Thursday, November 16, 2023

Can OP.GG Keep Up with Gamers? An Usibility Evaluation of OP.GG

By Tiancheng Qin

What is OP.GG? A Gamer's Helper

Every 'League of Legends' match is a new battle, demanding not just skill, but also knowledge. If you play it, you might use OP.GG to get better at the game. It's like a helper that gives you lots of information. You can find out which champions are strong right now, what items to buy, and how other players are doing. We've talked to real players and looked at how they use OP.GG to find out what works well and what doesn't.


Clarity or Confusion? Real User Feedback

The Ambiguous Search Box 

Issue

Fig 1 – search box in OP.GG’s homepage.
 When you go to OP.GG, the first thing you see is a search box on the main page (see Figure 1). But it's not clear what you should search for. Is it for player names or champion names? This confusion can waste your time, especially when you're in a hurry to choose your champion before a game.

Suggestion

Fig 2 - search box in OP.GG Esports’ home page

OP.GG could make this easier. For example, the esports stats page (see Figure 2) has a search box with clear labels. If OP.GG did the same, it would help a lot. They could add labels like "Search for Players" or "Search for Champions" to the search box. This will make players more aware of what the search box does to find the information they need faster.


The sorting function is not intuitive

Issue

Fig 3 – Tier table sorted by Tier (descending order) of champion statistics in OP.GG’s champions page
Fig 4 – Tier table sorted by Win Rate (descending order) of champion statistics in OP.GG’s champions page

Fig 5 – Tier table sorted by Win Rate (ascending order) of champion statistics in OP.GG’s champions page

Players often want to sort champions based on various criteria such as win rates or specific performance metrics to inform their game strategy (see Figures 3, 4, & 5). However, this crucial feature of OP.GG is not immediately noticeable. This lack of intuitive functionality can be a major obstacle, especially when time is of the essence.

Suggestion

Fig 6 – tier table of heroes sorted by pick rate (descending order) in HSReplay.net

OP.GG can make sorting champion stats much simpler. They could look at how HSReplay.net does it (see Figure 6) and do something similar. For example, they can add clear signs or buttons to show you can sort by win rate or other stats. They could also add up and down arrows to show if the stats are from high to low or low to high. These changes would help players find the best champions faster and make better choices in their games.


lack of a 'back' button

Issue

Fig 7 – Champion page for Zed on OP.GG
Fig 8 – Zed’s champion page in OP.GG desktop apps. (image from op.gg apps, which can be downloaded here)

Have you ever clicked on a champion's page in OP.GG and then wanted to go back to where you were before (see Figure 7)? It can be annoying. On OP.GG, if you go to a champion's page, there's no easy way to return to the previous page. Even if you use OP.GG's desktop app, which has a back button (see Figure 8), it doesn't remember the filters or the page you were on before. This means you might have to start your search all over again. 

Suggestion 

OP.GG can make things better by adding a 'Back' button that remembers where you were. This button should not just take you back, but also keep the same settings or filters you had before. So, if you were looking at a list of champions with certain filters, and then clicked on one to see more, hitting 'Back' should take you to that list with your filters still there. This would save time and make the site easier to use.


Confusing Matchup Descriptions

Issue

Fig 9 – The list shows Zed's win rate and the number of games played against a given champion on Zed's counters

Fig 10 – A list of counter champions indicating which are easy and which are difficult for Zed to play against

Imagine you're on OP.GG looking at champion matchups (see Figures 9 & 10). It's not clear whether the win rate shown is for Zed playing against other characters or the other way around. Furthermore, the terms "weak against" and "strong against" are ambiguous as to who is weak or strong - Zed or the other character. This confusion can be a big problem. When you're planning for a match, you need to know clearly who is strong or weak against whom. If you get this wrong, you might pick a champion thinking they have an advantage when they actually don't. This could lead to a tough game where you're already at a disadvantage from the start.

Suggestion 

OP.GG can make this easier to understand. They should change the way they show who's strong or weak against whom. For example, they could use clearer terms like "Champion A is strong against Champion B" or "Champion B struggles against Champion A."This way, players can quickly see which champions are good or bad against others, helping them make better choices for their games.


