Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: All You Can Optimize on Your Website in a Single Report (dareboost.com)
65 points by damienj on April 14, 2015 | hide | past | favorite | 46 comments



Feedback for @damienj: * Your Terms & Conditions link is a PDF that downloads vs displays inline. That's super weird, and I thought I got a report without signing up. Nope, it's your ToS.

* The paginated loading of HTML/CSS validation does strange things with the loading bar. I was trying to drag it to the bottom, and it kept moving itself upward as it loaded new pieces. Doing the deferred loading is cool, but it would be better to kick it off async and have it run to completion, versus waiting for scroll-per-page.

* In the downloaded report, there's no indication of what the grey circles are (average top-1000 result).

* Detected Technologies is odd, it only lists comScore for the site I tested - should that include Google Analytics, Facebook Graph, etc.?

* The Table of Contents isn't clickable to get to the related sections.

* The color-coding of sections doesn't mesh mentally with the counts presented. I have something in green (Success) with 7 tips, and something else in red (Issues) with 7 tips as well. Is that good or bad?

* I didn't check the "short" report option, but I'd like to see just things-to-fix in its own report, with details of how to fix them. As-is I have a 35-page report with no clear indication of what the biggest problems are, just the biggest per-category. I realize the webpage already does that, but unless I'm the sole developer I would want to send a hit list out to my team as a PDF.

* Your own page is only 99 out of 100 =)

All of that aside, this looks really useful. It's like a prettier version of http://www.webpagetest.org/ - a lot of the same info, but a much better / more modern presentation. A good pricing hook might be a free (scheduled) report once a month, with the top-level warnings included but a paid option for the full detail.

Good luck, and congrats on "SHIP IT"!


Thanks a lot for this great piece of feedback.

- T&C : that was a fast fix for something, it will be solve as soon as we find a few minutes

- lazyloading on W3C listings: agreed and noted

- PDF/average top-1000 result : indeed, we have to add that!

- Detected Technologies: Yes for Google Analytics, that's odd. No for social widgets. We are mainly focused on CMS and frameworks.

- Table of Contents: you're right

- color-coding: it will be changed soon. New policy will be: if you have one issue at least in a given category, the category will be red (for now is based on the average grade of the category).

- "short" report: you're not alone, we add you to the list, but still to few people

- 99/100 damned!

"A good pricing hook might be a free (scheduled) report once a month"

> you have that daily for free (from you're dashboard, you can create a "monitoring"). Paid options allow to use the tool on other pages than homepages and to use advanced settings (custom bandwidth, authentication, etc)

Thanks again! (and sorry for typos)


This would be great if you didn't need to register to get advice from tests. There's no obvious reason to require it (your website says it's free, after all). There are several other test sites that do not need registration to do so.


I know, that's a debate we regularly have with the other founders. Not sure if we'll keep that choice that way.

But still, I'm sure we offer checkpoints you won't find on other websites (cutomized tips according to your CMS, etc)


However, I did like that it unlocked the page without having to reload when I registered! Nice touch!

a few small spelling mistakes: En savoir plus. (is that being cute, or just not xlated yet?) This page do not provide information to social networks.

Still overall I am quite pleased with the report quality! And I like that it tells how to fix things like https, Open Graph, etc.

I am curious why my images load slowly though! It takes four of my images (png/jpg) from 1.56 to 1.95 seconds to load. They are 13kB to 32kB. One is 111kB. Any ideas? Is that just the server's fault (hostmonster)? Would it be worth it to host the images elsewhere?


Single Page Applications are great ;)

spelling mistakes > thanks a lot, it has been fixed! (correct answer was : "not xlated yet"!)

Thanks for this positive feedback too. About your issue with your images, have you test your website from the nearest test location from your hosting provider? (latency can hurt)


Apparently, you can look at the source and see all the recommendation without registering.


yes you can :) Feel free to do it, but once again we will not spam you so do not be afraid to register you'll save a few seconds rather than playing with your debugger!


This seems very similar to Google's Pagespeed [1], but after registering and diving in a bit, I see some very useful extra features.. recurring monitoring for one, cache/connection simulation, and some other niceties that GPS doesn't give you.

Very nice!

FWIW, I don't share the hatred of registration that the others here do. I don't think it's asking a lot for an email address and a password after you perform a useful service for someone.

[1]: https://developers.google.com/speed/pagespeed/


yeah we indeed share some checkpoints with GPS (because we're using it, and YSlow too)

Thanks for this great feedback, and thanks for your point of view about registration, that's very valuable for us to have both sides, that's a difficult choice (anyway, inorder to offer the free monitoring,we need an email address!)


You force me to register to see the blurred text.

Ok let's take some time to create a temporary email address because, well, your product looks refreshing.

I put down the email address (the one I just spent time creating!) and a password. Your now reject my temporary email address as being temporary !

If I eventually give up, and give you my real email address, will you send my any email ? That one would be unsollicited, and you can be sure that people who created a temporary email address will mark your emails as spam, which may destroy your sender reputation...

People who give their real email address may not mark your emails as spam. But people who take the time to create a temporary email address will, with very high probability.


The email address will be used if you create a free monitoring, to send you weekly digests and alerts in case of quality issue / slowdown on your monitored page. So yes, we reject temporary address as we don't want to analyze daily your webpage if you're not reading our e-mails.


Registering, we will be able to disable any email notification even before we send you any


What if I just want to see the blurred text?


