Recherche avancée

Médias (0)

Mot : - Tags -/images

Aucun média correspondant à vos critères n’est disponible sur le site.

Autres articles (38)

  • Support audio et vidéo HTML5

    10 avril 2011

    MediaSPIP utilise les balises HTML5 video et audio pour la lecture de documents multimedia en profitant des dernières innovations du W3C supportées par les navigateurs modernes.
    Pour les navigateurs plus anciens, le lecteur flash Flowplayer est utilisé.
    Le lecteur HTML5 utilisé a été spécifiquement créé pour MediaSPIP : il est complètement modifiable graphiquement pour correspondre à un thème choisi.
    Ces technologies permettent de distribuer vidéo et son à la fois sur des ordinateurs conventionnels (...)

  • HTML5 audio and video support

    13 avril 2011, par

    MediaSPIP uses HTML5 video and audio tags to play multimedia files, taking advantage of the latest W3C innovations supported by modern browsers.
    The MediaSPIP player used has been created specifically for MediaSPIP and can be easily adapted to fit in with a specific theme.
    For older browsers the Flowplayer flash fallback is used.
    MediaSPIP allows for media playback on major mobile platforms with the above (...)

  • De l’upload à la vidéo finale [version standalone]

    31 janvier 2010, par

    Le chemin d’un document audio ou vidéo dans SPIPMotion est divisé en trois étapes distinctes.
    Upload et récupération d’informations de la vidéo source
    Dans un premier temps, il est nécessaire de créer un article SPIP et de lui joindre le document vidéo "source".
    Au moment où ce document est joint à l’article, deux actions supplémentaires au comportement normal sont exécutées : La récupération des informations techniques des flux audio et video du fichier ; La génération d’une vignette : extraction d’une (...)

Sur d’autres sites (4374)

  • FFMPEG Incredibly Slow On Windows Server 2016

    1er août 2019, par Ben Gardner

    I have a t2.small (2GB RAM, 1 vCPU) Amazon EC2 instance running a process using FFMPEG. It runs just fine ( 30 fps) on my computer (i7, 12GB RAM), but at around 2 fps on the server. Here’s the command :

    rescale_command = f'ffmpeg -i {srcVideo} -filter_complex \"scale={owidth}:{oheight}, setsar=1:1, pad={dim[0]}:{dim[1]}:{oofx}:{oofy}\" {destVideo}'

    I’ve tried uninstalling and reinstalling ffmpeg/ffprobe

    Edited per llogan’s request, I’ve also discovered that it’s relatively speedy towards the beginning (I’m rescaling 100 times over the course of the program) and slows down towards the end.

    It utilizes 100% of the CPU even from the beginning, though.

    Example command/output :
    ffmpeg -i media/8-1-2019/hi/dl-107.mp4 -filter_complex "scale=607:1080, setsar=1:1, pad=1920:1080:656:0" media/8-1-2019/hi/sl-107.mp4