No Help with Spelling Mistakes

Issue


Fig 11 – search box at the top of Champions pages

When you use the search boxes on OP.GG, especially on the Champions pages, there's a problem (see Figure 11). If you make a spelling mistake, OP.GG doesn't help you fix it. This is really important in 'League of Legends' because the names of champions can be tricky. For example, if you type "Daruis" instead of "Darius" by accident, OP.GG won't correct you. This can mess up your search and waste your time, especially when you're trying to quickly find information about a champion.

Suggestion 
Fig 12 – search daruis in Google

OP.GG could improve by making their autocomplete feature more forgiving. For example, even if you type 'Daruis' instead of 'Darius', the system should still suggest 'Darius' as a possibility. This is similar to how Google works. When you make a typo in Google, it still understands what you mean and helps you find the right thing (see Figure 12). By making the autocomplete smarter and more tolerant of mistakes, OP.GG can help players quickly and accurately find the information they need, even if they're not sure about the exact spelling of a champion's name.


Understanding What Gamers Need

I talked to real 'League of Legends' players to see how they use OP.GG and what they think about it. This feedback is really important. Players told me that they need more than just numbers and statistics. They want OP.GG is easy to use and understand. When they're getting ready for a game, they often don't have much time. So, they need to find what they're looking for fast. My evaluation with gamers showed that a website that is easy and quick to use is just as important as having accurate stats. Players need a tool that helps them make decisions quickly, especially when they're about to play a game and every second counts.


Conclusion

For fans of 'League of Legends', OP.GG is a really useful site. It's full of data that can help you play better. However, our review found a few things that could be improved. The site could make it easier to find what you're looking for and move around. It could also help more with searches, like fixing typos. These improvements would make OP.GG is not just a place with lots of information, but a place where you can find exactly what you need quickly and easily. This is crucial because, in the game, fast and right decisions are key to winning. We hope that OP.GG listens to what players need and makes these changes. 

Acknowledge

This blog uses the Grammarly browser plugin to improve writing style.


Which is better? AppSmith or Retool - a walk through from UX perspective

 

Which is better? AppSmith or Retool - a walk through from UX perspective

By Haonan Chen

Appsmith and Retool, both renowned no-code platforms/tools, assert that they empower users with limited programming skills to develop applications and automate workflows. Typical applications that users can construct on these no-code platforms encompass:

  • App development, create common mobile/desktop/web applications
  • Workflow automation, automate repetive tasks and processes
  • Data integration, connect data sources to extr
    act or visualize wanted information
  • ...
In this blog, I will guide you through the CRUD tasks using the aforementioned two no-code platforms, illustrating the creation of a management tool for a blog website. This tool allows managers to perform Create, Retrieve, Update, and Delete (CRUD) operations on relevant data records. I will specifically focus on the data retrieval process as CRUD tasks share similarities in their implementation across these two platforms. I'll compare the functionalities that are crucial for non-technical users and highlight which platform offers a better user experience (UX) perspective.

Before delving into the development of CRUD applications on both platforms, consider watching the following video to gain a brief understanding of the capabilities of each platform if you're interested.

AppSmith in 100 seconds

How to Retool: A Getting Started Guide

Enviroment and Data

For both no-code platforms, I'm accessing their demo applications on the cloud. You can easily find and use their cloud applications by registering on their respective sites: Retool's at https://retool.com/ and AppSmith's at https://www.appsmith.com/.

To implement a management tool, it's essential to have our data prepared, which can be in the form of Excel sheets. In this case, I'm using structured data, as it more closely resembles what we might encounter in a real environment. If you're not well-versed in databases, Excel sheets can also serve as data sources. For both no-code platforms, there is no distinction in using Excel sheets or databases, except for importing the data.

For a blog website, common data entities include users, posts, comments, and likes. As a result, I've designed and populated four data tables with either random data or authentic data generated by the Large Language Model (LLM) - llama2-7b. If you're interested, you can find the codes here: https://github.com/bianyuanop/CSI5122-A2.

Below are the declarations of the data tables and their fields. The term "fields" here corresponds to the column headers in Excel, and "records" are akin to the rows in Excel.

Data tables

No-code App Development

Our objective is to establish a connection between our application and the current data source, followed by performing CRUD operations on it. This goal is divided into two sections: the first subsection involves development with Retool, while the second part entails development with AppSmith. The workflow is as follows:
  1. Connect data source
  2. Conduct CRUD tasks

Retool

Let's log into Retool and look around to see what we can do. 


