1
11 Comments

Web App Speed Optimisation (Paid Full Stack Project)

I own a web application that's designed to combat misinformation, by helping people crowdsource and moderate factual content. It's called òtító and can be accessed here:

www.otito.io

We already have just over a hundred people signed up and have hundreds of monthly visitors. One key challenge has been the speed of the site (it has a Google Page Insights speed of 4 out of 100, on mobile devices).

https://developers.google.com/speed/pagespeed/insights/?url=https%3A%2F%2Fwww.otito.io%2F

I need a developer (I suspect full stack) to help me optimise the code/site and take the speed of the application (on mobile and desktop devices) to over 90 out of 100.

Development Tech:

.netcore
AngularJS
MySQL as database

The site is hosted on Digital Ocean and has a couple of integrations (e.g. Firebase for single sign-on).

My budget is $500 (somewhat negotiable, depending on any additional value).

My biggest concern is that bringing on a new developer at this stage—when the application already has many features (some of which are hidden)—might lead to an inadvertent proliferation of bugs. So, it's really important to me that whomever picks up this job is very willing to learn in-depth how the application works (so they know what can be broken) and has a robust QA plan/approach.

I look forward to working with you. Thanks!

  1. 3

    Hey Timi,

    I happen to have quite a bit of experience with this. 90/100 is an ambitious goal. Mind you that this score is not actually calculated based on your website's performance. It is calculated based on your website's performance compared to the other websites. The other websites include blogs, personal websites, portfolio websites etc. which are easier to make fast due to their static content. It's not a coincidence the most of the major websites like Facebook, Airbnb, Reddit etc. don't make it over 50. I would recommend you to reach lower so that you can find help easier.

    1. 1

      Thanks for the comment. Agree with you (broadly)—I worked at Facebook until recently and the site speed was a constant bug bear (although it was successfully increased in the past year, you may be aware).

      However, my site is one of those that can/should have a high page speed score. I'll just paste my reply to another comment:

      "I'm aware not all components of page speed are direct referents for actual speed (in a performance sense). However, many of them are (e.g. first contentful paint). On most modern connections (like I imagine everyone on here has), the speed differences are barely noticeable. But in poor-connectivity situations (e.g. certain geographies or while on a train), these variables come into play.

      Most importantly though, Google factors many of these variables into search engine performance (which is the main reason why I'm optimising the site)."

  2. 2

    Timi, one of my company’s core services is website optimization (but only for WordPress, sorry).

    We’ve helped people with speed and loading issues since about 2015. I am not guessing or speculating when I tell you this:

    I looked at your site’s test results, and I agree with all the advice posted to this point:

    Google PageSpeed is a measure of best practices - not actual page speed.

    Your biggest problem is your images. To solve it you need to optimize what's already on the site, and put in place practices/workflow to ensure future uploads are optimal.

    Images aren’t your only problem, but in the quick review I did there wasn’t anything I would recommend as a first step that couldn’t be handled by someone with a basic understanding of website development and how servers and browsers work.

    Ideally this person would be someone who can suggest a better workflow so you and your helpers can adhere to best practices going forward.

    1. 1

      Hey, there, thanks for the comment. I'm aware not all components of page speed are direct referents for actual speed (in a performance sense). However, many of them are (e.g. first contentful paint). On most modern connections (like I imagine everyone on here has), the speed differences are barely noticeable. But in poor-connectivity situations (e.g. certain geographies or while on a train), these variables come into play.

      Most importantly though, Google factors many of these variables into search engine performance (which is the main reason why I'm optimising the site).

      1. 2

        Couple of things to note on your comment.

        I'll agree that loading speed is important, but its use as a ranking factor is dubious (see this: https://backlinko.com/search-engine-ranking ).

        In my experience, the correlation between speed and poor rankings is there because people with slow load times tend to have a number of issues they need to work on – all of which affect bounce rate, time on page and returning users.

        Re: "modern connections" – When most of your connectivity is via big, hard-wired pipes, it's easy to forget how huge a factor mobile data connections are, beyond just being on a train.

        In less developed countries, mobile data may be all that's available in places. You may not need to worry about those people if your users are not among them. But because the average person in more developed countries still spends a significant percentage of time connecting via mobile/cellular data, it behooves us to care about more than modern connections.

        Up until two years ago, I lived near Atlanta and big, big "pipes." I had a fast, hard-wired connection to them. Now, I work, travel and live all over the US, relying 100% on cellular data. Even in this first-world country the connection often sucks, and that can come as much from being in overpopulated areas as often as being out in the boonies.

        The bottom line is that lighter pages (talking kb here) with few or no calls for external resources will load faster for everyone. FCP and every other measure.

        Even if you want to continue searching for more-complex causes and solutions, you should address the basic issues previously mentioned in this thread.

        Best of luck to you.

  3. 1

    Hi, I'm not available to do the coding work to fix the PSI score. I will offer some advice as I have been fixing slow web sites for over a dozen years. Google did recently change the metrics and average weight of the timings when it announced web vitals. The critical metrics of LCP and TBT are weighted the most (see https://web.dev/performance-scoring).

    Your PSI score is not directly used by Google in its page experience ranking (which is a factor in its page ranking algorithm). But Google Web Vitals will sometime in 2021 (LCP, FID, and CLS). So it is MORE important to track your web vitals in Google Search Console than to score a 90/100 as that is explicitly going affect your ranking in 2021.

    So you will want to make sure web vitals are your focus for improvements, and they will have the secondary effect of improving your PSI score.

    Here are some things that might not take any developer time which can improve your PSI score...

    1. optimize your hero image as it is the driving long LCP times and it does not take any developer to do this. This is a web vital.

    2. While "animations" look great they will increase your CLS score. So removing them should help with your CLS score which is a web vital.

    3. Now web vitals track your FID (first input delay) but that is ONLY a field/RUM metric. So as a proxy, PSI (and webpagetest) tracks total blocking time. You can visualize total blocking time using webpagetest.org.

    So check out this...I ran your site through webpagetest.org...

    https://www.webpagetest.org/result/200616_ZS_e7174b08b587901d52e7075686a4aed0/1/details/#waterfall_view_step1

    You will want to look at graph at the bottom of the waterfall table (just able the connection chart). It is labeled "page is interactive" and is red and green. So the x-axis is time and it charts if the page is interactive, all of the green is good while the red means the page is blocked. The more red -- the higher your FID and TTI (time to interactive). So you want to limit the red. Almost guaranteed, scripts are blocking your page.

    Now I loaded this webpagetest into my little app https://waterfaller.dev/ which helps you analyze and annotate test results.

    https://waterfaller.dev/#200616_ZS_e7174b08b587901d52e7075686a4aed0

    So I used the range slider to identify the 21 requests being made before start render, and then filtered down to the 9 scripts being loaded. These scripts are what are increasing your FID or TBT.

    Suggestion -- combine these first party scripts

    https://www.otito.io/js/vendor.js
    https://www.otito.io/js/site.js

    Suggestion -- is GTA needed? Can you just add Google Analytics without needing a tag manager?

    https://www.googletagmanager.com/gtag/js
    https://www.googletagmanager.com/gtm.js

    Suggestion -- nothing here...

    https://www.gstatic.com/firebasejs/6.3.4/firebase-app.js
    https://www.gstatic.com/firebasejs/6.3.4/firebase-auth.js

    Suggestion -- do you need recaptcha?

    https://www.google.com/recaptcha/api.js
    https://www.gstatic.com/recaptcha/releas...PIBTU15LG6xPf/recaptcha__en.js

    Suggestion -- load this async or defer or get rid of this

    https://js.hs-scripts.com/6019738.js

    So that's a lot, with a lot of abbreviations and I assume a lot of things. I think some of these things can be done without hiring a developer. Let me know your thoughts.

  4. 1

    This is just an standard optimizing, resizing, compressing, htaccess gzipping/browser caching task, and whoever helps you don't need to know much about backend stuff. All actionable steps are listed here >> https://gtmetrix.com/reports/www.otito.io/fjTWFGU3 , click on the issue and do what is recommended.

  5. 1

    Hey,

    I'm a web designer who can handle HTML, CSS & jQuery. While checking your webpage, I guess some of the issues which causing the slow can be fixed by me. There are some issues which can done with some server support.

    The page speed displays only for the home page. So i can start with the home page first and if its going good, we can move further.

    Reach me on mail [at] surjithctly [dot] in

    We'll discuss.

    1. 2

      Thanks for the response. I'm looking for someone who can handle all aspects of site speed optimisation though.

  6. 1

    Looking at the PageSpeed stats page, you could be able to increase the mobile score a lot by reducing your image sizes, improving the font load by pointing to a local font family before loading by using font-display: swap (see https://web.dev/font-display/), and from there, probably looking at a faster replacement for Sumo.

    1. 1

      Thanks! I'm aware of some of these issues (I'm semi-proficient in web development).

      The issue is I have another full-time job (as well as this), so I definitely don't have the time to do the ads, social, content, outreach, analytics AND optimise the site.

      If you're able and willing to help me optimise the site speed, I'm willing to pay.

Trending on Indie Hackers
How I grew a side project to 100k Unique Visitors in 7 days with 0 audience 49 comments Competing with Product Hunt: a month later 33 comments Why do you hate marketing? 28 comments My Top 20 Free Tools That I Use Everyday as an Indie Hacker 14 comments $15k revenues in <4 months as a solopreneur 14 comments Use Your Product 13 comments