1. Pore over your data to segment your users' behaviour.
The first step when launching any product is to make sure that you are collecting good amounts of data so that you are able to segment your customers. Tools like Google Analytics, a good event analytics platform like Mixpanel or Heap and your own database are usually good enough to start with.
With the data in place you can then look at whether a particular segment of customers are performing worse than others and then focus your user research to get the maximum product gains.
For example, it might be that users coming in from a particular marketing channel are performing worse than another. In other cases, it might be that certain users are using the product in very different ways which lead them to drop out sooner.
2. Retargeting bounced users.
Finding users that have just visited your homepage and then bounced can be the hardest people to contact. They might have clicked on an ad and then left and it would be really valuable if you could just know why. One technique you might want to consider is to retarget these customers using paid ads asking them why they didn’t like the product. The ad either drives them to a landing page offering them a voucher or cash in exchange for their feedback via a survey or a call.
If you want to get a bit smarter you could configure your retargeting so that it only targets users who have been on the page for more than 20 seconds. This means that you will only retarget users who were more interested in what you were offering.
Pros: You connect with people that literally bounced from your site.
Cons: Potentially low conversion rates and can be expensive.
3. In person user testing with paid recruits.
Another way that you can gather insights into what is not working with your site is to recruit users through Gumtree or Craigslist. This is a very efficient way to grab a whole bunch of new users that you can bring into the office and observe them as they use the site and ask questions.
In order for this to be effective, make sure that you screen all responders to the ad with a small survey so you can get the most appropriate people. Ring the finalists for a sanity check.
If you are trying to work out first reactions to your product, it can often help to not mention your company beforehand and try to let the person use your product first without giving them lots of information. Watch their reactions carefully and help them to talk aloud. If you have a competitor in the market, then it is useful to show them their product as well so you can get a comparison (remembering to randomise the order to increase the fairness of the results).
The book Rocket Surgery Made Easy is a great primer on how to run these types of sessions effectively.
I am always surprised at the quality of users I find. For example, we placed an ad for Patch (Online Urban Gardens) and we had over 80 responses. In a couple of days we were user testing, and by the end of the week had lots of useful feedback.
Pros - Very quick to get users and not too expensive (we normally give £30 to each person).
Cons - Small sample size of customers that have not shown intent and takes a bit of organisation. Better when you have more product to show.
4. Live site remote testing.
If you have enough traffic,you could try live remote user testing with products like Ethnio. This tool will ask customers as they are using your product whether they would like to participate in a user study in exchange for a voucher. If they say yes then you get a message and can call them up immediately to get their feedback while they are in the moment.
Remember to make sure to only have the popup on when you can respond straight away and when you do call, try to put people at ease with some friendly chat before firing lots of questions!
Pros: Ethnio is great for talking to live site users so that you 100% know that they are the kind of customers you want to talk to.
Cons: The conversion rates on the screeners are very low, so you typically need higher volume traffic in order for this to be really effective.
5. Capture email address as soon as possible.
When you are in the early days of your product and you want to optimise for learning, it is important to have a way to communicate with your customers. In most cases, this means you should capture their email address as early as possible. Even if it means you lose some customers in the process. Without that email we can’t learn as fast.
For products where a user account is natural this is pretty straightforward, but I would also suggest this for ecommerce and marketplace businesses. If it really harms conversion,at least offer your customers a discount or a gift in exchange for their email as soon as possible.
For a lot of businesses email is not only great for product development questions, but also for dramatically improving CPAs and retention using email campaigns.
Secret Escapes have even made the email sign up a core part of their business strategy using email marketing to dramatically improve CPAs.
6a. Add a customer communication tool into your product for automating feedback messages.
I am a big fan of tools like Intercom or Drift that let you add live chat into the product experience and keep all emails and messages in a single interface. These tools are also great in that they allow you to drip feed in automatic communication based on a customer’s behaviour. This means you can create an automatic steady flow of responses as to why customers have abandoned your product, as well as emails or messages to encourage them to come back.
I will typically have at least a couple of automated emails running, one for asking new happy customers to get feedback on the service and also another email that targets lost customers to find out why. I will also send out NPS emails on a regular basis to get a score for the overall product experience.
These tools are most effective if you commit to using them properly. That means making sure all communications go through the tool, updating and tagging your customers (manually and automatically) with relevant information about where they are in the product cycle. This makes the data more powerful and ensures automated emails are going to the right people.
Pros: Amazing for creating a stronger relationship with your customers. Lots of powerful tools to have constant dialogue with your customers.
Cons: You need customers to sign up to get the full power of the platform and you have to keep it up to date.
6b. Send short, personal emails to your lost customers.
When you communicate with customers, it is best to keep your messages personal, open ended and very short. Here is a great example of a message sent with Patch when trying to figure out how we could improve the product. This message alone had a 7% response rate, which is pretty amazing and has provided a wealth of information to improve the product.
Pros: You get amazing insight just by sending a simple email.
Cons: None whatsoever.
7. Exit surveys (web only).
Another approach to think about is to use an exit intent popup. Exit intent popups only appear when a user is about to leave your website by moving the mouse towards the menu bar. Some users can find this annoying, but if your product is struggling then you can quickly gain insights without a lot of effort. Also, the fact that it appears only when they want to leave means you don’t affect conversion rates.
As an example, in the early days of working with Snaptrip (last minute self-catering properties) our conversion rates suddenly dipped for no apparent reason. We couldn’t figure out what was wrong, and we tried to see if we had broken something or if recent product changes had a delayed effect. We added an exit intent popup to the site and within a day we had 10 responses telling us that there were not enough properties on the site. We then dug into our data in more detail and saw that our supply of properties from their early partners was drastically reduced over holidays.
The company quickly reset their priorities to get more partners on board so that quantity of supply would not be compromised for these periods.
Pros: Easy to set up, can give you some information from real people leaving your site.
Cons: Hard to target specific customers, can be annoying for customers, low conversion rates.
8. Call them up.
If you have a phone number as part of the product flow, you should phone your abandoned customers and ask them how you could have improved the experience.
As with any cold calling, you will get people that are busy or don’t want to talk. However, I am always amazed that as soon as people realise you are not selling anything how much time they have to give you feedback on your service. 10 conversations and you will probably get so much information that you won’t know what to do with it.
Pros: Rich source of insight and the ability to actually have a conversation.
Cons: You need a phone number which might not make sense for every product.
9. Leverage your happy customers.
Lastly, when you are trying to understand why customers don’t want to engage with your product you shouldn’t forget to speak to the ones that do. It is quite normal when you are moving fast that the value you think is important is actually different to what your customers value. By having conversations with your happy customers, you can learn a great deal about what they actually find valuable with the product. You can then try to see if you can make this value clearer and more immediate in the product flow.
Another useful question to ask your happy customers is “What nearly made you not use the product?”. This can often highlight doubts and hesitations that customers that didn’t engage might also have.