Observing the layout, there are widgets and other tools panels on the left, with detailed functionalities/properties panels on the right. The main application is situated in the middle. The issue at hand is that upon opening the window, which takes up half of my screen, the interfaces appear too small for comfortable visibility on a 4K screen.

Connect data source 

Let's connect our data source. 


Confusingly, there aren't any other places in the project window that provide us with entries to add a data source except for adding from the "Code" panel.

Navigate through menu -> resources 

Furthermore, in comparison to the entry shown in the preceding image, it is evident that the data source configuration panel in the first .gif is not as comprehensive as the second one.

Configure & test data source but password field is missing

Therefore, when users attempt to add data sources that require a password for access, it proves ineffective and results in a connection error. Consequently, users must invest time searching for alternative entries that permit them to input the password for the database, leading to a wasteful exploration of the system.

Query and Displaying blogs

Instinctively, we would select the "Code" icon to initiate the data query, given that we previously used this panel in the preceding step to establish a connection with the data source. In the subsequent three images, I will demonstrate the process of adding a query and incorporating a "Table" widget to showcase the retrieved data using the GUI mode.

No select "Action type" is shown


Left "Action type" to be blank




Inserting a table to use content from "query1"

Nevertheless, we are unable to present the data. An error is reported, indicating that our query response is empty, even though we have left the "Action type" blank and received values after clicking the "Preview" button.

Manually writing SQL

Therefore, what I did was manually write SQL codes to accurately query the data, a task that is typically challenging for non-technical users.


Confiscated attributes panel


Refresh attributes


Upon completing the "query," our next step is to showcase the retrieved data using a "Table" widget. As we modify the data source of the table in the right panel, it becomes apparent that selecting the query doesn't automatically update the columns in the table. To address this, users must either manually edit each column using the provided panel or refresh the attributes by clicking the "refresh" icon. This could potentially lead to confusion, as users might associate the "refresh" icon with updating the response from the data source rather than refreshing the column names.

AppSmith

Let's sign in AppSmith and create a project. 
Clicking "New" to create an application

As observed, the interfaces are slightly larger compared to those in Retool, providing users with a clearer and more user-friendly experience.

Connect data source 

Including data sources in AppSmith has become simpler, as they've created a separate panel specifically for this purpose, as opposed to Retool, which combines "Query" and "Datasources."

Adding a new data source

Moreover, the pop-out panel includes all fields we need, including the password, along with the capability for testing configuration as Retool. 


The same panel after testing and configuring


Conduct CRUD tasks

Creating a CRUD application in AppSmith is an uncomplicated task, facilitated by its varied methods. Users can construct widgets directly from the widgets panel, akin to Retool, or utilize the Datasources panel, providing flexibility. Notably, AppSmith distinguishes itself with the powerful ability to swiftly produce complete sites using the "Pages -> generate from data" feature. With these approaches, users can have their applications and widgets ready in a matter of seconds.

Adding widget from data source

Generating CRUD template from data source


Problems 

The most of the problems I met happened on Retool, including 
  • Small interfaces on large screen, e.g. needing zooming in on 4k devices.
  • Mixing functionality of adding data sources with functionality of composing queries. 
  • No selecting function is found when composing a SQL query on GUI mode.
  • Widgets won't refresh themselves when data sources are changing.
  • Complicate widget interfaces 
While on AppSmith, seems all things are right and clear compared to Retool

Conclusion

In conclusion, the comparison between Retool and AppSmith reveals distinct advantages and challenges within the realm of no-code platforms. Retool, while powerful and versatile, presents certain drawbacks that might impede the user experience. Issues such as small interfaces on large screens, the mixing of data source addition and query composition functionalities, the absence of a selecting function in GUI mode for SQL queries, and complications with widget interfaces can be notable pain points for users.

On the other hand, AppSmith emerges as a more user-friendly alternative, offering a seamless and clear experience in comparison to Retool. Users may appreciate the absence of the issues encountered in Retool, finding a more intuitive and straightforward approach to building applications.

Ultimately, the choice between Retool and AppSmith depends on individual preferences, project requirements, and the specific challenges users are willing to navigate. However, it's recommended for a non-technical user to use AppSmith since it's simple and straight forward that serves as a really no-code platform. 


