How To Use Polls To Increase Conversions
Polls are the most effective tool to understand your customers needs, wants, desires and objections. The closer you align your store and marketing with these, the more sales you'll make. The best part? Polls can be created in 39 seconds.
On the surface, increasing conversions is simple. In fact, it can be broken down into 3 steps:
- Learn when and where your visitors leave your Shopify store
- Learn why your visitors leave
- Fix it
Unfortunately, you soon uncover the hidden complexities that leads to guessing and as a consequence, frustration.
Step 1 is relatively easy with the help of quantitative data provided by Google Analytics. Learning why your visitors leave on the other hand, can be infuriating.
For a lot of merchants it can feel like a never ending battle, based on a guess. They change headings, pictures, colours and product descriptions. They’ve read they need to build trust, but how do you know what’s affecting your perceived trust?
Step 1 can be understood with quantitive data, like what Google Analytics offers. Step 2 involves better understanding your visitors, and the best way to achieve this is with a bit of qualitative research. The easiest and most effective way is using polls (also known as on-site surveys).
On-site surveys are critical for conversion research
Alex Birkett, Growth Marketer at ConversionXL
Here’s what you can learn from on-site polls
- Understand the root cause of cart abandonment
- Understand who your customer is
- Understand why people leave without buying
- Learn what friction occurs in the purchasing process
- Understand demand for new products
- Understand different visitor motivators and how to best provide for them
- Learn visitor objections (what’s stopping them buying)
Did you know nearly 75% of commence website visitors put an item in their shopping cart and then leave, never to return. Wouldn’t it be great to understand why? That’s where exit intent polls can be useful. Before someone leaves your store, you can ask them why they are about to leave without buying.
What is a poll?
A poll is an unobtrusive pop-up box that appears to visitors based on rules: when a page is visited, a user spends a certain duration on a page or when a visitor decides to leave a page (know as an exit-intent).
A scale poll in Feedme with the material theme.
A multi-choice poll in Feedme using the flat theme.
A yes/no poll in Feedme using the original theme.
How Teespring increased conversions by 12.7% from a single change using polls
Teespring, a t-shirt printing company ran a poll with the goal of optimising their store for conversions. The responses they received suggested visitors lacked trust in the brand.
They were receiving responses like, “I don’t want to give my credit card information” and “I’m not sure if I’ll receive my shirt”.
Teespring's solution was to create 2 Call-To-Action (CTA) variations:
One with negative small print, stating:
The other with positive small print stating:
The second CTA increased conversions by 12.7%.
What questions to ask
Renowned digital marketer, speaker and author, Avinash Kaushik was asked by the VP of a Fortune 100 company: “Which analytics tool do you recommend because we want to improve our website and increase sales.”
Avinash’s response: “Don’t implement a web analytics tool, implement a short website survey that has just three questions”
The three question are:
- What is the purpose of your visit today?
- Were you able to complete your task today?
- If you could not complete your task today, why not?
Now this is a good starting point, but we can do better.
Choose questions based on your goal
Understand why people are visiting your website (motivators)
By understanding why people are visiting your store, you can make your visitors experience as sleek as possible via on-going optimisation. For example, maybe people visit because they want to solve a problem.
Here are some questions you could ask to discover motivators:
- Why did you visit our store?
- What are you hoping to find today?
- What are your intentions today? predetermined responses could included: look at specifications, buy, just browsing and other etc.
These are usually the main reasons people aren’t converting. Questions to ask to discover why visitors aren’t buying could include:
- Is there anything preventing you from completing your purchase?
- What’s the one thing that may have stopped you placing your order?
- Is there anything we could improve about the checkout process?
- Has anything you’ve experience on our store annoyed you?
- What is your biggest fear or concern about purchasing from us?
- What would have convinced you to purchase from us?
- Do you have questions?
- How could we make our store more useful?
- Is there anything on our store that doesn’t work like you’d expect it to?
Understand what you’re missing
Sometimes people won’t convert because you don’t offer what they’re after, or you don’t have the right content for them to make an informed purchase decision.
- Are there other products you would like us to offer?
- Were you able to find the products you were looking for?
- Is there any information missing on this page?
- Is there anything on the page that is difficult to use?
- Is the product description sufficient on this page?
- Are the images sufficient on this page?
- Is there anything we could do to make this page better?
Sometimes it’s important to consult your target market before making decisions. A decision like choosing new product offerings. By asking questions such as: which of the following products would you prefer? Can give you confidence that your next product will be a hit.
Here are some of our favourite questions we think everyone should ask:
- How could we make our store more useful?
- What additional products should we add?
Fixing a high bounce rate:
The more visitors that go to your store without leaving immediately, the more sales you’ll receive, and the less your customer acquisition will be.
Here are some questions you can ask to tackle a high bounce rate:
- Is there any content missing on this page?
- Can you easily navigate to other pages?
- Have you managed to find what you were looking for?
