I work at a retail company and am about to start a redesign project - I'll be looking at checkout process.

I don't have any personas in place and the company doesn't have the money to invest on this. So, I'll:

1.take a look at customer feedback

2.speak to the Call Centre asking about customer issues around the checkout.

3.speak to our Web Analyst and look at Google Analytics to understand Bounce Rate etc

I'm thinking that I'll create a number of provisional personas (probably 2), then conduct a Customer Journey Mapping exercise.

Does this approach make sense? Is there any other technique you would recommend?

  • I would also ask the call centre what the biggest problems in general are why users call. This might also give you some information what goes wrong, for instance a lot of users call regarding certain 'out of stock' items. This has nothing to do with your checkout, but on some websites you only see late in the checkout process that the item you selected is out of stock. This can cause a lot of bounces in the checkout, but the problem lies in the general design before checkout even begins.
    – Kevin M.
    Commented Mar 5, 2019 at 14:24
  • 1
    Is this a physical store or online checkout? Commented Mar 5, 2019 at 14:43

7 Answers 7


Speak to Marketing about who the company's target user is. Also, ask them who your main competitors are in the space and analyze how they execute their checkout process. Keep in mind they might be doing it wrong, but it's good to analyze yourself to see what would and wouldn't work for your case. If your Marketing team has an email list, ask for them to send a survey out (in exchange for some discount/coupon at the end of filling it out) that will ask them specifically about the checkout process. This should drum up enough tangible data to drive your redesign.

I don't believe that personas will help you any in regards to developing the best (most converting) checkout process. This will take a bit more trial and error. Speak with your Web Analyst to see if it's possible to A/B test different checkout processes simultaneously (at least 100 visits per example). This'll allow you to dial in the best one. Remember that it's critical to verify your assumptions.

Bonus tip: Be sure your Web team has implemented the ability to use ad or email retargeting for abandoned carts.


I think personas are not particularly useful for a checkout experience redesign. An user journey mapping could help if you discover that the problem is not on the checkout page per se, but somewhere between product display and checkout page (cart abandonment).

Now, barring the user journey need, you mention 3 tools you currently have at your disposal, and that's EXACTLY what you need: ask your customers, ask intermediaries and support and analyze user behavior with analytical tools.

While the creation of personas and user journeys can be useful, in my experience, they will not add any relevant data in an existing site where only one aspect will be modified. However, user surveys and data analysis are the tools that you can consider as indispensable for this task.

Therefore, I think that the 3 items that you mention as an approach to the problem in question is not only correct, but it will surely be more than enough. You could also add some visual user behavior tools such as Crazy Egg or Hotjar to create heat maps that give you some guidance on what users are doing and what they pay attention to when they are on the checkout page


You're definitely on the right track although I'm a little confused about not having the money for creating personas. I'm surprised that marketing doesn't have any information about its customer base? You could start gathering up information to fill that gap by putting up some simple surveys. That won't help you much for this project ... but it'll help for the next.

Your three points are spot on.

Keeping the Pareto distribution in mind I would focus on identifying those few problems that cause the greatest amount of issues.


Sounds a fairly good approach. GA is a great place to pick out bounce rates etc, call center staff will tell you what the pain points are and customer feedback will too.

You can create proto-personas to guide later stages of research. One thing you should be doing is building a first set of designs/IA to bring out to testing with users. This will drive needs and help you create better personas.

I take it there is a budget for user testing?

And use resources such as Baymard Institute https://baymard.com/blog

  • I'll probably be doing some guerilla research. Thanks for your comments
    – Caius
    Commented Mar 5, 2019 at 15:14

All of these are great things to work on.

Once you have personas using whatever research you have available, look at creating journey and experience maps (current state and future state) and Jobs to Be Done. What is your persona "hiring" your company to do, and what did they "fire" to do it?

Another approach to try is a heuristic evaluation of the current state. Look at every part of the experience and see how it maps against the 10 industry standard heuristics. You can then use this to prioritize things that are not usable.

  • Thanks for your comments Stacy. This is very helpful.
    – Caius
    Commented Mar 6, 2019 at 12:06

I actually disagree a little with looking at Marketing, though I still encourage you to contact the Call Center.

Reasoning: just because Marketing has an INTENDED user set in mind, that isn't always who ends up buying the products either, so use the information as who Marketing and management want their audience set to be, but take it with the proverbial grain of salt. (This doesn't always apply in retail, but I'm definitely thinking of other places where there's a lot of crossover in audience segmentation or there are unexpected audience sets.)

Also remember the adage that in customer service, many more people will remember something that they felt was a pain point than something they felt went okay or even well. I wouldn't exactly call it survivorship bias, either, but some of the data when you come across information from the call center is that it would be (to some extent) self-selecting: the user had enough of a problem to call in to the call center in the first place to speak their mind.


Some ideas for you;

  • Gather min 5 people you know and that havent used the site before, prepare some tasks, and run a small test
  • Number of steps doesnt play a big role in the checkout, clear structure does
  • Is good to communicate delivery time and cost early on cart overview page, as it is one of the common reasons users drop out in later steps
  • Transactional e-mails are not bad. If someone drops from checkout remind him in two weeks, ask if he needs help
  • Nicely design Thank you page, with proper thanks message can bring up satisfaction. You can use pictures, animations, or videos also.
  • Heatmap tracking tools provide additional data. They are free untill some limit, you will see where your users move they mouse, also video sessions are available.

Good luck


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.