Usability Evaluation of canada.ca

By Nima Meghdadi, a Student of Computer Science at the University of Ottawa
Conducted usability testing and evaluation of Canada's government website, Canada.ca, employing the think-aloud technique. Proposed recommendations to enhance user experience and usability based on the evaluation findings. You can reach him at nmegh082@uottawa.ca

Introduction: This blog tells you how good canada.ca is for the people who use it, the official website of the Government of Canada that provides information and services to the public. It also offers some recommendations for improvement based on the findings.
The evaluation method employed for this website was the think-aloud technique, if you want to learn more about how this technique works, you can visit this website. But in this post, I will explain enough about this technique, which involves asking users to verbalize their thoughts and actions while performing tasks on the website. The users' screens were recorded and their feedback was collected to analyze the usability and user experience issues of canada.ca. If you want to learn more about this method, you can check out this website.
The website is supposed to help people learn about Canada's rules and do some information like taxes and immigration. But the website does not do a good job. People have a hard time finding what they need and doing what they want. People are not sure what to do next and they keep clicking on different things to see if they are useful or not.
People did not like the website because it had many problems. They could not tell what the website was for and how to use it. They also said that the website had too many words and not enough pictures, which made them feel bored and not want to stay. The website needs to be more clear, easy, and fun to keep people interested and happy.

1. The problem: The website made people mad when they tried to sign up. The page had too many ways (82) to sign up, and people could not find the best one for them. The website gave people too many choices that they did not need, and people often quit or messed up. Here in this picture I just show four of them.


here are just four of 82 sign up option

Recommendation: The website could make it easier for people to sign up. The page could have only one way to sign up and then let people choose their work later. Or it could have two ways for people from Canada and other countries. This would make signing up less confusing and harder.
The website has a problem that makes people unhappy. People have to make many accounts for different things, which is hard and takes a lot of time.
People can get used to this problem after they try a few times, but it is not a good thing.
Some people may think this problem is not a big deal, because some accounts are not used often or only once a year. But some accounts are used a lot. This problem makes it hard for people to remember which account they need for each thing.

2. The problem: The first page has too many words and no good groups. The website does not tell people what to do next, it just shows more words and sends people to another page. This makes people confused and angry. At the end, I said what things I used to test the website. If you have time, please try it yourself. You will feel unsure at every step.
some main categories

Recommendation: The website could make it easier for people to get a visa. The website could have a visa option that lets people pick the kind of visa they want. The website now has an immigration and citizenship option that makes it hard for people to find what they need.
This part of the website is good for some people and not for others. People who want to live in Canada may use this part a lot, but it may not matter for people who are already Canadian.
This part of the website is what the website is for, but it is not done well and it makes people confused. People have a hard time finding what they want and knowing how to get back to it.
The website does not help people find their way around. They often lose track of how they got to the thing they wanted.

3. The problemThe website has a thing that shows where you are, but it is not good and wrong. You cannot believe the thing to show where you went. If you want to see for yourself, you can click on this site and try to go back and forth. The thing that shows where you are on the website is called a breadcrumb bar and you can see it in the below picture. But if you click on the link, you cannot go back and forth based on the breadcrumb bar. The breadcrumb bar should show the steps you took on the website. And if you made a mistake, you should be able to go back with the breadcrumb bar.
here you can see the bar


Recommendation: The website could make the thing that shows where you are better. It should show where you went right. This would help you remember where you went and go back if you want.
People have this problem all the time on the website. This means that the problem was always there and everywhere. The problem made the whole website bad for the people and stopped them from doing what they wanted and needed.

4. The problem: The website had a search bar that let people look for what they wanted, and it should have given them good and helpful information. However the website's search bar often gave them bad and useless information, and they did not match what people wanted. This made people not able to do what they came for, as they spent a lot of time and work looking at search, and often did not find what they wanted. The people also felt unhappy and angry by the bad results, and they did not trust or believe the website. 
search bar
I searched for "import pet" here, but in the picture below, you can see the two best results for bringing a pet to Canada. But they are not good enough, because the website's search bar should show this link as the first result. This link has the most relevant and helpful information for people who want to import a pet.
search result of import pet


Recommendation: The website has a search bar that lets people see the information that other people look at a lot, which means they are good and helpful. This is a good thing that can help people find the best information on the website. But the website should also do other sort criteria to show the information and make it easy for people to find what they want. For example, the website could have groups, words, choices, stars, comments, or tips to sort and show information in a better and more personal way. The website could also use smart techniques to know what people want and mean and give them more right and helpful information. The website's search bar should show information in different ways, and not only by how many people look at them. The website could also let people make an account and change sorting techniques for what they like. This would let the website show information for the people based on what they like, want, and need. The website could use cookies, history, or questions to get and use the people's information and give them their own helpful information.

