Criticism is good thing. It even helps me to see my product clearly)
So let's see
1. Taku is not just a chat popup window. You don't talk with only one user at a time. You have constant connection with all users.
2,3. Notifications won't be always on screen. Only when there is a new message. And yeah I'm thinking about enabling "Mute" for users who don't want to be annoyed)
I think it's a cool idea, and could be useful. In the best case it would be like a crowdsourced customer support, where the users will help each other. And a way to make the product more engaging.
Biggest doubt for me is that, I think this would require a lot of moderation, since surely there will be people advertising something, or recommending competitors. Or people impersonating as customer support agents and scamming your users
Yeah I think this is main problem with products where all people can post. But I think you would be able to limit who can comment if you receive a lot of comments. Because you'll collect data on users through polls and forms
I think this is a sensible idea. As a user I feel friction when needing to get in touch with people working on a software that I use.
Users don't necessarily wanna join or know how to join a discord server/forum just cuz they're using a particular software/product and need help from time to time.
An alternative to Taku would be forums hosted by the official group/co. in-charged of the product. These are a hit or miss. They usually require a sign up as well.. so now users need one account per good or service. (On Taku, a user could have a single account that spans across products).
Also, nothing stopping Taku from offering different kinds of chatting services in a single solution: Forum, Q&A, instant chat.
I suspect a selling point would be increased user engagement in terms of helping other users out... in the past, as a user I might come across a question on a forum i kinda an help out but I'd be too lazy to create an account just to post.
First of all, that's a very nice landing page and I can see this being useful.
I love that I can post a comment anonymously and don't need to sign up. Makes for a very low friction experience.
Some feedback for the landing page - it wasn't clear to me at first that the Taku box that pops up there is a live example. Maybe you can make that more clear with some copy change. For example:
"This is a Taku Box. All users receive this message. You can even chat with them in the comments below. Scroll down!"
Another small design feedback - it wasn't clear to me at all that "Yodelling R2-D2" (currently the top comment) was a user. I thought maybe it was a caption for the image, or an example product name, or something else. I found it confusing. Maybe a small auto-generated avatar or even just an icon of a profile next to the names would help, something to signal to me that it is a username. Or maybe even get rid of the names altogether, since they are all anonymous.
I suggest to change the title to something like "Taku: a better way to communicate with users" or "Taku: a widget to communicate with yours users".
I don't understand how it can be a widget in the webpage and in the app. Are you shipping two versions? Which framework do they use?
Email has the advantage that it can be send to people that didn't visit your site and perhaps forgot about it, in spite they may mark it as spam.
It's usually better to wait to post it here until it's available, and it looks like you are only building a waiting list. But IIUC the thing of the right is a live demo, so it's not yout vaporware or an idea without any implementation.
What happens if I press the button "Subscribe to Launch"?
After looking again carefully, if I press the button on the left side, then the cursor jumps to the form in the right side. The problem is that it's not obvious and foolproof (at least for me). Perhaps you can make the field in the form on the right blink once or twice, or some similar animation to catch the user attention.
I set up a public chat room once. It was incredibly good. We banned one guy for posting erotica but otherwise it worked. The problem was the chat was really just a custom "wall" connected to my back end. It was hard to maintain and moderate, and wasn't live.
I tried to replicate that recently with Discord, but Discord doesn't feel anonymous enough.
I would actually love to have something like that again. Just a public chat widget suited for mobile, without needing to learn XMPP or integrating IRC or setting up my own server. Actually an old school chat room would do. I just don't want reacts and profile pics and all that junk.
I'd be quite willing to pay for it too. Not $5/month, but something.
When you load up a webpage and there is a little box like this in the corner of it offering chat, how quickly do you close it so you can see the page it's obscuring? Do you even notice it or is it just another thing in the list of twenty popups you have to close before you can see the actual page?
I talked with some founders and they don't like they live chat. It designed to talk with only one user at a time. Taku is designed more like community building tool but you can collect data from users as well.
I like the idea but I am not sure if a discord server / whatsapp will get indexed by Google? If not you might loose out on getting free traffic. You might just consider a good old forum, e.g. discourse  which is pretty easy to implement and what we use .
My first thought was that this is chatbot like functionality. I'm now desensitized to chat bots on websites. As soon as I hear their chime I look for the chat window to close it.
My second thought was when Microsoft used to integrate feedback in their desktop apps.
But, it's an integrating concept to use radio buttons for quick engagement is a great approach. Compared to the previous two examples, this seems much more focused. Seems like a great way to get quick feedback!
That's what I saw as well. But I talked with some founders and they told me they have some trouble with discord as well. Like there are a lot of noise, it's not designed to handle multiple threads with different users, when you wanna do 1-on-1 chat you need some data on users (how often they use your product, from what country they are etc).
Something to dig into more might be the developing PLG software space. This idea is not dissimilar to PLG in general, though I think most PLG cases are for larger implementations. PLG covers a lot more but might be worth diving into a bit.
PLG, Product Led Growth, is similar to your implementation in that you are trying to observe how users, not buyers, are utilizing your software so you can sell/expand to them.
You'll need to build in content moderation in an open group chat (ala shoutbox).
If this is more a one to one chatbot, this is similar to existing chatbot solutions in the marketplace but you could evolve your own spin on it. It seems to be positioned for early stage websites/apps where you can keep up with a low volume of active users.
> then you’ve probably realized that there is no good way to talk to them.
How about not talking to them? Have you considered that they don’t want to engage in a communication-level relationship with every single app they use? How about respecting their time and emotional energy and not bombarding them with communications?
Telegram/Slack/Discord were mine inspirations but they have some troubles. First it takes time to get people there and not everyone have telegram, slack or discord.
They are also not build to collecting and analyzing data. With Taku you can collect data from users and then target them using this data. For example allow DM only for people who uses your product daily.
this is not so useful because your polls wont be search-indexed by search engines so people wont find them by searching app+keyword.
so usecase like uservotes to add/change feature will have duplicates.
plain polls in itself had their issues - people answering polls arent people using your app, unless poll is nagged in the app/app settings(so a/b running in parallel to poll can help evaluating user feedback - remember google chrome bookmark page fiasco),
if someone resorts to searching apps site they usually either evaluating if app has something that they need to accomplish with app; or if there's a workaround or a way to do something they dont understand how to accomplish with your app - so if they didnt find a way they will try another app or do it other way.
there's much simpler live chat/search forum(Discourse)/discord all in one approach:
allow users to create topics(akin to discord rooms) relating some aspect of your app.
create flat not nested, not collapsible live chat in each topic.
allow user either post new message in topic or reply with a button to existing message - creating a thread this way.
allow people to subscribe to threads, maybe in side panel.
auto color user messages and threads background - the way google wave did it
allow to +- rate messages in threads if user posted a message inside; if message rating is below threshold visually change it's backround - like paint half of message backround into different color, but dont hide it.
allow marking messages as offtopic - color them differently too.
for single messages without thread allow rating it only to people that answered at least a few times and doesnt have much of their replies as completely off topic.
this way you wont need much moderation besides removing obvious hate speech or creating topics not related to your app.
allow indexing every topic - either page them by day (hour if very chatty), then group day-pages into week pages, allow fulltext+exclude search.
no need for fancy discord/slack like webapp chats, their sole purpose is to limit outside access. your goal Is outside access - to make every aspect of your app(or its abilities/features) as discoverable and explained as possible, and to allow discussing user wants and approaches.
the main key is to make it as findable, searchable, readable(people often dont wat to fight search and just skim read) and approachable(by not having echo chamber hate community) as possible.