    ffmpeg version 3.2 Copyright (c) 2000-2016 the FFmpeg developers
     built with gcc 5.4.0 (GCC)
     configuration: --enable-gpl --enable-version3 --disable-w32threads --enable-dxva2 --enable-libmfx --enable-nvenc --enable-avisynth --enable-bzlib --enable-libebur128 --enable-fontconfig --enable-frei0r --enable-gnutls --enable-iconv --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libfreetype --enable-libgme --enable-libgsm --enable-libilbc --enable-libmodplug --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenh264 --enable-libopenjpeg --enable-libopus --enable-librtmp --enable-libschroedinger --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzimg --enable-lzma --enable-decklink --enable-zlib
     libavutil      55. 34.100 / 55. 34.100
     libavcodec     57. 64.100 / 57. 64.100
     libavformat    57. 56.100 / 57. 56.100
     libavdevice    57.  1.100 / 57.  1.100
     libavfilter     6. 65.100 /  6. 65.100
     libswscale      4.  2.100 /  4.  2.100
     libswresample   2.  3.100 /  2.  3.100
     libpostproc    54.  1.100 / 54.  1.100
    Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'media/8-1-2019/hi/dl-107.mp4':
     Metadata:
       minor_version   : 512
       major_brand     : isom
       compatible_brands: isomiso2avc1mp41
       comment         : vid:v09044ce0000bks3802jqrog167l3rf0
       encoder         : Lavf58.20.100
     Duration: 00:00:15.12, start: 0.000000, bitrate: 1046 kb/s
       Stream #0:0(und): Video: h264 (High) (avc1 / 0x31637661), yuv420p, 540x960 [SAR 1:1 DAR 9:16], 972 kb/s, 29.97 fps, 29.97 tbr, 30k tbn, 59.94 tbc (default)
       Metadata:
         handler_name    : VideoHandler
       Stream #0:1(und): Audio: aac (LC) (mp4a / 0x6134706D), 44100 Hz, stereo, fltp, 64 kb/s (default)
       Metadata:
         handler_name    : SoundHandler
    [Parsed_setsar_1 @ 00000000023ac840] num:den syntax is deprecated, please use num/den or named options instead
    [Parsed_setsar_1 @ 00000000023ad2c0] num:den syntax is deprecated, please use num/den or named options instead
    [libx264 @ 00000000023ad800] using SAR=1/1
    [libx264 @ 00000000023ad800] using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX FMA3 AVX2 LZCNT BMI2
    [libx264 @ 00000000023ad800] profile High, level 4.0
    [libx264 @ 00000000023ad800] 264 - core 148 r2721 72d53ab - H.264/MPEG-4 AVC codec - Copyleft 2003-2016 - http://www.videolan.org/x264.html - options: cabac=1 ref=3 deblock=1:0:0 analyse=0x3:0x113 me=hex subme=7 psy=1 psy_rd=1.00:0.00 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=0 deadzone=21,11 fast_pskip=1 chroma_qp_offset=-2 threads=1 lookahead_threads=1 sliced_threads=0 nr=0 decimate=1 interlaced=0 bluray_compat=0 constrained_intra=0 bframes=3 b_pyramid=2 b_adapt=1 b_bias=0 direct=1 weightb=1 open_gop=0 weightp=2 keyint=250 keyint_min=25 scenecut=40 intra_refresh=0 rc_lookahead=40 rc=crf mbtree=1 crf=23.0 qcomp=0.60 qpmin=0 qpmax=69 qpstep=4 ip_ratio=1.40 aq=1:1.00
    Output #0, mp4, to 'media/8-1-2019/hi/sl-107.mp4':
     Metadata:
       minor_version   : 512
       major_brand     : isom
       compatible_brands: isomiso2avc1mp41
       comment         : vid:v09044ce0000bks3802jqrog167l3rf0
       encoder         : Lavf57.56.100
       Stream #0:0: Video: h264 (libx264) ([33][0][0][0] / 0x0021), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], q=-1--1, 29.97 fps, 30k tbn, 29.97 tbc (default)
       Metadata:
         encoder         : Lavc57.64.100 libx264
       Side data:
         cpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: -1
       Stream #0:1(und): Audio: aac (LC) ([64][0][0][0] / 0x0040), 44100 Hz, stereo, fltp, 128 kb/s (default)
       Metadata:
         handler_name    : SoundHandler
         encoder         : Lavc57.64.100 aac
    Stream mapping:
     Stream #0:0 (h264) -> scale (graph 0)
     pad (graph 0) -> Stream #0:0 (libx264)
     Stream #0:1 -> #0:1 (aac (native) -> aac (native))
    Press [q] to stop, [?] for help
    frame=  453 fps=3.6 q=29.0 Lsize=    3016kB time=00:00:15.01 bitrate=1645.7kbits/s speed=0.12x
    video:2763kB audio:237kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.576820%
    [libx264 @ 00000000023ad800] frame I:6     Avg QP:20.93  size: 38467
    [libx264 @ 00000000023ad800] frame P:118   Avg QP:23.12  size: 11624
    [libx264 @ 00000000023ad800] frame B:329   Avg QP:25.19  size:  3725
    [libx264 @ 00000000023ad800] consecutive B-frames:  1.8%  2.6%  4.6% 90.9%
    [libx264 @ 00000000023ad800] mb I  I16..4: 16.4% 76.7%  6.9%
    [libx264 @ 00000000023ad800] mb P  I16..4:  2.3%  5.8%  0.9%  P16..4:  9.8%  3.8%  1.6%  0.0%  0.0%    skip:75.7%
    [libx264 @ 00000000023ad800] mb B  I16..4:  0.4%  0.7%  0.1%  B16..8:  9.5%  1.6%  0.3%  direct: 0.6%  skip:86.8%  L0:43.0% L1:50.8% BI: 6.2%
    [libx264 @ 00000000023ad800] 8x8 transform intra:67.4% inter:78.3%
    [libx264 @ 00000000023ad800] coded y,uvDC,uvAC intra: 32.5% 40.1% 9.4% inter: 2.7% 3.9% 0.1%
    [libx264 @ 00000000023ad800] i16 v,h,dc,p: 50% 25%  6% 19%
    [libx264 @ 00000000023ad800] i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 38% 11% 27%  3%  3%  5%  3%  6%  3%
    [libx264 @ 00000000023ad800] i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 43% 11% 12%  5%  7%  9%  4%  7%  3%
    [libx264 @ 00000000023ad800] i8c dc,h,v,p: 62% 11% 22%  5%
    [libx264 @ 00000000023ad800] Weighted P-Frames: Y:0.0% UV:0.0%
    [libx264 @ 00000000023ad800] ref P L0: 64.4% 14.3% 16.1%  5.2%
    [libx264 @ 00000000023ad800] ref B L0: 90.6%  7.5%  2.0%
    [libx264 @ 00000000023ad800] ref B L1: 97.7%  2.3%
    [libx264 @ 00000000023ad800] kb/s:1496.84
    [aac @ 0000000000628120] Qavg: 754.761
  • Alias Artifacts

