Recherche avancée

Sur d’autres sites (4563)

• 10 Key Google Analytics Limitations You Should Be Aware Of

  9 mai 2022, par Erin

  Google Analytics (GA) is the biggest player in the web analytics space. But is it as “universal” as its brand name suggests ?
  

  Over the years users have pointed out a number of major Google Analytics limitations. Many of these are even more visible in Google Analytics 4. 

  Introduced in 2020, Google Analytics 4 (GA4) has been sceptically received. As the sunset date of 1st, July 2023 for the current version, Google Universal Analytics (UA), approaches, the dismay grows stronger.

  To the point where people are pleading with others to intervene : 

  

  Main limitations of Google Analytics

  Google Analytics 4 is advertised as a more privacy-centred, comprehensive and “intelligent” web analytics platform. 

  According to Google, the newest version touts : 

  • Machine learning at its core provides better segmentation and fast-track access to granular insights 
  • Privacy-by-design controls, addressing restrictions on cookies and new regulatory demands 
  • More complete understanding of customer journeys across channels and devices 

  Some of these claims hold true. Others crumble upon a deeper investigation. Newly advertised Google Analytics capabilities such as ‘custom events’, ‘predictive insights’ and ‘privacy consent mode’ only have marginal improvements. 

  Complex setup, poor UI and lack of support with migration also leave many other users frustrated with GA4. 

  

  Let’s unpack all the current (and legacy) limitations of Google Analytics you should account for. 

  1. No Historical Data Imports 

  Google rushed users to migrate from Universal Analytics to Google Analytics 4. But they overlooked one important precondition — backwards compatibility. 

  You have no way to import data from Google Universal Analytics to Google Analytics 4. 

  Historical records are essential for analysing growth trends and creating benchmarks for new marketing campaigns. Effectively, you are cut short from past insights — and forced to start strategising from scratch. 

  At present, Google offers two feeble solutions : 

  • Run data collection in parallel and have separate reporting for GA4 and UA until the latter is shut down. Then your UA records are gone. 
  • For Ecommerce data, manually duplicate events from UA at a new GA4 property while trying to figure out the new event names and parameters. 

  Google’s new data collection model is the reason for migration difficulties. 

  In Google Analytics 4, all analytics hits types — page hits, social hits, app/screen view, etc. — are recorded as events. Respectively, the “‘event’ parameter in GA4 is different from one in Google Universal Analytics as the company explains : 

  

  This change makes migration tedious — and Google offers little assistance with proper events and custom dimensions set up. 

  2. Data Collection Limits 

  If you’ve wrapped your head around new GA4 events, congrats ! You did a great job, but the hassle isn’t over. 

  You still need to pay attention to new Google Analytics limits on data collection for event parameters and user properties. 

  

  These apply to :

  • Automatically collected events
  • Enhanced measurement events
  • Recommended events 
  • Custom events 

  When it comes to custom events, GA4 also has a limit of 25 custom parameters per event. Even though it seems a lot, it may not be enough for bigger websites. 

  You can get higher limits by upgrading to Google Analytics 360, but the costs are steep. 

  3. Limited GDPR Compliance 

  Google Analytics has a complex history with European GDPR compliance. 

  A 2020 ruling by the Court of Justice of the European Union (CJEU) invalidated the Privacy Shield framework Google leaned upon. This framework allowed the company to regulate EU-US data transfers of sensitive user data. 

  But after this loophole was closed, Google faced a heavy series of privacy-related fines :

  • French data protection authority, CNIL, ruled that  “the transfers to the US of personal data collected through Google Analytics are illegal” — and proceeded to fine Google for a record-setting €150 million at the beginning of 2022. 
  • Austrian regulators also deemed Google in breach of GDPR requirements and also branded the analytics as illegal. 

  Other EU-member states might soon proceed with similar rulings. These, in turn, can directly affect Google Analytics users, whose businesses could face brand damage and regulatory fines for non-compliance. In fact, companies cannot select where the collected analytics data will be stored — on European servers or abroad — nor can they obtain this information from Google.

  Getting a web analytics platform that allows you to keep data on your own servers or select specific Cloud locations is a great alternative. 

  Google also has been lax with its cookie consent policy and doesn’t properly inform consumers about data collection, storage or subsequent usage. Google Analytics 4 addresses this issue to an extent. 

