Faster Hosting Now Within Reach and User Experience Metrics

Scrolller Weekly Update October 30th, 2021

Intro

Hello and Welcome to our weekly update. This week's main highlights are solutions to faster hosting, the launch of the leaderboard, the launch of safe for work categories, and user experience metrics.

If you are new to the newsletter, please subscribe to keep up with our journey of building a 100 million MAU web application!

User Highlights

Our users pour their love by sharing their feedback about our service and by being open about their pain points. As a company with User-Driven Development Focus, we welcome this feedback and make sure to prioritize this.

User Feedback

This week's user highlight goes to "ISPHOD". The user feedback is as follows:

The Apollo app, their “pro” level has the ability to fast forward and rewind video and gifs just by swiping right / left them. It would be great if Scrolller had the same feature for videos and gifs.

Otherwise, I love the UI/UX on both mobile and desktop and its focus on content, love the auto-scroll feature as well. As soon as I can spare the extra change I’ll sign up for premium.

Scrolller's Action

"ISPHOD" We really appreciate you taking some time out to describe your pain point to us. Scrolller doesn't have an app ready for its users yet. We are planning on working on the IOS and Android apps as soon we provide our all users with a fast and enjoyable experience.

If you would like to share any feature feedback and contributions, don't hesitate to let us know. We greatly appreciate your taking valuable time to offer us feedback and support on this journey towards building Scrolller.

Go Here to help us with Scrolller Features testing.

User Experience

288856 → 292886

User Accounts Growth

SFW Categories:

Safe for work categories is now launched! They will allow our users to explore new content and enjoy a Scrolller experience tailored to their taste, by viewing only the content categories they love. SFW and NSFW categories can be switched with the same NSFW toggle at the bottom of Scrolller. If you see any errors while using this page, please reach out to us. Our development team will fix it on a priority basis.

Backlink to Scrolller for Traffic:

To increase our traffic on Scrolller from other sources, we are now adding the links to the original websites on our content collections. Our content partners can reach out to us for a link to their website on Scrolller collection in return for a backlink to Scrolller on their platform. This will result in a win-win situation for both of us!

Please reach out to us. Let's start a mutually beneficial relationship by promoting other areas on our website to our users, where they can enjoy the content that they are interested in.

User Experience

User Experience Metrics

To provide our users with the best user experience, it was crucial for Scrolller to understand what our current user experience is like for our users. Previously we only had three measures for user experience: Good, average, or bad.

The challenge was to identify how we can measure the quality of the user experience. and what determines a good user experience on Scrolller. Our team started to figure out multiple metrics to measure the quality of the user experience.

• Mean Content load time

  • Our team created a new tool to investigate the mean of content load time on the user end. We ran this experiment internally and found that at the same time, two people in different geographical locations were having a completely opposite user experience.

  • We established that mean of content load time is a direct measurement of the quality of user experience. Following are the metrics that we established:

    Good User Experience: < 1-second

    Average User Experience: 1 second to 2 seconds

    Bad User Experience: >2 Seconds

  • Our team is now working on integrating that tool in the website so we can have sample data from all over the world. As Scrolller has a user base in over 120 countries, we want to provide the optimal user experience in every geographical region.

• Mean time to first Favorite

  • If things are normally distributed, the mean & the median are pretty close. On investigating we found that mean and median of first time to favorite on Scrolller were far apart. This means that either the majority of people “like” almost immediately, and then those that don’t like immediately takes a while, pushing the mean out. Mean time to Like: < 18 Minutes Median time to Like: < 30 Seconds

• Mean time to first Follow

  • We observed the same pattern in follows as we observed with favorites. The median time is pretty low where as the mean is higher. People who follow a collection on Scrolller usually do it within 3 minutes. While for others it may take up-to 24 hours to follow a collection. Mean time to Follow: ~ 24 hours Median time to Follow: < 3 Minutes

Metrics Under Investigation:

• Most viewed pages

• The average number of errors on page

• The volume of new content added to Scrolller each day

• Average time for the user to create an account

• Average time for users to subscribe to premium

These necessary metrics are creating a nice base benchmark for our current user experience on Scrolller which will let us easily track the improvement by comparing the new values with them.

Please let us know what you think about these user metrics. If you have any metrics that might be a good measure of user experience, we would love to know more.

User Experience Issues - Resolved:

• Opening collections in leaderboard is always taking you to first collection.

• Prevents users from accessing the subreddits with <10 posts.

• Remove Empty Collections from Scrolller.

Roadmap for Growth for User Accounts

• SFW Categories (Completed)

• Unlocking Categories plan (In Progress)

• Content Recommendation (In Progress)

• Pro-Metric Advertising (In Progress)

• Scrolller APP (Back Log)

• Scrolller Extension and Enhancements (Blocked)

Revenue

Faster Hosting, Now In Reach:

Scrolller has finally found the key to provide faster hosting to our users. We have been searching for solution to this problem in every place when solution was right under our nose all the time.

SSD NVMe Experiment Results

Our developers tested how our new hosting server with SSD's and increased CPU's will help improve our average content load time. This metric really effects the end user experience and is the one of root causes of broken columns. At the end of result we found out that these new servers were in turn faster but not enough to provide our users with optimal user experience.

This test turned out to be a failure, and we started to look into other areas.

Using Faster Server as Cache:

Our developers came up with a new idea to use a faster host as a dedicated cache server for our slower servers. By doing that we started to serve content to our users from the cache using the slower hosts a pool to populate and refresh the cache server. We performed the experiment internally within various geographical regions. The results were very promising as we saw a ~175% increase in the quality of user experience using the metric "Mean Content Load Time".

