Humans are interactive by nature and they exchange opinions, thoughts and feedback almost inadvertently everyday. They reciprocate to sentiments and emotions and mostly react rather than act, in many cases.  The same behavior is exhibited while using any mobile app too. Every sentiment that is caused by some outcome is translated into their next steps. Keep a customer satisfied, it gives you engagement. Disappoint users, and they will leave your app. Such is the power of user emotion and sentiments. How are you keeping a tab on this?

Let’s understand this better: User Sentiment if anchored correctly can make metrics shoot high and if not taken into consideration they are going to say bye!

Every action done by a user on the app is conveying some sentiment if it is closely monitored. Every sentiment has a ripple effect on the end goal. The selling point in understanding them is if they are acted upon, it can boost performance metrics of the app immensely. When a user is at the last step before booking tickets and then he drops off, what could have been the problem? Why didn’t he complete the journey end to end? Understanding what every in-app navigation means  will help in iterating the product quickly to build a  powerful, customer-centric product.

How to understand user sentiments when data points don’t suffice?

Often there are use cases that don’t get concrete answers in spite of tracking granular level data. Consider a minimal onboarding funnel where a user has to log in with just one click after giving their mobile number or they can log in as a guest user. When you are trying to improve sign-ups, a common question that would get raised is “Why aren’t users registering?”. Some reasons can be OTP generation failed or network issues. But in spite of everything working fine, why aren’t they still registering? What are the challenges they have? Such questions can’t be answered with mere data points, but the job can be easily done using Customer Feedback - Surveys!

Surveys are one of the frequently used features to understand and analyse user sentiments. Whether it is for understanding if they are having a good app experience, or if they have understood the proposition of the app, or to understand the challenges that they are facing that stops them from being a happy customer - Surveys, Feedback forms, NPS Ratings help in evaluating where you stand for every user on the platform.

Surveys are not new these days! In fact, did you know there are almost 2 Lakh surveys being conducted every day! Out of this how many go unresponded? Almost 70% of them! After designing a survey to understand users better, you do not want it to go unresponded. How are you making sure it gets responded, and more importantly you are getting quality answers?

The perfect blend of UI, Audience, Relevance, and the value add the user gets is what marks the standard of the survey.

Having an end goal while designing the survey is a prerequisite before you start. This helps in asking the right questions to get the right answers. One of the biggest advantages of surveys is asking any questions regarding your product. But how will it help in building a powerful and more user-friendly app can be obtained when the right questions.

A lot of surveys don’t get the completed response from the surveys for the following reasons :

1. Irrelevant Surveys
2. Lengthy questions
3. Irrelevant options to select from
4. Never-ending surveys etc
   

Decide what type of feedback required

Once you have the end goal, decide on the template of the feedback survey. When you are trying to understand their overall experience at the time: an NPS style survey is the most efficient to get the results. When you want to check options that users would like to try in the near future, a multi-response survey helps. When you want to understand a single point why users didn’t complete a funnel (like an onboarding funnel, or purchase) a single response survey will get the one point answer on why they didn’t complete the process.

Recurring surveys over a longer period of time is also used to understand consistency. NPS surveys are often recurred over a period of time to understand if the user sentiment is the same or if it has improved over the course of time.

Decide what answers you are looking for

Next is the crux of surveys: Questionnaire. If a survey is very long, users will drop off. If it is not relevant, poof they are gone! Identifying short and fitting questions often helps. Surveys are often sent out regularly and having the impression that they are going to be long refrains from users taking up more surveys in the future. Anything that is non-essential is not worth the user’s time of answering and it should be removed. Crisp follow up questions will have better reasoning. For Example: Having a rating survey asking how the user experience was after making a purchase can tell the satisfaction level. It can be backed up by asking what made them give that rating. Based on the rating they give, they can be taken into two flows of options to select from. This will help in improving the user experience from the next time and from there take it to the next level.

Make the survey relevant

Most of the surveys go answered or users tend to dismiss due to lack of relevance. Industry standards for an excellent survey's response rate is anything above 50%. Relevance plays a critical role to achieve an above par response rate and almost accurate feedback.  Asking a user to rate the app right after a failed transaction is only going to pull down the rating. Whereas asking a survey to understand what went wrong during the transaction can fetch the right answers for why the user is disappointed with the app. This ensures that the user knows you are trying to serve them better which will help in giving more relevant thoughts as the options in the survey.

Contextualize your survey

Contextual Surveys are the new cool, and is the new hype around the regular surveys. This delivers in terms of response rates, better-answered questions which in turn gives a great opportunity for the product to grow and impress. Need to know more about the magic of context? Get a detailed explanation here and why it matters! Launching it to the right users at the right moment is where the gold lies and that is the mine you need to be drilling into.

It is important to remove order bias by randomizing choices when the users are shown the survey. The first step towards analyzing the responses would be to understand how many users responded ‘Yes’ and how many said ‘No’ to take up the survey.

Analyzing users who were not ready to take up the survey is an important information as well. The need for this? It can be because of the relevance point that we spoke about earlier. Users come to that app with a motive in mind and disrupting that flow can cause users to not take up the survey. This entirely depends on the contexts that are set while asking the users to take up the survey. Having the right context will help in boosting your response rates dramatically.

When coming to users who responded yes to take up the survey, it is critical to check how many completed the survey successfully and how many dropped off in between. This shows the threshold for the number of questions that have to be set for the user to complete the survey entirely.

Responses can be created into cohorts of users to dig deeper into their user behaviors and understand reasons for the given response. Patterns can be specific locations, discovering a functionality or even based on the internet they use. Let us consider a use case when a survey was shown to users who completed their first purchase on the app. The first question was an experience rating question where they were asked to rate between 1 - 5 and the second was a follow up question to ask why they had given good / bad ratings with a couple of options given to choose from. Let us assume that above 5 was considered as a good rating and below 5 it was a bad rating. While cohorting users who gave a good rating, it was observed that these users regularly came to the app and had checked out different trends before the purchase and were actively using the app. When the cohort of users who gave negative ratings was analyzed, it was observed that most of these users had a payment transaction failed. Drilling down deeper helped in identifying a specific bank was not successful in making the transaction.

Surveys are one of the most effective ways to understand user sentiment. They are not just about creating a form and throwing it on the face of the customer to fill anymore. Well thought out, designed and contextualized surveys are themselves a way of improving user experience. In some cases, they even deter the users who are on the paths of uninstalls. Let your survey reach out to your user, when they need it, rather than when you need it.