    26 avril 2013, par Multimedia Mike — General

    Throughout my own life, I have often observed that my own sense of nostalgia has a window that stretches about 10-15 years past from the current moment. Earlier this year, I discovered the show “Alias” and watched through the entire series thanks to Amazon Prime Instant Video (to be fair, I sort of skimmed the fifth and final season which I found to be horribly dull, or maybe franchise fatigue had set in). The show originally aired from 2001-2006 so I found that it fit well within the aforementioned nostalgia window.


    Alias (TV Series) logo

    But what was it, exactly, about the show that triggered nostalgia ? The computers, of course ! The show revolved around spies and espionage and cutting-edge technology necessarily played a role. The production designer for the series must have decided that Unix/Linux == awesome hacking and so many screenshots featured Linux.

    Since this is still nominally a multimedia blog, I’ll start of the screenshot recon with an old multimedia player. Here is a vintage Mac OS desktop running an ancient web browser (probably Netscape) that’s playing a full-window video (probably QuickTime embedded directly into the browser).


    Old Mac OS with old browser

    Click for larger image


    Let’s jump right into the Linux side of things. This screenshot makes me particularly sentimental since this is exactly what a stock Linux/KDE desktop looked like circa 2001-2003 and is more or less what I would have worked with on my home computer at the time :


    Alias: Linux/KDE desktop

    Click for larger image


    Studying that screenshot, we see that the user logs in as root, even to the desktop environment. Poor security practice ; I would expect better from a bunch of spooks.

    Echelon
    Look at the terminal output in the above screenshot– it’s building a program named Echelon, an omniscient spy tool inspired by a real-world surveillance network of the same name. In the show, Echelon is used to supply plot-convenient intelligence. At one point, some antagonists get their hands on the Echelon source code and seek to compile it. When they do, they will have access to the vast surveillance network. If you know anything about how computers work, don’t think about that too hard.

    Anyway, it’s interesting to note that Echelon is a properly autotool’d program– when the bad guys finally got Echelon, installation was just a ‘make install’ command away. The compilation was very user-friendly, though, as it would pop up a nice dialog box showing build progress :


    Alias: Compiling Echelon

    Click for larger image


    Examining the build lines in both that screenshot and the following lines, we can see that Echelon cares about files such as common/db_err.c and bt_curadj.c :


    Alias: Echelon used Berkeley DB

    Click for larger image


    A little googling reveals that these files both belong to the Berkeley DB library. That works ; I can imagine a program like this leveraging various database packages.

    Computer Languages
    The Echelon source code stuff comes from episode 2.11 : “A Higher Echelon”. While one faction had gotten a hold of the actual Echelon source code, a rival faction had abducted the show’s resident uber-nerd and, learning that they didn’t actually receive the Echelon code, force the nerd to re-write Echelon from scratch. Which he then proceeds to do…


    Alias: Rewriting Echelon

    Click for larger image


    The code he’s examining there appears to be C code that has something to do with joystick programming (JS_X_0, JS_Y_1, etc.). An eagle-eyed IMDb user contributed the trivia that he is looking at the file /usr/include/Linux/joystick.h.