  By default, GA4 relies on first-party cookies, instead of third-party ones — which is a step forward. But the user privacy controls are hard to configure without losing most of the GA4 functionality. Implementing user consent mode to different types of data collection also requires a heavy setup. 

  4. Strong Reliance on Sampled Data 

  To compensate for ditching third-party cookies, GA4 more heavily leans on sampled data and machine learning to fill the gaps in reporting. 

  In GA4 sampling automatically applies when you :

  • Perform advanced analysis such as cohort analysis, exploration, segment overlap or funnel analysis with not enough data 
  • Have over 10,000,000 data rows and generate any type of non-default report 

  Google also notes that data sampling can occur at lower thresholds when you are trying to get granular insights. If there’s not enough data or because Google thinks it’s too complex to retrieve. 

  In their words :

  

  Data sampling adds “guesswork” to your reports, meaning you can’t be 100% sure of data accuracy. The divergence from actual data depends on the size and quality of sampled data. Again, this isn’t something you can control. 

  Unlike Google Analytics 4, Matomo applies no data sampling. Your reports are always accurate and fully representative of actual user behaviours. 

  5. No Proper Data Anonymization 

  Data anonymization allows you to collect basic analytics about users — visits, clicks, page views — but without personally identifiable information (or PII) such as geo-location, assigns tracking ID or other cookie-based data. 

  This reduced your ability to :

  • Remarket 
  • Identify repeating visitors
  • Do advanced conversion attribution 

  But you still get basic data from users who ignored or declined consent to data collection. 

  By default, Google Analytics 4 anonymizes all user IP addresses — an upgrade from UA. However, it still assigned a unique user ID to each user. These count as personal data under GDPR. 

  For comparison, Matomo provides more advanced privacy controls. You can anonymize :

  • Previously tracked raw data 
  • Visitor IP addresses
  • Geo-location information
  • User IDs 

  This can ensure compliance, especially if you operate in a sensitive industry — and delight privacy-mindful users ! 

  6. No Roll-Up Reporting

  Getting a bird’s-eye view of all your data is helpful when you need hotkey access to main sites — global traffic volume, user count or percentage of returning visitors.

  With Roll-Up Reporting, you can see global-performance metrics for multiple localised properties (.co.nz, .co.uk, .com, etc,) in one screen. Then zoom in on specific localised sites when you need to. 

  

  7. Report Processing Latency 

  The average data processing latency is 24-48 hours with Google Analytics. 

  Accounts with over 200,000 daily sessions get data refreshes only once a day. So you won’t be seeing the latest data on core metrics. This can be a bummer during one-day promo events like Black Friday or Cyber Monday when real-time information can prove to be game-changing ! 

  Matomo processes data with lower latency even for high-traffic websites. Currently, we have 6-24 hour latency for cloud deployments. On-premises web analytics can be refreshed even faster — within an hour or instantly, depending on the traffic volumes. 

  8. No Native Conversion Optimisation Features

  Google Analytics users have to use third-party tools to get deeper insights like how people are interacting with your webpage or call-to-action.

  You can use the free Google Optimize tool, but it comes with limits : 

  • No segmentation is available 
  • Only 10 simultaneous running experiments allowed 

  There isn’t a native integration between Google Optimize and Google Analytics 4. Instead, you have to manually link an Optimize Container to an analytics account. Also, you can’t select experiment dimensions in Google Analytics reports.

  What’s more, Google Optimize is a basic CRO tool, best suited for split testing (A/B testing) of copy, visuals, URLs and page layouts. If you want to get more advanced data, you need to pay for extra tools. 

  Matomo comes with a native set of built-in conversion optimization features : 

  • Heatmaps 
  • User session recording 
  • Sales funnel analysis 
  • A/B testing 
  • Form submission analytics 

  

  9. Deprecated Annotations

  Annotations come in handy when you need to provide extra context to other team members. For example, point out unusual traffic spikes or highlight a leak in the sales funnel. 

  This feature was available in Universal Analytics but is now gone in Google Analytics 4. But you can still quickly capture, comment and share knowledge with your team in Matomo. 

  You can add annotations to any graph that shows statistics over time including visitor reports, funnel analysis charts or running A/B tests. 

  10. No White Label Option 

  This might be a minor limitation of Google Analytics, but a tangible one for agency owners. 

