Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Need your opinion–what would you use this for?
5 points by giordanobd on Feb 25, 2016 | hide | past | favorite | 13 comments
Hey there!

Over the past month I built a service that allows you to see your users browse your website / web app in real time. It's kinda like "session recording" offered by various services, but live, mainly to offer support.

To make it easier to market I built a Chrome extension that integrates it with Intercom and other live chat services, so that when someone chats you up you can see their screen live and troubleshoot problems a lot quicker. Some early users utilize it this way.

However, I personally like a lot more the idea of having a "monitor" with say 6 screens and being able to send a message to people you see struggling, to offer "proactive" support. This comes from seeing way too many of my users on other apps make dumb mistakes and seeing myself losing money after they leave frustrated. I would really have paid to send them a message on screen saying "hey, let me know if you need help with this :)".

I'm not sure whether: 1. this "proactive support" is creepy; 2. it is scalable; 3. it only works in my mind; 4. companies would pay for it; 5. the ease of marketing something "for intercom/livechat/..." outweighs the coolness of the monitoring station.

What do you guys think? Would love your opinion!

Website is https://peekin.io/




1. If users are NOT aware of being tracked, it is HELL creepy 2. Scalability is now not an issue :) I think your concern should be to find the first customer. 3. not sure 4. I think they would, but you need to make sure users are not creeped out by it. 5. Unsure. You might want to try both ways. Start marketing a specific use-case and see what happens (if it works, etc..).


By the way, what I meant with scalability is that if you have up to say 10-15 users at a time you can keep track of them easily, if it's more you either need more people or the whole monitoring station becomes worthless


Agree but you could still easily gather events and metrics, which once aggregated would provide BI insights. Managers do love this kind of things


Thank you for your opinion!

Regarding it being creepy—I'm not sure how I see it myself.

In a way it's a lot more creepier that advertising companies know your tastes and have a shitload of info about you--sometimes more than you do.

At the end of the day we only track the HTML the site owner provides in the first place, and stuff like mouse movements within the page, without it ever being recorded.

On the other hand, I agree that the fact that you "see" the page being used and not just collect metadata could be deemed very creepy.


"Proactive" desktop support is very common in larger organizations where the time needed to personally dispatch a technician to each user in meatspace would be prohibitive. Usually such remote proactive support is initiated with the user's direct approval.

In my opinion the "consenting adults" principle applies here: it doesn't hurt to ask for a user's permission, and if they agree, it's not creepy. If they don't agree, no harm is done.

One last thing: you're right to worry about scalability. High-touch support is fundamentally not scalable (someone please prove me wrong!), which is why you see companies like Google providing automated "support" that makes customers feel alienated and not valued.


I get what you're saying—I guess what I'm trying to understand is whether a regular user has an unspoken expectation that browsing or using a web app is "private"--in which case that expectation is already broken most of the times by usage of analytics and, less often by session recording tech--or whether it is like walking in a store: no one is bothering you but you know that there's someone that might be seeing what you are doing, and if you stand in front of an item long enough someone is gonna come over asking if you need any help.

Apps like intercom already allow you to specify "triggers" so that when the user does something you can send them a message. I'm not sure how many people actually realize that it is an automated message.


Considering how much brouhaha there has been recently in the media (legitimately or not) regarding privacy issues, I doubt anyone expects true privacy when using SaaS products or any old websites these days. There are obviously those people who really care about privacy, however those are a minority and probably aren't worth worrying about in your case.

I think making this work well (assuming underlying tech is solid) is a matter of UX technique. Iterate on a solution that makes the user aware of your support capabilities, then make it easy to summon help. Perhaps it'll be more natural if the user must initiate each support session, i.e. by highlighting a section of their screen and twiddling a control which makes it clear that human help is on the way.


Woah! Take my money!

Though I really wish you would consider selling the software and allowing it to be self hosted for security reasons - I think this is great!

My users really aren't the brightest (that's not to put them down; computer savvy users just aren't my target demographic) - so it would be awesome to see what their struggling with.

That - and for months I've been having an incredibly hard time hunting down 1 specific bug - I've setup all kinds of traps, triggers - still nothing.

Maybe if I can actually see it happen, it'll shed some light on it.

1 problem - in an AngularJS 1.x app - this is throwing a lot of JS errors. The site is working fine with it in place - but the console has a lot of red in it pointing to your script with no specific error message.

Message me if you wanna see for yourself.


Hey there! Thank your for your feedback!

Are you seeing errors on your website or peekin.io? There are a lot of console error on peekin.io that are due to the html being sandboxed and there is no way around that.

Otherwise I can take a look :)

--

As for the product itself: I'm not sure—are you saying you'd use it to watch your user "proactively" or when they ask for support?

Thank you!


The errors are occurring on my site, after including the script.

I intend to use the app to watch the user proactively. I have various triggers in place that let me know when a user is on the site / what their looking at / etc... (the usual metrics) but 1 bug continues to elude me. Letting me see what the user is actually doing might help me find it.


If a "Stuggling with X?" window came up on bottom right hand corner I might not be as creeped out if I thought it was just help


Its creepy. Creepy enough to never come back. I did that once and i am pretty sure the user never came back.


In most cases creepy, creepy enough to probably leave a site and never come back, especially if it is things like payment details or a service where I store data (so basically, nearly everything that is complex enough to make this service "useful")




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: