6 ways to get honest user feedback on your product

Building a successful product is all about repeating this cycle as many times as possible:

“Speak with your users” is one of the most common pieces of advice you’ll hear. The problem is that it’s not that easy. You’ll usually hear back from users who had a really good experience with the product, or from those on the other side of the spectrum – users who experienced technical issues. However, the most important segment are the ones in the middle – those who tried out the product but didn’t buy, who thought it was just OK. How do you get these users to tell you what they don’t like about the product?

Over the last year, thanks to the amazing work of Ophir, Chen, and Alex, we were able to increase the percentage of users who communicated with us and to get their feedback: we went from around 10% to over 50%. Here are the main changes which led us there:

Getting users to reply your emails 

1. Best subject lines:

• Personal subject lines – One of the best-performing subject lines we use is “Iris from Takipi reaching out”  or just “Iris from Takipi” (or any other name among our teammates, of course). A very common approach is to add the user’s name in the subject line:  for example, “Hi David, we have some updates for you.” From our A/B tests, these user’s name emails haven’t received a much higher reply rate, likely because everybody is using this method today and people don’t associate it with personal emails any longer. However, using our name in the subject line increased the number of replies.
• Add the company name to the subject line – While adding the user’s name to the subject line didn’t perform well, using the company name led most of the recipients to open and reply to the email. “Takipi @companyname” is one of my favorite subject lines.
• Contextual subject lines – One of the worst performing subject lines was “Welcome to Takipi.” That was a huge surprise. We were sure that users who had just signed up were looking to get more info about the product. We were wrong. From tests on different emails, we learned that users are likely to ignore ‘general’ emails, but will open up emails they think they need. We changed the subject line to “Your secret key is ready” and had an increase of 25% in the open rate. Here is a great post by KissFlow: Ditch your Welcome mail. By changing their welcome mail to a more contextual one and changing the subject line to “KissFlow is *not* for everyone” they managed to increase engagement by 10X.

• Interesting subject lines – One email we struggled with a lot with was asking users who stopped using the product what had led to their disinterest. The subject line which worked best was “Did we do something wrong?”. Most users opened this email in order to understand what we’re asking, or if we had done something specific which was wrong.
• A/B test your subject lines – A/B testing subject lines requires minimal effort and can make a huge difference. For our main emails, we run at least five tests on the subject line. We use intercom.io for the test; its metrics are great and it’s super simple to create the tests. The only caveat is you can run only two tests at a time.  Here’s an example of an A/B test we’re running this week, and how easy it is to track the different stats (open, click, reply)

Here’s a great video illustrating a part of Daniel Pink’s book To Sell Is Human, which deals with picking the right email subject line. Researchers from Carnegie Mellon University found the top two subject lines categories users are more likely to open:

2. Keep it as personal as possible

The first challenge on the way to getting feedback is having people believe that there’s a real person on the other end and that someone will read their reply. So, to start with the obvious, always email users from a personal mail with your name, rather than ‘Support Team,’ for example. If you’re using a third-party email service, connect it to a real email address. I like the sender name to be as simple and standard as possible – “Iris Shoor” rather than “ Iris from Takipi.” Make sure you have a picture associated with your account.

Stop using boring email signatures. Here’s a small tweak we tried a few months ago which helped us increase our reply rate. Instead of going with the standard signature – “name, title, phone number, Twitter …” – we decided to personalize it. This is how our signatures look today:

A cute picture, explaining what we do at the company, a personal detail (Is this person a Game of thrones or Breaking bad fan? what’s his favorite band?) and link to a post this person wrote. The key thing to remember is that people answer people, not companies.

3. Automated emails did right

Understanding the context better is the key to starting more conversations. We’ve created a system of automated mails based on different behaviors or on problems users often face. For each technical problem, users will get a different automated mail, referring to the problem and asking if there’s some way we can help. Sending emails at the right time can also make a huge difference. If we want to send a feature update, we’ll wait until the user is using Takipi and then send the email. If we want to ask a user for his feedback, we’ll wait for a certain engagement level (for us, for example, that comes when a user has opened his tenth Takipi event). What have we learned so far about automated emails?

• All the automated emails are written as a personal email. There’s always someone online who can pick up the reply.
• Avoid ‘The Big Brother’ approach. When we started writing event-driven automated mails, we used all the information we had. The user has been using Takipi for over a month, has 4 servers and is using Linux? Let’s mention it! These emails got a really low reply rate. Even though we usually have lots of data on the problems users are facing, we send out a simple email asking if there’s anything we can help with.
• We usually wait a while (10-60 min) until we send the automated mail. Otherwise, it doesn’t look personal. For example, we send an automated email from one of our co-founders asking users right after installation if they need any help. When we sent it right after registration, the open and reply rate was lower than waiting for an additional 20 minutes.
• Make sure you don’t land at the promotion inbox, as that can decrease the open rate by 70%. Use plain text and avoid including too many images and links.

Alex from GrooveHQ shares some great insights about using event/ time-triggered support emails: How A Single Email Made Customers 350% More Likely To Convert.

4. Using online surveys

I have to admit that I was a bit skeptical at first about using a survey; it seemed a bit old-fashioned and tedious. However, it is effective. We have been sending an email to engaged users if we don’t see them for over ten days, asking them to fill up a short survey.

Best practices:

• The goal of this survey is to really understand why users stopped using the product in order to improve it. We don’t try to sell the product there or to convince users to come back. We ask direct questions and get the answers we’re looking for.
• We ask only four questions; that’s usually enough to get the most value. The most important question is an open-ended one: “We’d love to hear more about your experience.”
• I believe a simple Google form can be a good solution. However, I do feel a more complex tool can look more professional. We’re using Typeform. The things I like best about it are its customization options and ability to connect it well to the brand.

Here is another great post from GrooveHQ – How We Grew Our Customer Exit Survey Responses by 785% on how the changed the main question to get more valuable feedback.

5. Online chat

One of the reasons we postponed integrating live chat was the concern we would need a 24/7 presence. The truth is that even if you turn it on for only 2-3 hours a day, you can get super valuable feedback. Choosing whether the chat option is displayed to users requires just one click.  Another thing which surprised us was the simple integration.

Why choose an online chat over a feedback form? There are lots of great feedback solutions (intercom.io, service), and I’ve used some in the past and learned that feedback forms provide very different data than live chat. When you ask users to type their feedback or ask questions you can’t start a real conversation. Live chat definitely consumes more time and resources, but it enables you to learn much more about the users – how they’re using the product and which features they’re missing.

Best practices:

• We found that customizing the default message to the stage the user is at right now can increase engagement. For example, a user who hasn’t installed the product yet will see the following message: “Need help installing Takipi?”. A user who’s visiting your pricing page will see an “Any questions about our pricing?” message.
• One little-known feature of live chats is the ability to initiate a chat with a user. If you receive an email from a certain user or see that a user is running into a problem, you can easily start a chat. It’s a great tool to offer better support and start conversations with users.

6. Using giveaways/ discounts to get feedback

 Last week I signed up for Slack and for the first time had a really good experience providing more details about me and the company. Here’s what it looks like:

I was asked to fill out a personal data which will obviously be used for sales (company size, title, tools we’re using today) but this step was optional, and I got a $100 discount for filling it. The funny thing is that I had just started using the tool and had no intention of buying it, so the $100 discount shouldn’t have been so compelling. But, I guess I went through the same process most people do: $100 for 6 simple questions – I’m in!

 Another great way to get feedback is to give away something for free and then ask for optional feedback. It’s a basic reality of economics – when you give something for free, the receiver is much more likely to give something back. It doesn’t have to be something physical, either. Here’s a great example – This email from 99bitcoins highlights for the user that he/ she received value for free and asks them in return to tell the sender which content they’re most interested at. It’s a great way to engage with users and learn more about them.

This is an email we send to users we’re interested in and whose feedback we want:

Notice that we don’t ask them to answer the questions in order to get the T-shirt. Yet, over 50% of the recipients send back their feedback, usually with very detailed answers.

We hope you enjoyed reading this post. If you wish to analyze your website and marketing efforts, and really improve results, you are welcome to try our new analytics tool for Free.