  Offering an on-brand, embedded web analytics platform can elevate your customer experience. But white label analytics were never a thing with Google Analytics, unlike Matomo. 

  Wrap Up 

  Google set a high bar for web analytics. But Google Analytics inherent limitations around privacy, reporting and deployment options prompt more users to consider Google Analytics alternatives, like Matomo. 

  With Matomo, you can easily migrate your historical data records and store customer data locally or in a designated cloud location. We operate by a 100% unsampled data principle and provide an array of privacy controls for advanced compliance. 

  Start your 21-day free trial (no credit card required) to see how Matomo compares to Google Analytics ! 

  21 day free trial. No credit card required.

  Your email address
  

  Your website address
  

  Your Analytics subdomain will be .matomo.cloud (change)

  Choose your analytics subdomain
  

  .matomo.cloud

  I hereby accept the terms and conditions and the data processing agreeement (DPA).

  
  Your information will be used to create an account on our cloud service. For more information please consult our privacy policy.
  
  

  » Start improving your websites now

  

• opencv does not find ffmpeg functions during compilation (make)

  17 avril 2022, par titicplusplus

  I am currently trying to compile OpenCV with CUDA.
So I downloaded opencv 4.5.5 and opencv_contrib and followed this tutorial : https://gist.github.com/raulqf/f42c718a658cddc16f9df07ecc627be7

  


  cd opencv-4.5.5/
mkdir build
cd build
cmake -D CMAKE_BUILD_TYPE=RELEASE \
-D CMAKE_INSTALL_PREFIX=/usr/local \
-D WITH_TBB=ON \
-D ENABLE_FAST_MATH=1 \
-D CUDA_FAST_MATH=1 \
-D WITH_CUBLAS=1 \
-D WITH_CUDA=ON \
-D BUILD_opencv_cudacodec=OFF \
-D WITH_CUDNN=OFF \
-D OPENCV_DNN_CUDA=OFF \
-D CUDA_ARCH_BIN=7.5 \
-D WITH_V4L=ON \
-D WITH_QT=ON \
-D WITH_OPENGL=ON \
-D WITH_GSTREAMER=ON \
-D OPENCV_GENERATE_PKGCONFIG=ON \
-D OPENCV_PC_FILE_NAME=opencv.pc \
-D OPENCV_ENABLE_NONFREE=ON \
-D INSTALL_PYTHON_EXAMPLES=OFF \
-D INSTALL_C_EXAMPLES=OFF \
-D BUILD_EXAMPLES=OFF \
-D OPENCV_EXTRA_MODULES_PATH=../../opencv_contrib-4.5.5/modules ../


  


  The cmake command generated these lines :

  


