Feature addition to Instagram to solve for a common User dilemma.

INSTAGRAM

For work inquiries please feel free

to contact me

Let’s work together !

Scope:

Adding a Feature

Researcher / Visual Designer / UI / UX Designer

Researcher / Visual Designer / UI / UX Designer

Role:

Figma / Survey Monkey / Chat GPT / MidJourney / Maze , Instagram

Tools:

3 weeks

Duration:

Project Overview

Background

Instagram is a social media platform that allows users to share photos, videos, and stories with their followers. It was founded in 2010 and acquired by Facebook in 2012.


The platform is primarily mobile-based, with a user-friendly interface that allows users to easily navigate through their feeds and explore new content. It also has a search feature that enables users to discover new accounts and hashtags related to their interests.

Instagram has become a popular platform for individuals and businesses to promote their products and services, build brand awareness, and engage with their audience. It has also become a hub for influencers, who leverage their large following to partner with brands and earn income through sponsored posts.

However, users are finding difficulty navigating their Home Page and their Explore Page, as they may be recommended content they have no interest in, may deem inappropriate, or just don't relate to their personal interests.


The applications algorithmic feed can fail to show users the content they are interested in, leading to frustration and disengagement. Additionally some of the advertising and sponsored content can be intrusive and irrelevant, leading to a negative user experience and decreased engagement.

Objective

  • Develop a way to help Instagram Users have more control over the

content that appears in their Home Page and Explore Page feeds to give them a more curated and safer content selection.

Define

Research Plan

To begin the project, we performed competitive and comparative analyses, and market trends. With this data we can evaluate Instagram’s strengths and weaknesses regarding content moderation and content suggestions.

Research Goals

  • Understand User sentiments regarding recommended content on the Home Page and the Explore Page.


  • Develop a curation process that allows Users to customize their Home Page and Explore Page based on promotion rather than restriction.


  • Develop a method to make Instagram use more enjoyable for Users who want to see content that is more relevant to their interests.

Methodologies

Competitive Analysis.


User Survey - Gain greater insight into User reasons for using Instagram and current issues they may have with its algorithm.


User Testing via Maze - Review site iterations to determine success in content curation.

Secondary Research

Market Research

Background: Instagram. Current Research and Current Content Controls


In 2021 Instagram introduced a new toggle called “Sensitive Content Control” that allows anyone to screen posts that they think could be offensive, hiding them from the Explore tab. The new feature appears in the settings menu and lets users choose to either allow more content that could be “upsetting or offensive,” limit that content or “limit even more.”

In 2022 Instagram expanded that setting, letting users turn off that content in recommendations throughout the app. Rather than letting users mute certain content topics, Instagram’s controls only have three settings, one that shows you less of this bucket of content, the standard setting and an option to see more sensitive content. Instagram users under the age of 18 won’t be able to opt for the latter setting.


Per Instagram the content controlled includes:


Content that may depict violence, such as people fighting. (We remove graphically violent content.)

Content that may be sexually explicit or suggestive, such as pictures of people in see-through clothing. (We remove content that contains adult nudity or sexual activity.)

Content that promotes the use of certain regulated products, such as tobacco or vaping products, adult products and services, or pharmaceutical drugs. (We remove content that attempts to sell or trade most regulated goods.)

Content that may promote or depict cosmetic procedures.

Content that may be attempting to sell products or services based on health-related claims, such as promoting a supplement to help a person lose weight.”


In 2023 Instagram is introducing a handful of features designed to give people more control over the photos and videos they see on the platform. Specifically, you can now tell the app the types of content you don’t want to be recommended to you. Starting with the Explore tab, you can select multiple tiles and tap “Not Interested” to shape Instagram’s content algorithm. Doing so will also affect what you see when you search for posts. Additionally, in an expansion of a feature that was already available for comments and direct messages, you can list specific words, hashtags and emoji you want Instagram to filter for when recommending content. You can access that tool through the “Hidden Words” section of the app's privacy settings.

Facebook:

Tumblr:

Twitter:

TikTok:

Pinterest:

Competitive Research

Competitive Research

Discussion

Application Propositions

This year we will be able to determine how much clickbait or graphic content we want to scroll through, Facebook's parent company Meta Platforms told USA TODAY. Soon we will be able to decide if we want to see less of that content in our feeds by choosing “reduce more.” If we want to see more “low quality” or “unoriginal ”content in our feeds, we can select “don’t reduce” content.


That option will not be available for “sensitive” content, according to Facebook. In the United States, Facebook users will also be able to decide if they want to see less or more content that was fact-checked by one of the company’s third-party fact-checking partners. The fact-checked content will still be labeled, Facebook said. Users will be able to decide whether certain content is “demoted,” and they will also have greater control over fact-checked content, though the fact checks cannot be turned off entirely.

Tags come into play when a user searches for a term or follows a tag actively. The latter allows you to keep up with one tag or another, and Tumblr will include these tagged posts in your feed and the blogs you follow.

You can do the same with post content, as well—the difference is that a post content filter indexes the entire post for your flagged tags and keywords, not just the tags attached. This is the way to go if you would like to eliminate the topic from your Tumblr feed with the highest possible degree of certainty.


Community make the site’s moderation tools more granular. Tumblr is asking users to tag posts that depict or reference three categories of content: drug and alcohol addiction, violence “similar to what you might see in an age-restricted movie,” and “sexual themes” including “erotic writing or imagery.” Adding any of them will add a “mature” label to the post. You can also add a “mature” label for anything you think is sensitive but doesn’t fit these categories.Once tagged, mature posts will either be blurred or hidden altogether on a viewer’s dashboard, depending on their own content settings. Those settings also let you turn off both options and see the posts normally — unless your birthday indicates you’re under 18, in which case all mature posts are hidden and the settings locked. The filtering is opt-out, so if you want to view mature content, you’ll need to manually change your settings.

Twitter blocks some tweets with a “potentially sensitive content” warning, including when you’re retweeting them. You can disable this warning—even on an iPhone or iPad. You can also disable sensitive content warnings on your tweets without making your account private. By default, Twitter restricts this media with a warning like, “This media may contain sensitive material,” “This profile may include potentially sensitive content,” or “The following media includes potentially sensitive content.”

TikTok has a reputation for showing you exactly what you want to see, which can keep you scrolling for hours. It’s thanks to the app’s powerful algorithm, that learns your behaviors in a way no other social media platform does. According to The New York Times, the equation for scoring videos and including them in your recommendations is a combination of “likes, comments, and playtime, as well as an indication that the video has been played.” In 2021 TikTok added new warnings to its in-app search that will alert users when results may include “distressing content.” The app has employed “sensitive content” warnings on individual videos since 2020, but the updated alerts will appear in search results for terms that could include such content. In 2022, TikTok announced the rollout of three new content curation and control features: content filters, age restrictions on content, and limiting content recommendations for certain topics. TikTok is adding a tool “people can use to automatically filter out videos with words or hashtags they don’t want to see from their For You or Following feeds.” TikTok also limited content recommendations for certain topics “that may be fine as a single video but potentially problematic if viewed repeatedly.” Examples given of such topics included “dieting” and “extreme fitness.” So basically, you may still see such content recommendations but they’ll show up less often to make room for other topics so that you aren’t watching the same type of video over and over again about topics that could be harmful the more you watch them.

In 2023 Pinterest announced further safety measures for its platform, including a new set of parental controls and updated age verification policies, among other things. However, the company also said that it would soon re-open some of its previously locked-down features for teens to allow them to once again message and share content with others after approving contacts. For starters, it says it will expand its age verification process. By the end of this month, if someone who entered their age as under 18 tries to edit their date of birth on the Pinterest app, the company will require them to send additional information to its third-party age verification partner. This process includes sending a government ID or birth certificate and may also require the users to take a selfie for an ID photo.

Leveraging algorithms to automatically detect adult content, hateful activities, medical misinformation, drugs, graphic violence, and more before it’s reported, the company says that policy-violating reports per impression have declined by 52% since fall 2019, when the technologies were first introduced. And reports for self-harm content have decreased by 80% since April 2019.

