Ask HN: How to effectively get feedback from users?
I mainly use email to ask for feedback and the response rates are terrible, 2-3%.
Are there any effective ways of getting feedback from customers? Tools or templates?
I mainly use email to ask for feedback and the response rates are terrible, 2-3%.
Are there any effective ways of getting feedback from customers? Tools or templates?
[+] [-] grandvoye|5 years ago|reply
* If you're pre-product/market fit, just reach out to users via email, in-product messaging etc. You can incentivize them with Amazon gift cards as well. At this stage nobody actually needs your product so you'll have to do it by hook or by crook.
* If you're post-product/market fit then people actually _need_ your business. Once you make it clear that you're going to action on their feedback they're highly incentivized to help you out – so use the same strategies as above but make it clear what you'll use the feedback for.
* Set up automated surveys, especially ones that are unobtrusive in order to not create a shitty user experience – this allows you to baseline your product's effectiveness which is the first step towards continuous iteration. Eg last quarter we were at 3.3/5 satisfaction, let's get to 4.0/5 this quarter. These surveys are also a great way to find customers to interview.
* FullStory and Hotjar both allow you to view user sessions, highly recommend them as well.
The most important point: if you're an enterprise business, I highly recommend identifying the "best" customers that you have, building relationships with them, and favoring their guidance over others'. Once you hit scale all of your customers will want to give you feedback, but only some of them will have the wisdom/intelligence/creativity/whatever to have a great sense for what _you_ should build for _your_ business to succeed. When you find these customers, get them onto a customer advisory board / meet with them a lot as they will help you in the art of pulling a great product out of your market.
[+] [-] redman007|5 years ago|reply
[+] [-] inetknght|5 years ago|reply
If you value your user's privacy (as well you should if you ever want to do business in the EU) do not be tempted by this privacy-hostile path.
[+] [-] XCSme|5 years ago|reply
[+] [-] aktionman|5 years ago|reply
[+] [-] diggan|5 years ago|reply
[+] [-] arjunvpaul|5 years ago|reply
Beauty of it is they sneeze, we know. Everyone in the team kinda knows what's important what's not. we can confidently say things like "nobody cares about that fancy feature". It super valuable for a startup to get immedeate feedback. Especially valuable has been knowing - the first things the notice when they start using: where is sort? how do i add team members? do i have to prepay? - feature requests: it's painful to keep closing windows, need the ability to adjust windows, I also want to do this other thing you guys havent thought of. - Bugs: why am i seeing double messages.
Disadvantages are that some customers expect immedeate answers, unrealistic expectations. and its cringey to lay down the rules and set expectations every time. no matter how many times you do it.
Our take is its 80% good and 20% bad. As we grow the company, the idea is to keep doing this for a "representative sample of customers" . Customer feedback is gold. If you can setup some kind of process like this, its great.
[+] [-] DenisM|5 years ago|reply
[+] [-] ignoramous|5 years ago|reply
[+] [-] devxpy|5 years ago|reply
[+] [-] disillusioned|5 years ago|reply
https://amzn.to/383bo92
It talks at length about the challenges, especially, of getting _honest_ feedback from prospective users. And it has a lot of ideas to that effect.
[+] [-] SquareWheel|5 years ago|reply
https://www.amazon.com/Mom-Test-customers-business-everyone/...
[+] [-] dgr582systems|5 years ago|reply
[+] [-] kareemm|5 years ago|reply
At the early stage my experience has been:
1. You should install SmartLook, FullStory, or similar to watch users. Qualitative feedback beats quantitative b/c your raw numbers are too low to be useful.
2. Identify key points in the customer journey when you want to collect feedback. For B2B SaaS I ask for feedback via email at these stages:
- When a customer signs up for a trial
- When a customer converts to a paying sub
- When a customer doesn't convert
- After 1-2 months of paying
- When a customer churns
First I'll make sure that I get system-generated emails when these events happen so I can send them manually.
I'll eventually automate sending these emails once volume gets too high. I'll automate them by sending from our app or from a tool like UserList or Intercom.
And when I get feedback I centralize in the tool I'm building (www.savio.io) and use it to prioritize what to build next.
[+] [-] mxstbr|5 years ago|reply
We have found that our customers collect up to 35x more feedback than before once they add the widget to their websites! It does depend on where you place the "Give feedback" button though — the more visible it is, the more feedback you'll collect.
[+] [-] Aldipower|5 years ago|reply
[+] [-] tobiaslins|5 years ago|reply
[+] [-] harrybr|5 years ago|reply
The responses on here are reasonable but a bit short sighted. What works in one context will not work in another.
[+] [-] brokebroadbeat|5 years ago|reply
Heaps of information on how to get started in this blog: https://userresearch.blog.gov.uk
[+] [-] Jugurtha|5 years ago|reply
If you don't have any, log user actions when they succeed or fail to do something. Even a simple log can tell you many things.
It also tells you if they're failing to discover or use a certain feature, or if they're using something in an unintended way.
You can set up Slack channels for your product to help users troubleshoot issues. We have workspaces where users can get help. Whenever the 20th user complains about something, what to do next becomes clear. We tend to focus on issues with high frequency and high severity as opposed to say stylesheets, but it's because the field is struggling not for lack of good CSS. But it may be different for you and you may have a product where the value of the product is the better interface. Adjust accordingly.
- [0]: https://posthog.com/
- [1]: https://github.com/PostHog/posthog
[+] [-] toast76|5 years ago|reply
We had this problem with our product (BugHerd) and built a solution for it. Our success rate was ~1% and we thought we could do better.
We created a JS embed that you install in your web app. It allows you to segment your users and place a call with them from within your app (using webrtc). Unlike email/calendly/zoom square dance, Smidge has a 17-20% success rate to call. It's been huge for us and we're now opening to beta. Would love feedback
(Screensharing, "request a call" and page/user calling coming soon!)
[+] [-] Aldipower|5 years ago|reply
[+] [-] aliakhtar|5 years ago|reply
Edit: Found the answer on the video. Looks like nothing is shown until the founder starts a call with the user.
Suggestions:
- It'd be great if the user could initiate calls too, e.g if they fit in a certain group, let them initiate a call as a form of 'customer support'
- Can you allow audio only calls?
I LOVE this idea though. Will definitely be a user.
[+] [-] oleks637|5 years ago|reply
- I have installed a chat popup on my landing page and I always try to answer any question within 1 hour, and then follow up with those users afterward.
- I have a feedback portal (similar to canny). Whenever anyone leaves any kind of feedback, I write them a thank you email and try to elaborate more on the feedback they left, no matter if this is negative, positive feedback or feature request. For example, if they ask for a feature, I tell them when I plan to develop it.
- Whenever user purchases a license for my app, I also write them a thank you email, and ask for a small feedback. I try to do a short investigation on what user is doing and include some personal touch in it. Usually, 50% of such emails get an answer.
- If I see some bug report coming from a known user I immediately send them an email explaining why that bug has happened, how and when I'm gonna fix it.
I also keep a page in my notes called 'Followups' where I put the user's name and the reason why I should follow up with them. For example next week I'm preparing a new release of the app that has changes which I know are used heavily by a couple of users and I will reach out to them this week, give them an alpha build and ask for their feedback. Whenever I get an idea of how I can bring someone's attention to the product, I add their name to this list.
[+] [-] oakst|5 years ago|reply
For qualitative data such as user interviews/calls the solution for most teams is either to spend a lot of time on find users to talk to and set up calls or to pay companies like usertesting.com for access to their panels.
These tools don't allow you to get quick (and for most teams affordable) access to your own users for user interviews though, which is something that was enough of a pain point within my own team that we've built a tool exactly for that (https://tryribbon.com).
The Mom Test is also a great read on how to get useful user feedback in general (https://amzn.to/383bo92).
[+] [-] hermitcrab|5 years ago|reply
"We are very interested in how you are getting on with <product>. Please email us if you have any suggestions about how we can improve <product>."
But that only works if you are getting sales.
Also the more responsive you are to customers, the more likely they are to report bugs or make suggestions.