5. The problem: The website's colors did not make people happy, they thought they were boring and not fun. They said that the website should use more bright and different colors to make it more fun and nice. The website's look should show what it is and what it does, and make people want to see more. The people did not like the website's color behind everything, which was #fff (white). They said that this color was too shiny and hurt their eyes, making them feel bored and sleepy. The flag of Canada and the background color on this website are the same. It is hard to see the flag clearly. 
this is the main page

Recommendation: People said that the website should use a color that is not too bright and warm for the color behind everything, like light brown or yellow, to make it more cozy and friendly. The website's color should think about what the people like and how they see, and make the website easier to read and use.
The website should change and make things better, and make sure they are easy, simple, nice, and friendly. The website should also ask the people what they think, and use their ideas and likes to make it better. The website's goal should be to make a website that people like and are happy with, and that does what they want and need.

Method: Here I tell you what I did to test the website. Think-aloud is when evaluation takers show how they think when they read a text out loud. They say what they do to understand the text, like guessing, asking, summing up, or explaining. This way was tried on three people who were 23, 23, and 51 years old. All the people were given the same taskto do and asked to find the best way to do them. If they had any problems, the person who tested them gave them small tips. All the people knew about the website and what it was for. The website was tried by three people, two in person and one online. The people who were in person were given cards with tasksto do on the website, from easy to hard, to make them want to do them and see how they did. The person who was online was connected by Google Meet and got the tasks to do in the chat. Their screens were seen and saved by the person who tested them, who watched what they did and how they felt. The website trying ways were made to see how good the website was, how well it worked, and how happy the people were.
The people were asked to do five tasks on the website, about coming and working in Canada. The tasks were:
Task 1: Ask for a visa to visit.
Task 2: Find out how to get a permit to work after finishing school in Canada.
Task 3: Sign up for a permit to work anywhere and go to the last step before sending the request.
Task 4: Find out about the new rule about how many hours international students can work in Canada.
Task 5: Find out how to bring a dog to Canada.
The tasks were made to test the website's words, ways, and tasks, and to see how happy and sure the people were. The tasks were also good and real for the people who wanted to study, work, or visit Canada. The website tried different times for each person, depending on how fast and right they were. But the average time of the trying was one hour. This shows that the website's tasks were not too easy or too hard, but rather hard and fun for the people. The trying time also shows how big and hard the website was, and how much the people wanted to know. The trying time should be seen as a way to see how good the website was and how happy the people were.
Conclusion: The website has many helpful recommendations, but they are hard to find. People get frustrated and use Google instead. The website does not explain well, and people do not know what to do next.
AcknowledgmentI am grateful to the Bing AI tool for improving the grammar of all my sentences. The tool helped me to write more clearly, correctly, and professionally.

What are Feedly's problems?

By Ka Chun Sie

Feedly as an RSS Reader

Since you are browsing this blog, scroll down to the bottom and you will find the link "Subscribe to: Posts (Atom)". By clicking on it, you will be able to view all the simplified information in XML format under uottawacsi5122.blogspot.com.

Yes, this is RSS. If you have used tools such as Blogger, WordPress, Github Pages, etc. to set up your own blog site, or if you have ever been a Google Reader user, then you must be familiar with RSS. RSS is a web feed that allows users and applications to access updates to websites in a standardized, computer-readable format. RSS works as a publishing protocol as millions of websites release their latest content and push them to users in the form of RSS feed. Technically speaking, there exist different web feed specifications such as RSS 1.0, RSS 2.0, RSS 3.0, Atom, JSON Feed, etc., however, since RSS has become so popular as a synonym for web feed, all of them can be regarded as RSS and are represented by the classic icon as .

In addition to clicking RSS's link or icon to get the feeds, users need RSS readers to aggregate and convert feeds to readable content from different websites. When users copy the RSS addresses and add them to the RSS reader, the RSS reader automatically grabs the latest contents and pushes them to the user. By using only one RSS reader instead of browsing different websites or applications, users can track the latest content from all subscribed websites. All in one is the essence of RSS and RSS readers. It may sound similar to email newsletter service, but RSS readers are better at managing thousands of subscriptions rather than cluttering up your inbox. Moreover, today's RSS readers have more features and a better reading experience than newsletters. To convert newsletters into RSS feeds, people can use this site.