    Getting back to the plot, how could the bad buys possibly expect him to re-write a hugely complex piece of software from scratch ? You might think this is the height of absurdity for a computer-oriented story. You’ll be pleased to know that the writers agreed with that assessment since, when the program was actually executed, it claimed to be Echelon, but that broke into a game of Pong (or some simple game). Suddenly, it makes perfect sense why the guy was looking at the joystick header file.

    This is the first bit of computer-oriented fun that I captured when I was watching the series :


    Alias: Java on the mainframe

    Click for larger image


    This printout purports to be a “mainframe log summary”. After some plot-advancing text about a security issue, it proceeds to dump out some Java source code.

    SSH
    Secure Shell (SSH) frequently showed up. Here’s a screenshot in which a verbose ‘ssh -v’ connection has just been closed, while a telnet command has apparently just been launched (evidenced by “Escape character is ‘^]’.”) :


    Alias: SSH/telnet

    Click for larger image


    This is followed by some good old Hollywood Hacking in which a free-form database command is entered through any available command line interface :


    Alias: Intuitive command line interface

    Click for larger image


    I don’t remember the episode details, but I’m pretty sure the output made perfect sense to the character typing the command. Here’s another screenshot where the SSH client pops up an extra-large GUI dialog element to notify the user that it’s currently negotiating with the host :


    Alias: SSH negotiation dialog

    Click for larger image


    Now that I look at that screenshot a little more closely, it appears to be a Win95/98 program. I wonder if there was an SSH client that actually popped up that gaudy dialog.

    There’s a lot of gibberish in this screenshot and I wish I had written down some details about what it represented according to the episode’s plot :


    Alias: Public key

    Click for larger image


    It almost sounds like they were trying to break into a network computer. Analyzing MD5 structure… public key synthesized. To me, the funniest feature is the 7-digit public key. I’m a bit rusty on the math of the RSA cryptosystem, but intuitively, it seems that the public and private keys need to be of roughly equal lengths. I.e., the private key in this scenario would also be 7 digits long.

    Gadgets
    Various devices and gadgets were seen at various junctures in the show. Here’s a tablet computer from back when tablet computers seemed like fantastical (albeit stylus-requiring) devices– the Fujitsu Stylistic 2300 :


    Alias: Fujitsu Stylistic 2300 tablet

    Click for larger image


    Here’s a videophone from an episode that aired in 2005. The specific model is the Packet8 DV326 (MSRP of US$500). As you can see from the screenshot, it can do 384 kbps both down and up.


    Alias: Packet8 DV326

    Click for larger image


    I really regret not writing down the episode details surrounding this gadget. I just know that it was critical that the good guys get it and keep from falling into the hands of the bad guys.


    Alias: Gadget using Samsung and Lexar chips

    Click for larger image


    As you can see, the (presumably) deadly device contains a Samsung chip and a Lexar chip. I have to wonder what device the production crew salvaged this from (probably just an old cell phone).

    Other Programs

    The GIMP photo editor makes an appearance while scrubbing security camera footage, and serves as the magical Enhance Button (at least they slung around the term “gamma”) :


    Alias: GIMP editor

    Click for larger image


    I have no idea what MacOS-based audio editing program this is. Any ideas ?


    Alias: Apple MacOS-based audio editor

    Click for larger image


    FTP shows up in episode 2.12, “The Getaway”. It’s described as a “secure channel” for communication, which is quite humorous to anyone versed in internet technology.


    Alias: FTP secure channel

    Click for larger image


  • 11 of the Most Effective Conversion Rate Optimisation Best Practices

    14 février 2024, par Erin

    Driving more traffic to your website is hard work, but it’s still only half the battle. 

    You don’t just need to acquire new users ; you need to make sure as many convert as possible to make your digital marketing efforts worthwhile.

    That’s why improving your site’s conversion rate is so important. It will also help you get more value from your existing traffic source and keep you in line with your competitors. It’s also probably a lot easier than you think — especially if you adopt optimisation strategies that have been proven to be profitable time and time again. 

    In this article, we’ll show some of the most powerful, innovative and tried-and-tested conversion rate optimisation strategies you can implement immediately. 

    What is conversion rate optimisation ?