These applications all share the ability to restrict sensitive content based on user preference, and can present content based on a series of restrictions the User has opted to apply. What I notice is there is no ability to specify which content you want to see promoted or the accounts you wish to see boosted within your feed. The algorithm is based around retracting and restricting content rather than adding more prompts to educate the AI. The type of content curation I am attempting to implement revolves around promoted accounts and hashtags a User wishes to see more of. Explore Pages can be shuffled around like tiles to promote the latest content from a User’s favorite accounts and feature hashtags that the User has curated themselves. Algorithmic suggestions will continue to be made.

Priority Ranking: Instagram can provide users with the option to rank certain accounts or hashtags as a priority, which would ensure that their content appears at the top of their feed.


Improved Search: Instagram can improve its search functionality by providing more relevant results to users based on their interests. This could include options to filter search results by category or by date.


User Feedback: Instagram can allow users to provide feedback on the content they see on their feed. This could include options to rate posts or accounts based on how relevant they are to the user's interests.

Primary Research

User Testing

Trends

Additional Responses

I created an online User survey via Survey Monkey, which garnered 8 responses.


The questions were as follows:


How often to do you find the Explore Page makes relevant content recommendations ?

How often do you find that the ads marketed to you are relevant to your interests ?

How often do you find that the recommended accounts are relevant to your interests ?

Do you find that you are recommended sensitive or suggestive content (Politics, Violence, Sexually Explicit Material) ?

If the answer to the above question is (Yes) does it negatively impact your use of Instagram ?

Do you apply restrictions or filters to what appears on your Instagram Hone and Explore Page feeds ?

Overall, are you satisfied with the type of content you are recommended ?

Do you feel Instagram can do a better job of content regulation and recommendation ?

Do you feel your favorite accounts and content you’re most interested in may not always be prioritized in your feed?

If Instagram were to provide feedback options to rate posts or accounts based on how relevant they are to the user’s interests, would you utilize this system ?

Responses highlighted the ads targeted to consumers are hits-or-misses. The most typical responses for this question were Frequently (37.50%) and Occasionally (37.50%).

Here we see half our users are not being recommended accounts that offer the greatest relevance to their interests ? Why ? Could account prioritization and AI responses based on those prompts manifest better recommendations ?

Users responded evenly to the question regarding the type of content directed their way. Half responded Yes, half responded No.


This only makes sense due to me knowing the kind of content the users engage in. Had I not known the content my users view, this data would lead me to want further inquiry.

Overall users are satisfied with the type of content they are recommended. But we will notice an interesting response that follows..

The majority of users although found satisfaction in the content recommended, felt that:

Instagram could can do a better job with content recommendation

Feel their favorite accounts are not being prioritized properly

They would utilize a rating system to prioritize the types of posts and accounts recommended to them.


It’s evident the the algorithm could use work…

I asked some users who were willing to provide additional feedback to describe their experience with Instagram. Here are their responses:


User 1:


“General feedback is sometimes Instagram will recommend in explore page content I didn’t otherwise show any previous interest in. If I look at one politically charged post, it thinks I want my entire explore page to be politically motivated. This concerns me because I think there needs to be a way to discern interest or just simply watching.”


User 2:


“On this account (personal) I tend to like the suggestions. On my music account (professional), somehow the algorithm thinks I like reels featuring individuals with deformities, medical oddities, and graphic content.”


User 3:


“The Home Page and Explore Pages recommend accounts irrelevant to my interests. The Explore Page also has a habit of recommending content that feels repetitive and I’m starting to see less original posts. The Explore Page isn’t feeling as inspiring as it used to, and I don’t even look at the Explore Page because it’s feeling bland to me.”


User 4:


“The Explore Page keeps pushing stupid stuff like violence, silliness, and “sexy stuff”. All of the good, talented content, the educational stuff isn’t really showing up on my feed, it feels like it’s censored. I want to see more art, fitness, and holistic health, less explicit material”


Ideate

User Persona

With the data collected from my research and survey, I developed 2 Persona’s to define the perspective of User’s dissatisfied with some of the content recommendation aspects of Instagram.

Project Goal

With the user assessments and an evaluation of the businesses needs, I was able to define goals for the project which it will also be a guide to the project.

Account Prioritization

Preferred Content

Reviewing the data, I developed two new features :


  • Account Prioritization


  • Preferred Content

User responses indicated that preferred accounts were not being prioritized properly.