You might probably heard "RSS is dead." before, just like this blog stated. It's true that RSS's usage has declined over the years. Some attribute the death of RSS to the rise of social media apps such as Twitter and Facebook, while others believe it is due to the shutdown of Google Reader and the lack of good enough alternatives, or, as some say, it's a result of Internet giants attempting to build more closed and centralized ecosystems for the sake of profit. But wait a minute, is RSS really dead?

I don't think so. As this answer says, there are still enough users, RSS readers, and content platforms to keep maintaining RSS. Feedly, the largest RSS reader on the market is the top successor to Google Reader. After the announcement of powering down Google Reader, Feedly claimed more than 500,000 Google Reader users joined the Feedly community over the last 48 hours. As of 2018, Feedly had 14 million users, making it the most popular RSS reader. You see, many "stubborn" people still use RSS and RSS readers from the heyday of RSS and Google Reader until the present. These devoted and sticky users are the core user base of RSS readers like Feedly.

In fact, I've never used Google Reader and experienced the early days of the Internet before, but started using RSS reader in the era of social media. In order not to be trapped by the filter bubbles created by recommendation algorithms, to gain more control of privacy and information, and to have a transparent and pure reading experience, I find Feedly as an RSS reader is a great tool for me. I'm sure lots of people like me searched and found Feedly when looking for an efficient way to manage information subscriptions to get rid of social networks. These people became new RSS readers' users and regarded them as productivity tools. They are a significant part of Feedly's incremental users.

Main UX Problems

Let's further discuss Feedly's user experience(UX) problems as an RSS reader. I chose to use Feedly as my RSS reader because of its large user base, good reputation, and clean UI. With a simple UI, Google Reader-like layout, rich features, and lower learning costs, Feedly is able to meet users' most demands and expectations. However, over the years of use, I have found some UX issues, some of which have been further exposed after heuristic evaluation. I need to mention that I have been using Feedly Web with the free plan, but my review of Feedly will cover some Pro/Pro+/Enterprise Plan features.

Specifically, I list some of Feedly's main UX problems as follows.

Unable to Access All Topics & Top Feeds

Users can explore different topics, search hashtags directly, and browse different RSS feeds at Find Websites. When exploring this page, users can see four featured topics, Tech, Cyber Security, Marketing, and Business, with background images, while scrolling down reveals four topic categories for Industries, Trends, Skills, and Fun. Generally, users will find the RSS feeds they wish to subscribe to under the interesting topic.

However, perhaps there are users like me who would like to browse all topics via the pagination component, as well as find the ranking of RSS feeds under each topic. Yes, Feedly has both of these pages, but unfortunately, they are not obvious. All Topics and Top Feeds are two hidden pages, which can only be accessed through the routes https://feedly.com/i/all/ and https://feedly.com/i/top/. It's really weird that users have to remember both paths instead of visiting both pages by clicking on any link. Adding hypertext links or buttons to All Topics and Top Feeds below the four featured topics in the Follow Websites page may be a good idea to enable users not to recall but recognize.

Difficulty in Creating AI feed

With Feedly AI, users can track specific topics and trends. This is a Pro+ feature but users in the free plan can experience how to construct a query (AI feed). When a user tries to create an AI Feed, there are too many options in the search box. Although common combinations of conditions are shown below as examples, it is difficult for users to fully understand the meaning of each option without consulting the documentation.

When users visit the documentation they will find that Feedly defines a very large number of AI models to flag instances and actions. These AI models are options in the query, and their permutations and combinations are too complicated for users. Although the user is supposed to filter the feed by AND/OR in the query, too many options will only make it more difficult to use and reduce the frequency of users' usage. In addition, when users select some options, they will find that the icons of some options have no relation to their actual meaning, and many different options often use duplicate icons. To make it less difficult to create an AI feed, Feedly should simplify the options and replace the icons to make them concrete and familiar.

Bad Internationalization

Users can select different languages at Find Websites as well as search or read feeds in languages other than English in Feedly. But that's it, Feedly's i18n just has this stuff.