A high exit rate from collection pages:
- Can you find what you’re looking for?
- Are our products appealing?
- Do you have any reservations about buying from us?
High cart abandonment rate
- Is there anything preventing you from completing your order?
- What made you change your mind about buying from us?
Watch out for acquiescence bias when phrasing questions
Cognitive biases can cause you to receive skewed results and also effect your judgment of responses.
A well known bias, confirmation bias, is consequential when you're phrasing questions and analysing results.
Confirmation bias is when you focus on responses that backup your current belief, while rejecting those that don’t. This first presents itself when creating a poll, causing you to phrase a question positively or negatively depending on your preconception. For example, if you already think the product images of a product are bad, you may say “Do you think the product images are unclear?” or, if you think they are clear, you may say “Do you think the product images are clear?”. This subtle change can create very different results due to another cognitive bias, acquiescence. Acquiescence bias refers to a phenomenon when people find it easier to agree with a statement, thus skewing results.
In this scenario the best way to combat this is by using a scale poll. In Feedme they look like this: “Please rate these product images. 10 is best.”
How to objectively analyse responses
When analysing responses it’s essential to keep your business goals in mind since they can determine your conclusions.
Take the question “are our products appealing?” for instance. Negative responses could ether suggest: you’re selling the wrong products to your target market, or you’re selling the right products to the wrong target market. The first would require you to change your range of products and the other would necessitate a target market change. Which one is right for you will depend on your business. For example, if you have been in business a while and know your target market well, then changing the products you offer would be the optimal solution. Whereas new businesses who cannot easily change the products they sell, should find a new target market.
Analysing responses objectively can be challenging due to pesky cognitive biases
Confirmation bias pops up again during analysis. When you’re looking at responses it’s easy to pick out the ones that confirm your preconception. For example, the first 3 responses may confirm your belief that your product images are unclear. So, you take action and hire a photographer. In the mean time, you receive another 500 responses that you ignore that suggest the images are fine.
How to create a poll
Step 1 - choose a tool
Feedme works straight away with Shopify without having to copy-and-past any code. Simply, press the try for free button above.
We created a poll with Feedme in 39 seconds:
It’s important to plan your poll before you rush and create one.
Step 2 - define a single business goal
A good goal is essential because it’ll determine the questions you ask. Without a clear goal, you’ll struggle and receive less responses with actionable insights, ask the wrong questions and ask questions at the wrong time. But ultimately, you’ll start to annoy visitors.
In this article our goal is to increase conversions.
Step 3 - think of questions to ask
Asking the right questions is pivotal to the success of a poll. The general rule is to ask questions that are relevant, timely and as specific as possible. The specificity of a question generally correlates with the current information you have to hand. It’s much better to create lots of specific polls than a single generic poll. It’s a similar principle to advertising. For example, if you already know from Google Analytics people are dropping off at a certain page, it’s much better to have a poll on that page about that page, rather than a poll on every page.
To get started see the section above ‘What questions to ask’ for a starting point. I recommend customising these questions so they better relate your your store.
Step 4 - choose when and where to ask the question
I recently visited a store that asked me about their product range as soon as I landed on their homepage - I hadn’t had time to look yet! This shows the importance of showing the poll in the right location and at the right time. A better time to show the poll would have been when I clicked the back button after scrolling through their collection pages.
In Feedme creating a poll that trigger on exit is simple. Choose a page (in this case a collection page) and then set the trigger to exit intent:
You shouldn't ask a visitor on the homepage why they didn’t buy from you today. But you should ask someone who goes to leave from the checkout page.
Step 5 - Wait for responses
How to increase response rate
Asking a single open-ended question immediately to a visitor can be daunting and cause people to skip the poll. This is a shame since open-ended questions are where the most insight can be extracted.
Feedme manages to combat this by taking advantage of a psychological principle known as commitment. This helps to increase the number of responses. This is how it works: Feedme asks an easy question to begin with, like a yes or no . Then Feedme asks an open-ended follow-up question to get a more detailed response. By asking a simple question first, respondents feel more compelled to answer the second question because they feel committed. In fact, structuring a question like this with Feedme can increase the number of response you receive by 500%.
How you word your questions can also have an impact. For example, changing “Why did you cancel?” to “What made you cancel?” could double the number of responses you receive. Unfortunately, there's no clear rule on how to phrase questions and it takes a bit of experimentation. However, unless you're chasing high response rates, you should receive enough responses without spending time optimising questions.
Hopefully, you now understand how fundamental polls are to increase conversions. But, make no mistake: just creating a poll with Feedme isn't enough. Take 5 minutes to create a plan for each poll and remember:
- to have a distinct goal
- don't let your preconceptions influence your question
- make sure you have enough responses to paint a clear picture before making changes
- experiment to increase response rate
- A/B test changes (optional but recommended)
If you would like to increase your conversions, you can try Feedme for free now.