I wrote about my experiences getting feedback here: https://successfulsoftware.net/2008/04/28/getting-customer-f... (old article, but still relevant)
[+] [-] JNRowe|5 years ago|reply
My assumption being that if you're sending unsolicited mails so quickly, then I should be expecting a truckload to arrive by the end of a year. In one of those cases I felt vindicated as I received a bunch of huge exit interview type replies.
[+] [-] philippz|5 years ago|reply
Our idea is to keep the perceived effort to give feedback as small as possible and that works quite well. Also, it can embedded and triggered in any software to collect feedback right in place.
Feedback is collected publicly by default. Here is our own page: https://www.stomt.com/stomt but of course you can also turn your page private.
[+] [-] sillysaurusx|5 years ago|reply
A discord server is useful for far more than most people understand. You can show daily progress; show new features; highlight community contributions; have a centralized bug report channel; and the other users will start solving other users' problems when it attains critical mass.
The nit is "critical mass". But if you make something that people find remotely interesting, people will join. Broadcast what you're doing on twitter, and eventually you'll get noticed.
Surveys and cold calls are like trying to make fire by banging stones together. Tools from a different era. Sure, they work, but I've found the above process to be far more effective.
It helps to make your discord a general topic, by the way, rather than a specific product. Our ML discord recently passed 1,000 users. https://discord.com/invite/x52Xz3y
Community building is tough work, and I've had several failed attempts. You shouldn't expect yours to take off immediately. You should do it because you like it, and eventually something will work.
But it depends entirely on the business. Most businesses don't need a community. But I can't think of a single business with a community that would have been in such a strong position without it.
[+] [-] amarghose|5 years ago|reply
[+] [-] hartator|5 years ago|reply
It's pretty effective at getting very early bug reports, gathering interest on potential features, and their initial use was for feedback gathering. Random users won't give you feedback however until they came across a bug or a need though in my experience!
[+] [-] tobiaslins|5 years ago|reply
* Hitting up active customers on Twitter, just writing a short message: "Hi xyz, just saw that you are using Splitbee quite actively. Do you have any feedback for me, that would help a lot"
For this method I got answers from 9/10 people with amazing feedback. Very detailed and long messages. Be sure to target "power" users.
* I sent automatic emails if someone embeds splitbee successfully to their page & a mail if they don't do it within 24 hours. People often engage with them and explaining why they did not embed it yet. Also pretty helpful
* We added the feedback.fish widget to all of our https://splitbee.io/docs/embed-the-script and in our app. We get about 8-10 submissions a month which is quite a lot for our size.
[+] [-] tbran|5 years ago|reply
Go where your customers already are and become a part of their community. Be a long-term person in a long term business. If they don't have a place, make it.
If you have that, you're on your way to Product-Channel Fit [1].
Low level:
I've also sent pointless emails and made useless phone calls and most people default to doing nothing when you contact them.
So:
If you think phone/email are where your people are, you need to follow up a lot. Steli Efti has lots of great content on this type of thing [2]. This is an unpopular view on HN because most people see this as spam, but this is what CRMs are built for.
[1]: https://brianbalfour.com/essays/product-channel-fit-for-grow...
[2]: https://blog.close.com/follow-up/
[+] [-] CharlesIvia|5 years ago|reply
Instead, ask what is wrong with a product and you will receive genuine feedback. You can ask this question though email.
[+] [-] mindhash|5 years ago|reply
Write an email to talk share your research, how market is doing a certain thing, what's coming up ahead for you and them in terms of trend, etc. Engage in active discussions around the business that your customers operate in. Post that seek opinions.
John does a wonderful job of explaining importance of this here: https://www.indiehackers.com/post/the-power-and-importance-o...
Start a thread and see where it goes. Build an active conversation around things before seeking feedback. Good luck
[+] [-] k1w1|5 years ago|reply
Just knowing that someone is satisfied, or unsatisfied, with your product is not actionable. Often the reasons why a customer might be unsatisfied are subtle and would be impossible to discern with a questionnaire or survey. We think the answer is "empathy sessions" - kind of like focus groups meets modern online communication.
https://www.aha.io/blog/introducing-empathy-sessions
[+] [-] daniel_levine|5 years ago|reply