When asked “ Do you feel your favorite accounts and content you are most interested in may not always be prioritized in your feed ? “ 62.50% of respondents agreed, and 87.50 % of respondents agreed that Instagram could do a better job of content recommendations. Again 87.50% agreed they would use a ranking system that prioritize desired accounts.


So, I decided to develop such a feature.


Account Prioritization has users go through the list of of accounts they follow and place them within the “Prioritized Account: category. When categorized the algorithm will not only promote these accounts within your feed, and put their content ahead of the content shared by other accounts, but it will make stronger content and account recommendations based off of these “Prioritized Accounts”. These profiles selected can be changed at any time, and it will create a chain-effect that will alert the AI to modify the content recommended now that alterations have been made to this category.


Users would be able to promote the content they want to see at the top of their feeds, whether it’s based on a system of prioritizing friends over any other account, or the preferences are genre-based for users who may want to see their favorite artists promoted over any other type of content they see on their feed.

So we go back to our statistical results regarding content prioritization, content recommendations, and ranking systems. I also reflected on User commentary, and statements such as:


“General feedback is sometimes Instagram will recommend in explore page content I didn’t otherwise show any previous interest in. If I look at one politically charged post, it thinks I want my entire explore page to be politically motivated. This concerns me because I think there needs to be a way to discern interest or just simply watching.”


“The Home Page and Explore Pages recommend accounts irrelevant to my interests. The Explore Page also has a habit of recommending content that feels repetitive and I’m starting to see less original posts. The Explore Page isn’t feeling as inspiring as it used to, and I don’t even look at the Explore Page because it’s feeling bland to me.”


Preferred Content is the ability to look at the tags pooled that are generated as a result of the media you interact with and the accounts that you follow. Traditionally, you can block or exclude tags that you no longer wish to see, and while with this feature you can continue that practice, with Prioritized Content, you can search tags to now include within your content recommendation for the AI to now include in the kind of content you are recommended and the tags that show up in your Explore Page. If you wish to include Tesla as a tag that you are recommended, the AI will begin to show that type of content and now make media and account recommendations that highlight ‘Tesla” as a focus of its content.


So how does this all look and function ?

Site Map

With the help of product roadmap, I was able to visualize primary features’ relationship between content along and arrange the hierarchy.

Mid-Fi Wireframes / Dirty Prototype

Design

Design

Before diving into designing wireframes, I decided to identify task flow that will guide the direction of pages needed and to see the overall flow in the website.

The task flow below addresses the pathway users will take the first time they explore the functions as a result of Instagram advertising the new feature

Task Flow

Before iterating the final prototypes, I created these ‘dirty prototypes’ to set the design and use as platforms for users to test my design. These are the barebones that are an amalgamation of screenshots and some initial designs to push forward with the usability testing.

Account Prioritization

Preferred Content

To use the Account Prioritization feature, go to your Profile Page. Then select the ‘Following’ Tab.

For the first-time use, Users will

find the prompt for the ‘Preferred Content’ feature in their ‘Explore Page’. To access this feature at any other time, you can visit your ‘Settings and Privacy’ tab, and select ‘Suggest Content’.

In this screen selection the entirety of all hashtags, words, phrases etc that are associated with the content you follow and/or engage with is shown. From here you can use the ‘Search’ bar to add new hashtags to be recommended. Of existing hashtags, you may select the ‘X’ next to the phrase to eliminate that hashtag from being recommended as part of your AI calculation.

This screen shows the altered selection which includes a series of the new hashtags added to the AI recommended content, while also highlighting that some of the previously shown hashtags have been eliminated.

Upon first-use, you will see the new feature advertised. You can select ‘Try Now’. To access this feature

at any other time, you can visit your ‘Settings and Privacy’ tab, and select ‘Suggest Content’.

When in the ‘Suggested Content’ page, the new features

are specified in their own section.

The ‘Account Prioritization’ screen shows a full list of the accounts you follow. From here you will follow the prompt:


“Review the accounts you wish to prioritize. This will

increase their importance and visibility within your

feed, and adjust the accounts and content you will

be recommended.”