    First, let’s look at what conversion rate optimisation means. Conversion rate optimisation is the practice of improving elements of your website to increase the number of users who take a desired action and turn visitors into customers. 

    Common conversion goals include :

    • Making a purchase
    • Adding an item to a shopping cart
    • Signing up for a newsletter
    • Registering for a free trial
    • Downloading an ebook
    • Watching a video

    It doesn’t matter what your goal is. Using one of the following conversion rate optimisation best practices can send your conversions soaring. 

    11 conversion rate optimisation best practices 

    Are you ready to roll up your sleeves and get to work ? Then use one or more of the following best practices to improve your return on investment. 

    Set a clear goals and hypothesis

    When running an A/B or multivariate test, you need a clear idea of what you are testing and why. 

    A goal (a statement about what you want to achieve) and a hypothesis (a statement about what you expect to happen) clarify the problem you are trying to solve and give you a definitive way to judge the experiment’s results. 

    Confused ? Just use this template :

    We aim to [insert goal] by testing [insert test] on [insert page]. We expect that [insert test] will increase [insert metric] because [insert reason].

    Make sure your goals are directly related to the experiment. If you are testing your CTA button, the goal should be getting more users to click the button. It shouldn’t be a goal further down the conversion funnel, like making a purchase. 

    Start with A/B tests

    A/B testing is one of the easiest and most effective ways to run experiments to improve your current conversion rate. So, it’s no wonder that the A/B testing software market was expected to be worth $1.2 billion in 2023 and hit $3.6 billion by 2033. 

    Also known as split testing, A/B testing allows you to directly compare the conversion performance of two elements on your page, like the colour of your CTA button or your headline copy.

    A screenshot of an A/B test using Matomo

    You can go even further with multivariate testing, which lets you test two or more changes against a single control. 

    For example, the screenshot above shows the results of a multivariate test between a standard header, a wide header and a small header using Matomo’s A/B testing tool. As you can see, the wider header has a much higher conversion, and the increase was statistically significant. 

    Try Matomo for Free

    Get the web insights you need, without compromising data accuracy.

    No credit card required

    Tweak your CTAs

    Calls to action (CTAs) are page elements that prompt users to respond immediately. They are usually buttons but can also be images or plain text links. 

    What your CTAs say, how they look, and where they are placed can greatly impact your site’s conversion rates. As such, this is one of the elements you’ll want to optimise first. 

    There are several tweaks you can test, including your CTA’s :

    • Colour 
    • Length 
    • Copy
    • Placement 

    You can even test the impact of removing CTA banners and using text-based CTAs on your conversion rates.

    You should test out personalising CTAs, too. Research shows that personalised CTAs perform 202% better than standard calls to action. 

    Revise your web copy

    You can use several strategies to improve your website’s copy and generate more conversions. 

    Optimising copy for search engines can increase traffic and generate more conversions, for example. But that shouldn’t make your copy any less impactful. Bear search engines in mind, by all means, but make sure you are speaking to the needs and desires of your potential customers. Your copy needs to convince users that your product can solve their problems. 

    Nowhere is this more important than your headlines. These will be the first thing users read, so make sure they sell your USP and highlight pain points.

    Don’t just guess at the kind of messaging that will move the needle, however. Constantly test new headlines and continue doing so even after you’ve started seeing success. The results may surprise you. TruckersReport, a site that helps people become truck drivers, boosted opt-ins by 21.7% by revising its landing page headline, among other changes. 

    Make sure there are no spelling mistakes in your copy, either. Misspelt words, poor grammar and bad formatting make your website look unprofessional and untrustworthy. Even if the rest of your copy is incredibly enticing, these rookie errors can be enough to turn customers off. 

    Simplify your site’s navigation

    A website’s navigation is an often overlooked factor in conversion rate optimisation, but simplifying it can make it much easier for users to take action. 

    If you’ve ever used a poorly designed e-commerce store, you know how confusing and overwhelming bad navigation can be. Research shows that a whopping 82% of stores don’t divide their navigation into manageable chunks. 

    The trick is to simplify your navigation as much as possible. As you can see in the screenshot below, our navigation only has five headers and a call to action. It’s easy to find exactly what you’re looking for, and you can’t miss the big green CTA button. 

    A screenshot of the navigation menu on Matomo

    Alternatively, you can test what happens when you completely remove your navigation. Brands usually do this on landing pages where the only action they want the user to take is to make a purchase. 

