You Have 5 Minutes To Test A Web Application: How Do You Do It?

Questions
#1

I was speaking to a software tester recently who was asked by their manager:

If you had 5 minutes to test a web based application, what would you test and why?

Now, before people start saying “What do you mean by a web based application?” “What does it do?” “Why do I only have 5 minutes to test it?” etc, note that this is third hand.

The person who was asked this gave an answer suggesting to start with a login page and potential issues that might arise around that because they thought that if users can’t log in to an application then it’s a show stopper.

They were worried that this wasn’t a software tester/QA mindset.

In relation to this persons login answer, I suggested that they take a look at Test Ideas For A Login Screen. Or some other areas in the How To Test category as it’s not a bad approach to flesh out. (I also assured them that this is a pretty good software tester/QA mindset).

I’ve come to learn that every tester will approach the same challenge in a completely different way so here is my answer for the 5 minute challenge:

I have Spectrum on Chrome. As soon as I open the application, I would get that running. I would also start up the screen reader that I use. If I only have 5 minutes I’m going to whack a lot into it!

I would then open developer tools on my browser and note any strangeness there, are there cookies I can edit, does that matter? You could spend a while on this depending on the site so I’d try to time box that if I could.

Is the site on http or https? Does that matter?

While I’m doing all of this I’m also thinking: How intuitive is this site? Do I feel like I know what I’m doing?

I think this links quite well to something that @simon_tomes shared on the Club Take the 5 minute challenge! What are Risks? But I was wondering, if you were in this testers position what answer might you give to that manager? How would you spend the 5 minute window you have testing that application?

2 Likes
#2

Run the site through an automated link checker, highlighting any clearly broken elements. (hopefully not a big site so it’ll be done in less than 5 minutes!). Maybe try and run through the OpenVAS tool set if time.

While the link checker is busy, run through any login elements manually to test those with a limited set of tests (right, wrong, obscure entries)

Review the link check results and if possible target testing at the clear issue points.

Provide dip checks into the remaining site.

5 minutes isnt long, so it’s a very tricky question @heather_reid :wink:

2 Likes
#3

Run it through: GTmetrix* for web speed analysis https://gtmetrix.com

*: as recommended by the podcast PerfBytes:

2 Likes
#4

Bearing in mind that the answer to this is massively contextual and the tester in question would have [some] understanding of their business already I think I’d ask, ‘what is the key goal of the application and the biggest risk it has/introduces (asking them to speak quickly!)?’ From that answer I’d target the value/risk described first then probably open WAVE to have a quick scan for some fast accessibility feedback.

Heather, I love the idea of quickly opening tools or Andy’s using tools to generate some quick feedback.

2 Likes
#5

Given that 5 minutes isn’t even enough time to get a first impression, I would first ask for more time and details.

Since this is a theoretical question, I will assume the answer to more time and details is “no”.

Then I would look at the site for areas which are obvious for data entries like links, forms, menus. I would click on everything as quickly as I can.

While I’m clicking, I’m looking at everything with the questions, “Does this make sense?” If I have only 5 minutes, everything has to be intuitive, otherwise there is an issue.

By the time I’m done clicking, my time is probably up. But if there happened to be a few extra seconds, I would start filling in form data, such as search.

I’m still not looking for bugs or performance problems (unless they’re apparent through my random clicking), but only trying to get an idea of what needs to be more carefully examined.

2 Likes
#6

I think there is a few options for this, all based on circumstance:

  1. I think this could very much depend on your previous knowledge of the web app. if you know it well, you would know where the key features and business value lies and aim to ensure they are smoke tested (as full testing is not possible in 5 minutes), and then any spare time, move down the priority order. if you don’t know it and this is the first time you see it, ask what it is meant to do, and aim to get at least the key journey tested.
  2. If it is desperately needed to be tested in 5 minutes and it is possible, gather everyone available (the whole team, whether dev/tester/manager etc…) and assign critical steps to each person, either by section or step in the journey, having at least one member go through the happy paths through the journey.
  3. Refuse/Plead for more time!

I do like the idea of using tools to get the most value out of your time, but need to ensure they are the right tools for the right circumstance.

3 Likes
#7

First I will test the main functions the main requirement of user and than check those one recently fixed by developers …

#8

This sounds like an interview question. The manager is looking to see how you approach testing. Are you going to go through the application systematically? I have seen testers who run tests in the order they were documented, over and over and over, for each test candidate. Not very efficient.

Now in this case is sounds like an interview question but you point out this is someone’s existing manager asking the question. So we might be able to assume the tester is familiar with the application. So the manager is trying to see if his employee understands what is most important. Are they prioritizing what they test? Are they going to find the most critical defects first?

Alternatively, as a manager I might want to make sure different testers will find different things. If we start having too much group-think I’ll need to deal with that.

As someone who used to ask these sort of questions, I also look for different approaches to testing. Is there something they do what I have never seen someone do before?

For example, when I first started testing web applications I realized that things could be happening on the backend and I wouldn’t see the ramifications for many steps on the frontend. So leaving a terminal open and tailing the logs on the application server while I interacted with the web frontend let me know when the defect actually occurred. I didn’t see many other people doing that 15 years ago.

Open the developer tools and see if the app is silently throwing javascript errors or ignoring when resources or links are broken.

A negative for this sort of question is that the manager might have assumptions. If they think performance is the most important thing, because so many tester miss performance defects, they might reward someone who talks about performance testing even if there are no real performance issue.

3 Likes
#9

Will use the following tools

  • Webspeed
  • Run Through Zap Proxy
  • Pairwise to get better coverage in limited time
1 Like
#10

I always wanted an excuse to try this out…

1 Like
#11

First of all, its not possible to test any application in 5 minutes unless it is a static single page application or so. But, if we need to test in 5 minutes then we will test one of the below approach that software testing services primarily focus:

  • Functional testing: Functional testing is the very important section of the application which needs to be tested on areas (URL testing, Login page testing, Navigation among main pages).
  • User creation: New user should successfully created on the application.
  • Interface testing: To check that interfaces in the application is consistent (General appearance (Look and feel), Color and frames etc).
  • Content checking: To check the content should be logical and understandable.
  • Compatibility Checking: Main objective for this to check that application should be compatible with important Web browser and do not gives browser specific errors.

These are the basic areas needs to be covered while testing a web application in very short time. Hope this information will be helpful for you.

#12

I disagree. You can always test something. You might not be able to test it as much as you would like to or feel comfortable with but you absolutely can do some form of testing in 5 minutes.

1 Like
#13

Yes, you are right. We can test something in 5 minutes. But. in my actual post, I was trying to say its not test full application in 5 minutes.

I am agree with your thought. Thanks

#14

Lot of context missing here to give a reasonable thoughtful answer.
Initial thoughts:
Is there a user/login flow? Test account creation/recovery.
Is it more of a e-commerce site? I am searching for items and loading up the shopping cart and going through purchase flows.

Perhaps pulling up browser tools to restrict my screen size to check for responsiveness.
If relevant I may look at search engine optimizations (meta tags, title tags, etc).

5 minutes is tough to do without more context about the site but those are some quick things to start any site with.

1 Like