  -- General configuration for OpenCV 4.5.5 =====================================
--   Version control:               unknown
-- 
--   Extra modules:
--     Location (extra):            /mnt/704E048C4E044D72/build/opencv/opencv_contrib-4.5.5/modules
--     Version control (extra):     unknown
-- 
--   Platform:
--     Timestamp:                   2022-04-17T16:01:44Z
--     Host:                        Linux 5.4.0-107-lowlatency x86_64
--     CMake:                       3.16.3
--     CMake generator:             Unix Makefiles
--     CMake build tool:            /usr/bin/make
--     Configuration:               RELEASE
-- 
--   CPU/HW features:
--     Baseline:                    SSE SSE2 SSE3
--       requested:                 SSE3
--     Dispatched code generation:  SSE4_1 SSE4_2 FP16 AVX AVX2 AVX512_SKX
--       requested:                 SSE4_1 SSE4_2 AVX FP16 AVX2 AVX512_SKX
--       SSE4_1 (18 files):         + SSSE3 SSE4_1
--       SSE4_2 (2 files):          + SSSE3 SSE4_1 POPCNT SSE4_2
--       FP16 (1 files):            + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 AVX
--       AVX (5 files):             + SSSE3 SSE4_1 POPCNT SSE4_2 AVX
--       AVX2 (33 files):           + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2
--       AVX512_SKX (8 files):      + SSSE3 SSE4_1 POPCNT SSE4_2 FP16 FMA3 AVX AVX2 AVX_512F AVX512_COMMON AVX512_SKX
-- 
--   C/C++:
--     Built as dynamic libs?:      YES
--     C++ standard:                11
--     C++ Compiler:                /usr/bin/c++  (ver 8.4.0)
--     C++ flags (Release):         -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -O3 -DNDEBUG  -DNDEBUG
--     C++ flags (Debug):           -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=non-virtual-dtor -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wundef -Winit-self -Wpointer-arith -Wshadow -Wsign-promo -Wuninitialized -Wsuggest-override -Wno-delete-non-virtual-dtor -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -fvisibility-inlines-hidden -g  -O0 -DDEBUG -D_DEBUG
--     C Compiler:                  /usr/bin/cc
--     C flags (Release):           -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -O3 -DNDEBUG  -DNDEBUG
--     C flags (Debug):             -fsigned-char -ffast-math -W -Wall -Werror=return-type -Werror=address -Werror=sequence-point -Wformat -Werror=format-security -Wmissing-declarations -Wmissing-prototypes -Wstrict-prototypes -Wundef -Winit-self -Wpointer-arith -Wshadow -Wuninitialized -Wno-comment -Wimplicit-fallthrough=3 -Wno-strict-overflow -fdiagnostics-show-option -Wno-long-long -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections  -msse -msse2 -msse3 -fvisibility=hidden -g  -O0 -DDEBUG -D_DEBUG
--     Linker flags (Release):      -Wl,--exclude-libs,libippicv.a -Wl,--exclude-libs,libippiw.a   -Wl,--gc-sections -Wl,--as-needed  
--     Linker flags (Debug):        -Wl,--exclude-libs,libippicv.a -Wl,--exclude-libs,libippiw.a   -Wl,--gc-sections -Wl,--as-needed  
--     ccache:                      NO
--     Precompiled headers:         NO
--     Extra dependencies:          m pthread cudart_static dl rt nppc nppial nppicc nppicom nppidei nppif nppig nppim nppist nppisu nppitc npps cublas cufft -L/usr/lib/x86_64-linux-gnu
--     3rdparty dependencies:
-- 
--   OpenCV modules:
--     To be built:                 alphamat aruco barcode bgsegm bioinspired calib3d ccalib core cudaarithm cudabgsegm cudafeatures2d cudafilters cudaimgproc cudalegacy cudaobjdetect cudaoptflow cudastereo cudawarping cudev cvv datasets dnn dnn_objdetect dnn_superres dpm face features2d flann freetype fuzzy gapi hdf hfs highgui img_hash imgcodecs imgproc intensity_transform line_descriptor mcc ml objdetect optflow phase_unwrapping photo plot python2 python3 quality rapid reg rgbd saliency sfm shape stereo stitching structured_light superres surface_matching text tracking ts video videoio videostab wechat_qrcode xfeatures2d ximgproc xobjdetect xphoto
--     Disabled:                    cudacodec world
--     Disabled by dependency:      -
--     Unavailable:                 java julia matlab ovis viz
--     Applications:                tests perf_tests apps
--     Documentation:               NO
--     Non-free algorithms:         YES
-- 
--   GUI:                           QT5
--     QT:                          YES (ver 5.12.8 )
--       QT OpenGL support:         YES (Qt5::OpenGL 5.12.8)
--     GTK+:                        YES (ver 3.24.20)
--       GThread :                  YES (ver 2.64.6)
--       GtkGlExt:                  NO
--     OpenGL support:              YES (/usr/lib/x86_64-linux-gnu/libGL.so /usr/lib/x86_64-linux-gnu/libGLU.so)
--     VTK support:                 NO
-- 
--   Media I/O: 
--     ZLib:                        /usr/lib/x86_64-linux-gnu/libz.so (ver 1.2.11)
--     JPEG:                        /usr/lib/x86_64-linux-gnu/libjpeg.so (ver 80)
--     WEBP:                        build (ver encoder: 0x020f)
--     PNG:                         /usr/lib/x86_64-linux-gnu/libpng.so (ver 1.6.37)
--     TIFF:                        /usr/lib/x86_64-linux-gnu/libtiff.so (ver 42 / 4.1.0)
--     JPEG 2000:                   build (ver 2.4.0)
--     OpenEXR:                     /usr/lib/x86_64-linux-gnu/libImath.so /usr/lib/x86_64-linux-gnu/libIlmImf.so /usr/lib/x86_64-linux-gnu/libIex.so /usr/lib/x86_64-linux-gnu/libHalf.so /usr/lib/x86_64-linux-gnu/libIlmThread.so (ver 2_3)