Feedly is not able to change the language globally or translate the article, like its competitor Inoreader. This means that the users have to know English to understand the meaning of each setting. In Find Websites, users will also find that the layout of the default English interface is different from other languages. Meanwhile, topics are shown without prefaced hash symbols by default, while choosing other languages, topics are shown in the form of hashtags. The lack of interface language options and inconsistency in design can be frustrating for international users.


What's more? For right-to-left languages, Feedly does make an adaption on i18n but leaves the HTML exposed to users. The exposure of the code will likely overwhelm the user and make them wonder if the current state is correct.


Documentation & Navigation are in Chaos

Feedly's documentation, tutorials, and case studies are spread out on different pages, and even though these pages can jump to each other, users have to spend more time searching for answers. For example, if users can not find the definition of the AI Feed search options in the Documentation, they have to click the Blog & Case Studies, then click the drop-down menu in the upper right corner, and finally click Feedly AI to find them. Feedly should definitely organize all tutorials, documentation, case studies, new feature descriptions, etc. into a single Help & Documentation page with good search capabilities and categorization to better help users.


About the navigation and page organization, it's also suck. Create AI Feed and Power Search are two completely similar features and reuse the same components, so why doesn't Feedly merge them into one page?

What's even more confusing is that there are two sets of options linked to the same pages in the pop-up menu after clicking the avatar in the corner. When users click Support, the main page shows the promotion of paid plans. Users must click the button in the bottom right corner to open the chat box. Perhaps Feedly should remove the duplicated options and simplify the features & pages, so as to keep itself simple and clean.

Imperfect Read & Unread

I think Feedly is doing well in the aspect of reading, but there are still a few problems. After subscribing to many RSS feeds, users may not read all of them, so Read/Unread flags will help users manage what they want to read. However, the color contrast between Read (Grey Title) and Unread (Black Title) is not obvious enough.

When the user selects Magazine View and Cards View to display the feed, the checkmark can flag the content as Read or Unread, but under Title-Only View, the same checkmark can only flag the content as Read.

Meanwhile, in any feed, to mark articles as Read in batches, the user has only three options to choose from without any customization option.


I think Feedly should improve these Read/Unread problems and thus make users' reading experience smoother.

What else

Target Audience

I believe there is also a business strategy issue with Feedly, which is related to the target audience. Feedly aims to retain its early RSS enthusiasts inherited from the Google Reader era, attract corporate users through industry insights features, invite information security professionals by offering vulnerability tracking, and even establish specific biopharma options to cater to biomedical researchers. It's obvious that Feedly is eager for various customer segments and is making efforts to downplay its identity as an RSS reader and transit into an insights-tracking tool.

The result of trying to develop specialized features for a wide target audience is that Feedly has to bundle these features together within the same interface and segregate users' access to these features based on different paid plans. I think mining business users is a good profit driver. But trying to serve all types of users at the same time is difficult and not a wise choice for Feedly. Perhaps narrowing down the customer base and customizing different versions for distinct user groups is a better choice (Like Personal/Industry/Academic).

User Expectation

RSS readers should provide users with complete control over feed management. Users should be able to group feeds by specific topics, establish personalized folders, archive read articles, and determine different settings of reading. My basic expectation of an RSS reader is to have a modern and minimalist UI, making it convenient to explore RSS feeds and read articles, enabling my subscriptions to be all in one. For the requirements mentioned above, I think Feedly succeeded in satisfying them.

Some RSS fundamentalists have a strong preference for absolute control, de-advertising, and anti-algorithms. However, inevitably, there are advertisement and basic ranking algorithms in Feedly, and Feedly has been marketing its AI features. While I don't care that Feedly is calculating popularity for articles and embedding feedback mechanisms on AI recognition in articles, those who expect more from Feedly may be displeased with it. Besides, I'm just a free plan user looking for an effective content reader, so I can tolerate Feedly's constant reminders to upgrade my plan to enjoy more advanced features. But I also agree with others about Feedly not finding a balance between driving income and enhancing user experience. The complex new features, pop-ups, ads, and reminders of upgrades indeed ruin the original pure reading experience.

I'll probably keep using Feedly as my RSS reader. Feedly's basic features still meet my needs, and its UX problems don't affect me seriously for the time being. However, Feedly still needs to address those issues in order to meet the expectations of its paying users and prevent user attrition.