The issues with column stop loading and video playing errors were immediately resolved with this new approach.

Our development team is working on the tool to measure the mean content load time from our users across all geographical locations. We will share the results with you in next news letter what we found in this investigation.

Estimated Launch for Premium Users:

Our experimentation with this new approach will last in 2 weeks after which we will launch the new faster hosting. We are running various tests to ensure a fast and resilient service to our users.

Monetization Optimisations:

Refreshing Native Advertisements

For a couple of weeks, we started noticing that the revenue from the native ads is gradually decreasing. We contacted our native advertisement partner to see why our profits from these ads were going down. They provided us with fresh batch of native ads that are supposed to be high performing. We have updated the site with new Ads, removing the old ones to get back on track with our revenue generation from native ad Partners.

Adding New Advertisement Partners

Scrolller is also getting in contact with other native ad networks to partner with. Addition of new ad partners for native advertisement will ensure a nice increase in our revenue from the native ads. We are getting in contact with EvaDav and A+ Advertisements as they claim to be offering the highest profit rate in the market.

Utilizing Scrolller for Better Advertising

There are various areas in Scrolller that are heavily used by our users but are not being used for advertisements. We found out that the Ads are usually getting served on the main page of Scrolller.

The areas which can be utilized for efficient advertisement are as follows:

• Search page

• More like this page

• Favorites / Following

While we are making these tweaks to optimize our monetization, we are also ensuring that these monetization optimizations don't have any negative impact on our user experience.

Please let us know in comments what you think about these monetization improvements on Scrolller.

Roadmap for Growth in Revenue

• Faster Hosting for Premium Users (Testing)

• Monetization Optimisations (In Progress)

• Multiple Payment Processors (Design Stage)

  • Bitcoin

  • Paypal

  • High Risk Payment Processor

• Exclusive Features for Premium User (Design Stage)

Social Traffic

Social Media Team Update

Our social media team is now leaving Instagram and Twitter management. Why? Even with the amount of time invested in these platforms, the results are not promising compared to Facebook and other methods to outreach potential users.

Because of this, our social media team decided that these social media are no longer worthy of our time, and will stop posting & engaging with users there. Despite that, our Facebook Page keeps giving results to outreach users, much more aligned with our team's goals.

Social Traffic Update:

Our overall traffic from social media channels to the website is still dropping this week. This is one of the main reasons why Twitter and Instagram aren't going to stay with our team's plan.

Total Social Sessions 11,728 → 14,396

FB, TW & IG Sessions 4,230 → 3,749 (11.3% decrease)

Currently, our team changed the aim to reach a stable traffic source coming from Facebook, as this channel has x5 more reach than our other platforms.

Social Media Team is now the Outreach Team

Our social media team is now changing its name to the outreach team, since driving traffic to Scrolller was the main objective since its launch; this name is more accurate for the tasks they work on and the methods they need to develop. Social media is just one of the methods that could work for our vision of success but can be accomplished in several ways.

Right now, our outreach team is aligned with the objective; reach users and get them to know Scrolller. How are they going to make it? Our team has already started working those outreach muscles and considering the possibility to implement Reddit as one of the platforms for them to work in.

CTA Test

Our outreach team keeps running tests on different Call-To-Actions to enter our website. We still maintain the objective of maximizing the amount of traffic & engagement on our website we can get from the users we reach with posts, generating interest in them.

So far, they found out that:

• Generic CTA's bores users.

• Slight changes can give a higher conversion rate.

• It's important to transmit our website haves something exclusive.

• Generate feelings with content and CTA combined means engagement.

Transmitting an impression of value to users is key. If not, why would someone click on a link?

Follow us to not miss anything: 🔘 Instagram 🔘 Facebook 🔘 Twitter

Thank you for being our fans and followers on social media.

Do let us know anywhere if you have strategies that have worked for you on social media or have any valuable feedback so we can learn from them. Looking forward to hearing from you!

Roadmap for Growth in Social Traffic

• Hashtag experiment (Done)

• CTA Test (In Progress)

• User Generated Collections: Create and share your favorites and follows (In Progress)

Team Highlights

Team Updates

Scrolller Team Highlights for the week!

Nick — Working on user experience issue tickets

Mannan — Working with developers to solve OVH issues.

Karol — Testing new hosting server and transferring data across hosts to maximize performance

Valentin — Hash tag experimentation, social media traffic analysis, Flo Hiring social media channels

Tim — Working with Nick to develop emailing system for recommendation system.

Matt — System performance analysis & metrics

Chris — Investor outreach, team feedback sessions, hiring

Have any feedback for us? Don't hesitate to share it with us in the comments sections. We like hearing any feedback you have for us.

Road Map

You can share feedback with us about any of the new features below. Using this Airtable Form. Looking forward to your comments

This week

• Faster Hosting for Premium Users (Testing)

• Content Recommendation System (In Progress)

• Exclusive Features for premium users (Back Log)

• Monetization Update and Optimization (Back Log)

• Enhancements to Auto Scroll Feature (Blocked)

• New native advertisements (Done)

• Categories page for SFW collections (Done)

• User Generated Collections: Leaderboard**(Done)**

P.S. As a reminder, subscribe to our newsletter if you are yet to walk with us on this Building Scrolller exciting yet challenging journey.

Also: Sign Up To Help → https://airtable.com/shrrRkFqMSZBqF24Q us translate Scrolller to Russian.

To help us with Scrolller features testing, Go Here

Until next time, happy scrollling!