You will select the blank circle of every account you wish to highlight. Once completed you will select ‘Done’. This will take you back to your ‘Following’ Screen.

Once the Prioritization has been completed and you are returned to the ‘Following’ screen, Users will note a new Category titled ‘Prioritized Accounts’. Here you can see the accounts you have highlighted, and then edit the ranking you’ve applied.

When in the ‘Suggested Content’ page, the new features

are specified in their own section.

The original name of this feature was

‘Content Ranking’

Usability Testing

These “dirty” prototypes were used to conduct our Usability Tests via Maze.


Evaluating the Prototype via Usability Tests:


Before testing to participants, I created a usability testing plan to define what I want to test and what I need to iterate. I conducted 2 online unmoderated usability tests via Maze wherein each participant had to complete the following:


Account Prioritization


Please prioritize the following accounts:


Mark

Daniel

Lyn

Armani


Select each account to filter into the ‘Prioritized Account’ category.


Preferred Content


Configure the settings in this prototype removing all existing tags. Add tags to new present content relating to the following:


Tesla

BMW

iWatch

Apple



Results:


8 participants performed the Usability Test.

Feedback on Account Prioritization:


Feedback on Preferred Content:

“Regarding the verbiage for the Account Prioritization feature, decrease the amount of words to diminish cognitive load:


Review the accounts you wish to prioritize. This will not only increase the account importance and visibility within your feed, it will also adjust the types of accounts and content you will be recommended.”

- Participant #3

“I would love a feature like this because it would allow me select who I’d like to see more content from rather than muting or unfollowing other accounts.


I like that there is visualization of who is shown in the feed”

- Participant #4

“This feature is another fantastic opportunity to prioritize the accounts of my friends and artists that I love. “

- Participant #1

“These features all blend seamlessly within IG’s existing design and function. This additions look and feel like something the application can already do, which is great when wanting to design an add-on feature to an existing program !”

- Participant #2

“The Content Ranking feature is very helpful. Educating the AI to direct refined recommendations will hopefully get me to see more of the art content I feel is missing from my feed.”

- Participant #1

“The feature was easy to locate and the pathway made sense but the term “Content Ranking” was a bit misleading. A better term may be “Preferred Content” as there is no “ranking” involved.”

- Participant #2

“I personally love this feature because IG likes to saturate me with content and profiles I don’t follow or memes that distract me. Narrowing down the content I see can inspire me rather than distract me. The navigation was very intuitive”

- Participant #3

Final Prototype:


The final prototype noted some of the criticisms, particularly the two notes on reducing the verbiage for the Account Prioritization feature, and changing the name of “Content Ranking” to “Preferred Content”, as one participant noted there was no ranking valuation within the feature so the name was misleading.


The final prototype reflects the Light Appearance of the application, a generic profile, and a cleaner appearance that aligns with Instagram design standards..

While developing this feature, my main goal was to seamlessly blend my design into Instagram’s UI to ensure the feature was integrated into the existing appearance of the application.


I envision the application will utilize a growing AI to identify themes, content, and values to better reflect the desires of the users. Human staff would be involved in the moderation process, but ultimately the new filters would increase the intelligence of the AI.



Next Steps:


One user pointed out an interesting feature that could be added on in a future iteration.


“A great feature to add in the future may be numerical ranking to this feature so the AI gets an

even better reading on who is important within your selection.”


With this the AI will get a better estimation of the true priority user’s place on the accounts they follow. The AI could determine the values user’s place on the content they ingest. The AI could determine if the priorities are based on interests, career, or friends and family. Ads can now reflect that value as well, making determinations on the types of products you’d invest in that can speak to the kind of lifestyle you’re cultivating. For examples, users that place high priority on engagement with friends and family will now be recommended content, accounts, and products that serve the experiences one has with friends and family. The direction of the AI would now be intended to help users cultivate a better social media experience and foster a lifestyle conducive to their values, for example.

Reflection

Account Prioritization

Preferred Content

The ‘Final Screen’ showing the new content expressed by the revised AI as a result of adjusting the ‘Preferred Content’

The ‘Final Screen’ showing the new Category titled ‘Prioritized Accounts’. Here you can see the accounts you have highlighted, and then edit the ranking you’ve applied.