    It’s exactly the strategy we’ve used on our free trial landing page. 

    Leverage heatmaps

    Analytics tools — and heatmaps in particular — can help you understand user behaviour and optimise accordingly. 

    Heatmaps are a visual representation of user interaction on your page. Red and yellow represent high levels of user interaction, and blue and green represent low levels of interaction.

    Screenshot of Matomo heatmap feature

    As you can see in the screenshot above, our CTA button has some of the highest levels of engagement on the page, telling us that it’s well-positioned. Given the focus on the site’s navigation, we can also assume we are correct to have a CTA button in there — something we can confirm using our web analytics to see how many users click on it.

    Reduce load time

    Speed matters when it comes to conversions. Fact. 

    Research shows a huge difference in conversion rates between quick and slow sites. For example, a site that loads in one second converts three times better than a site that loads in five seconds. 

    That’s why using a web analytics tool is vital to understand page load times and act accordingly if you think slow speeds are hampering your conversions.

    A screenshot of page load times in Matomo

    Identifying your slowest pages is easy with Matomo. Just sort your pages by the Avg. Use the page load time metric on the page performance report to identify the pages you want to drive conversions. 

    Next, take steps to improve your page’s load time by :

    • Compressing images
    • Compressing code files or using a more lightweight theme
    • Removing unnecessary plugins
    • Using a content delivery network
    • Improving your hosting

    Try Matomo for Free

    Get the web insights you need, without compromising data accuracy.

    No credit card required

    Add more trust signals

    Trust is essential when you’re trying to convince customers to make a purchase. In fact, consumers rate trust as one of the top three buying factors, far above a brand’s reputation and whether they love the brand. 

    Adding trust signals to your landing pages, such as customer testimonials, customer reviews, case studies, and other forms of social proof, can transform your conversion rates. If consumers see real people and businesses buy from you, they’ll feel reassured to do the same.

    Trust signals on the Matomo website

    It’s a strategy we use ourselves. Just look at the screenshot from our homepage above. Immediately after our free trial CTA, we display the logos of well-known brands that use our product. 

    Security-focused trust signals are also powerful if you are an online store. Installing an SSL certificate, showing logos of trusted payment providers (like PayPal and Mastercard) can convince people they are spending money at a legitimate store.

    Improve your site’s mobile experience

    More and more people are accessing the internet via their smartphones. In 2022, for instance, there were five billion unique mobile Internet users, meaning more than 60% of the internet population used a smartphone to browse online. 

    Moreover, 76% of U.S. adults make purchases using their smartphones. 

    That means you need to ensure your site’s mobile experience is on-point to increase conversions. 

    Your site should use a mobile-first design, meaning it works perfectly on smartphones and then scales up for desktop users. 

    Trust the data

    Opinions are a fantastic form of inspiration for new A/B tests. But they should never be trusted over cold, hard data. If your test shows the opposite of what you and your team thought would happen, then trust the data and not yourself.

    With that in mind, ensure you collect qualitative and quantitative data during your experiments. Web analytics should always form the backbone of conversion tests, but don’t forget to also use heatmaps, screen recordings, and customer surveys. 

    Keep testing

    There’s no such word as “finished” in the world of A/B testing. Continual testing is key if you want to convert more website visitors. 

    Make sure you aren’t stopping tests prematurely, either. Make sure every A/B and multivariate test reaches a sample size that makes the test statistically significant. 

    Understand your users better with Matomo 

    Whether you run an e-commerce store, a SaaS company, or a service-based business, implementing these conversion rate optimisation best practices could be an easy way to lower your bounce rate and boost your conversion rates.

    But remember, best practices aren’t clear-cut rules. What works for one website may not work for yours. That’s why running your own tests and understanding your visitors’ behaviour is important. 

    Matomo’s web analytics platform is the perfect tool for doing just that. Not only does it come with the tools you need to optimise your conversion rate (like an A/B testing tool, heatmaps and session recordings), but you can also trust the data. Unlike Google Analytics 4 and other tools, Matomo doesn’t use data sampling meaning you have 100% accurate data from which to make better decisions. It’s GDPR compliant and can run cookieless, so no need for cookie consent banners (excluding in the UK and Germany).

    Discover how you can improve your website’s conversions with Matomo by starting a free 21-day trial, no credit card required.