Then just freakin' sign up already! With the amount of time you've spent typing these screeds, you could have registered, decided you didn't like it, and killed your account about 10 times over.

This is not helpful criticism in the least - most people don't object to sharing something as innocuous as an email considering they just did something nice for you.


It's asking me to sign up to read the suggestions, Quora-style. Yucky yucky. Consider removing that please.


please read last comment!

We're not really Quora-sized (4 founders - 25yo).


Cool.

There's never enough of these type of tools.

An alternative of this site which I have used in the past is: http://www.webpagetest.org/


Thanks! WPT is great and we love it, but we would want to go further (with quality checkpoints, not only performance ones) and with a easy-to-use GUI (lots of our users are not very technical people)


Great start. I work a lot with web performance monitoring tools. This is definitely a fresh approach.

I would definitely build out features/checks that other tools aren't doing at the lower price-point. For example:

* Geo-location speed comparisons (e.g. performance from New York -vs- Sydney). You seem to be halfway there, just need to add the comparison graphs.

* Bandwidth speed comparisons (e.g. 4G connection -vs- 100Mb Fibre). Most people don't test their websites on slow connections.

Well Done. Ping me if you want some assistance.


Thanks! comparisons are indeed part of our roadmap, we have some mockups, but nothing we are really proud of for now! I would be glad to have more of your feedback, thanks!


> DAREBOOST may use the User’s trademarks and logos for commercial purposes on any document or in any medium throughout the term of the Contract.

Why not just politely ask a business for their permission when you want to use their logos rather than requiring it in your ToS?

I'm disappointed because this looks _really_ nice and we're just getting starting on a new revision of our site.


I understand what you're saying. I take full responsability of the inconvenience - as I'm the one that have added this line :)

If you don't want us to do so, just drop us en e-mail, and we will not and never do so.

About the background of this line: "politely ask a business" > unfortunatelly, it takes time, we are 4 young guys and are dealing with bigger companies. We were facing issues to get answers, we were losing time, and for the one and single goal to add a logo on our homepage and to help us to make a living from this startup...

If you want to be more confident about us: https://www.dareboost.com/en/about-us


Thanks for sharing your reasoning but you're asking for far more than a logo on the homepage. Perhaps you could add a checkbox at signup that says something like, "Will you allow us to use your company name and/or logo on our homepage?"

That will give you a good idea of how your customers feel about this. It's probable that many people are signing up for your service and are not authorized to give you permission in the first place.

Good luck to you and your team. I'll think about sending in the email you mentioned.


Thanks for your answer. That would be an idea, but it's still an additional checkbox and we like to keep it simple, because asking thousands of users just to use 10 or 15 logos does not really worth it.

But I agree that theoretically T&C allow us to do more than using logos on our own website, I'll try to find a way to add a restriction on what we can do without explicit consent.

Thanks! (my nickname is simply 'damien' in my pro email address if you want to send me directly the mentioned email)


Seems really nice - signed up.

But - you really, really need a native english copywriter to go through all of your english text. It's full of small mistakes which detract from an otherwise good first impression.


Thanks for the advice. We look forward finding someone able to help us to grow the business (we're all engineers...) and to be native english, and still be able to speak a technical english. Not easy, but we will as soon as possible!


There are a couple misspellings that I've noticed, plus some awkward phrasing. Maybe get someone else to look over the copy on your site? This looks really helpful otherwise.


Sorry for that. We're some french guys and we try to do our best, but for now we did not have the chance to hire someone to correct that. There's a "red thumb" on the left of each tip, you can warn us about misspellings with it, it will be widely appreciated :)


These automated tools are always cool, but from where did you take that we should "Define at least <h1> and <h2> titles in your content"?


SEO purpose, semantic way to define important content


Heading tags also have extreme usability and accessibility features. Screen readers love using them to jump around pages.


You're totally right! Thanks for the update.


Yes, but it is totally arbitrary. Why h1 and h2, not h3? Why not only h1? Why not <header> and <footer>?


It's kind of annoying and hypocritical when you give my website the accessibility suggestion that you shouldn't use target="_blank" when the page I am on is riddled with links opening in new tabs. I dunno I don't see that as getting a 0/100 for having 20 when your results page has 56 of them.


This suggestion only apply when using mobile testing with our tool. We do not penalize target="_blank" on desktop websites.

And unfortunately, for now, our service still needs big updates for mobile UX, but most of our users only use us on desktop.


I would appreciate the ability to add a custom user agent to the request so that this could run on sites that detect a mobile user agent and change accordingly.

This is good stuff. Great performance checklist.


From our homepage, under the URL field, you can select a mobile device (Nexus) rather than Firefox.

With a paid account, you can customize any HTTP Header (and more)


Hey guys! you're a lot from the U.S., feel free to test your website from our New York test location !


Just curious, what is the technology used to control browsers programmatically to carry out such tests?


We use Selenium and some proxies to collect data


Thanks!


Congrats guys!

Just a quick suggestion, try to do something server side instead to use `.blur` css class.


Thanks! We know, that's currently not an issue for us if you want to bypass it, as the advice are available after free registration!


FYI: your text blurring on the signed-out analysis page can be defeated by some very simple twiddling with the page source. You may want to render the blurred text to PNGs or something instead of trying to hide it with CSS.


yes we know thanks. Not an issue for us as registering is free. The idea is just to get you to create an account and to try our monitoring feature!




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: