I think it's an interesting idea, but I'm not convinced enough to put in the work required.

[CrazyPython]

Agreed! I've been meaning to do this for a while (the screenshots were too far down on mobile). I moved them to the top on small screens in ActivityWatch/activitywatch.github.io@fd785e7.

Good work. 50% of Hacker News readers are on mobile. Screenshots are still too small for me on desktop though:

You don't have to modify your main landing page if you don't want to, a custom /hn.html just for Hacker News is fine.

[CrazyPython]

Actually, I think the GitHub Readme is a better place for the Hacker News- See my pull request #576

[2br-2b]

I feel like having some built-in categories would be nice. That way, new users don't have to start building their categories from scratch when installing the app. Since there are many ways to customize it, maybe make a few different default categorizations that the user can choose between?

[ErikBjare]

@2br-2b Improving the defaults has been on the list for a while (see ActivityWatch/aw-webui#174). We've recently merged category export/import to easier enable people to share their categories which is a good first step in that direction.

Once we've made a release that includes the feature we'd gladly accept people's categories and link to some good community presets in the docs!

[ermalbaj]

Hi Erik,
When will be the next release? Any beta soon?

[ErikBjare]

@ermalbaj Working on it right now.

If all goes well, we'll have a beta release ready today or next week.

[Albonycal]

Nice.. I've been using this for a day now & it's awesome...
I was looking for FOSS time trackers for so long but couldn't find a good one
BrodieRobertson did a YouTube video on activitywatch & I found this..
Thanks

[ShayanYaseen]

Hi
The site omgubuntu is an established site frequented by a lot of Linux and tech users. It also mainly focuses on open projects and gives them much needed boost in users and SEO as Google regularly indexes this site in it's search results.

A single writer runs the site, so I implore you to send him an email or a tip at https://www.omgubuntu.co.uk/tip

[jfunction]

Adding my 2c - I found this by googling "screen time app ubuntu" and clicking the itsfoss article. I was intrigued and really liked what I saw but a little disappointed to see there's no snap for it nor any deb repo providing it. I am a dev, so I can figure out how to install it with a little extra investment on my part, but for less technical users I imagine they get stuck here.

I'm sure there are technical reasons that there is no snap and honestly I haven't bothered to learn about all that, I just wanted to relay my initial impressions.

[TarryDan]

I found it being recommended by a challenge on Habitica, & chose it of the available options due to the privacy focus.

The barrier to entry is installation. I'm using Lubuntu, and am an inexperienced but confident user (happy to type out whatever I'm told to in the terminal, but probably won't fully understand what the commands do). And I can't make head or tails of how to get it running.

It needs a one-click installer; and as a stopgap, an installation guide that assumes this is the very first time the user has ever done any of these steps.

[maikol-solis]

I found this software because RescueTime is dropping Linux support with their new interface (RTX).

Maybe this is an excellent opportunity to bring new user to this project.

Amazing work though!

[kazerniel]

As a new user - who's not a programmer - my main barriers to adoption are AW's bugginess and that the GUI (especially the timeline and settings pages) seem very barebones.

Reflecting on previous comments:

• I didn't have any problems with the installation process on Win10.
• I found the app via AlternativeTo, looking for an alternative to RescueTime. (I used that 5-10 years ago, but now that I want to start some kind of tracking again, I don't feel comfortable with a private company owning that much of my data.)

[brendantay]

I accidentally stumbled across this app when on alternativeTo looking at something completely different. As a IT Guy with ADHD that runs his own businesses (from home) I get easily sidetracked and go off and do other things that aren't relevant to work but that's just just a trait of ADHD... Unfortunately I often end the day asking myself "what did I do for the last X hours?"

I'm yet to test ActivityWatch, but today I am going to set it up on a few devices and I'll check back in next weekend to let you know if /what benefit it has been to me!

[isbur]

Sync out of the box is a must (including mobile app)! Don't think about reaching out people without this feature. Maybe temporarily drop the idea of fully distributed architecture and/or just allow later aw-servers to choose one of them as main depending on online topology.

Ok, maybe it's just me, however, I'm still sure this is a high demand feature.

[thetwo222]

Hello, have you considered supporting more languages? I want to translate it into Chinese

[SpongebobSquamirez]

If this let me import from rescuetime I'd switch in a heartbeat...sigh

[jifalops]

FWIW I have a lot of experience with Flutter and could help open up an avenue to iOS and native UI across platforms.

[jifalops]

After exploring the state of things a bit more I have some thoughts:

iOS activity can be watched using Device Activity APIs. There's also a Screen Time API but it seems more geared towards parental control. Users complain the activity API has problems but there are existing apps for iOS so I'd expect it to work but be sort of painful to develop.

--

If we went the Flutter route I would rewrite port the Android app in Flutter before expanding it to iOS. There are a couple of plugins for monitoring app usage (1, 2), but they only work on Android so I (or someone) would be on the hook for submitting PRs to support iOS.

Maintaining a Dart client wouldn't be necessary because the Rust client can be automatically translated using the Flutter Rust Bridge. It'd probably be a good idea to have a public package for activity_watch though to encourage adoption. Maintenance would be scriptable for the most part if not completely. That raises the question of including the server in such a package, allowing AW functionality without the AW app.

The Flutter route could consolidate desktop UI and browser extension efforts as well, one day. The general consensus is that it works well for apps, but not for websites (slow first paint and feels different). Web apps are okay though.

--

For Linux on Chromebooks I made ActivityWatch/aw-qt#84.

Finally I think a good manual 'watcher' would help reach more people. At least my idea for a hobby project revolves around easy manual entry of things you did recently. I'm considering writing one in Rust, although heartbeats don't really make sense for that.

[nirurin]

My main feedback, and the reason why I may not end up being able to use activitywatch, is I need a time tracker in order to track and populate invoices.
I currently use timecamp, which automatically (I don't need to activate it or anything) will see when I'm working on projects by using tags in window titles, start logging time, log that time -specifically- to the correct project. Which means that I could spend a week working on 10 different projects, without even looking at or touching timecamp, and at the end of the week I can go to the reports screen and it will have correctly logged all time I spent on each individual project, all correctly assigned so each client can be billed individually.

This took about an hour to learn how to do and completely set up.

On Activitywatch.... I spent a couple hours just reading documentation to figure out how to get it to put stuff in buckets. Never bothered actually doing it though cos half of it didn't make sense, and I still had no idea how to actually -do- anything with these buckets of data. From what I can tell, it would just be a list of all activity recorded, which is useless.

I believe I need to create a series of scripts, which will then find the right data and.... export it to json (which is useless for 99% of uses, especially for laymen). And I'd have to manually make a script for each new project.

The activity watch setup and dashboard is good, great even. It's good for people who just want to make sure they aren't wasting too much time, and you can go "ooh look I spent 2 hours on youtube, better do less of that in future". It's a nice procrastination prevention gimmick. But for actual production purposes (which is the important step tbh, as thats what 99% of people will grab something for long-term usage) it just doesn't.... work.