--     HDR:                         YES
--     SUNRASTER:                   YES
--     PXM:                         YES
--     PFM:                         YES
-- 
--   Video I/O:
--     DC1394:                      YES (2.2.5)
--     FFMPEG:                      YES
--       avcodec:                   YES (58.54.100)
--       avformat:                  YES (58.29.100)
--       avutil:                    YES (56.31.100)
--       swscale:                   YES (5.5.100)
--       avresample:                YES (4.0.0)
--     GStreamer:                   YES (1.16.2)
--     v4l/v4l2:                    YES (linux/videodev2.h)
-- 
--   Parallel framework:            TBB (ver 2020.1 interface 11101)
-- 
--   Trace:                         YES (with Intel ITT)
-- 
--   Other third-party libraries:
--     Intel IPP:                   2020.0.0 Gold [2020.0.0]
--            at:                   /mnt/704E048C4E044D72/build/opencv/opencv-4.5.5/build/3rdparty/ippicv/ippicv_lnx/icv
--     Intel IPP IW:                sources (2020.0.0)
--               at:                /mnt/704E048C4E044D72/build/opencv/opencv-4.5.5/build/3rdparty/ippicv/ippicv_lnx/iw
--     VA:                          NO
--     Lapack:                      NO
--     Eigen:                       YES (ver 3.3.7)
--     Custom HAL:                  NO
--     Protobuf:                    build (3.19.1)
-- 
--   NVIDIA CUDA:                   YES (ver 10.1, CUFFT CUBLAS FAST_MATH)
--     NVIDIA GPU arch:             75
--     NVIDIA PTX archs:
-- 
--   OpenCL:                        YES (no extra features)
--     Include path:                /mnt/704E048C4E044D72/build/opencv/opencv-4.5.5/3rdparty/include/opencl/1.2
--     Link libraries:              Dynamic load
-- 
--   Python 2:
--     Interpreter:                 /usr/bin/python2.7 (ver 2.7.18)
--     Libraries:                   /usr/lib/x86_64-linux-gnu/libpython2.7.so (ver 2.7.18)
--     numpy:                       /usr/lib/python2.7/dist-packages/numpy/core/include (ver 1.16.5)
--     install path:                lib/python2.7/dist-packages/cv2/python-2.7
-- 
--   Python 3:
--     Interpreter:                 /usr/bin/python3 (ver 3.8.10)
--     Libraries:                   /usr/lib/x86_64-linux-gnu/libpython3.8.so (ver 3.8.10)
--     numpy:                       /home/famillevincent/.local/lib/python3.8/site-packages/numpy/core/include (ver 1.22.3)
--     install path:                lib/python3.8/site-packages/cv2/python-3.8
-- 
--   Python (for build):            /usr/bin/python2.7
-- 
--   Java:                          
--     ant:                         NO
--     JNI:                         /usr/lib/jvm/default-java/include /usr/lib/jvm/default-java/include/linux /usr/lib/jvm/default-java/include
--     Java wrappers:               NO
--     Java tests:                  NO
-- 
--   Install to:                    /usr/local
-- -----------------------------------------------------------------
-- 
-- Configuring done
-- Generating done


  


  CMake have detected ffmpeg libraries, but when I run make -j8. I have this error :

  


  [ 39%] Building CXX object apps/interactive-calibration/CMakeFiles/opencv_interactive-calibration.dir/calibPipeline.cpp.o
Scanning dependencies of target opencv_cudafilters
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwframe_transfer_data »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « avcodec_get_hw_config »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwdevice_get_hwframe_constraints »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « avcodec_send_packet »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwframe_get_buffer »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwdevice_ctx_create_derived »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwdevice_ctx_create »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_bsf_alloc »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_bsf_receive_packet »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_bsf_free »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « avcodec_send_frame »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « avcodec_parameters_copy »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_packet_free »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwdevice_find_type_by_name »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_bsf_init »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwframe_ctx_alloc »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « avcodec_receive_packet »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_codec_iterate »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwframe_ctx_init »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwdevice_get_type_name »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwframe_constraints_free »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « avcodec_receive_frame »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_bsf_get_by_name »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « avcodec_get_hw_frames_parameters »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_bsf_send_packet »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_hwframe_ctx_create_derived »
/usr/bin/ld : ../../lib/libopencv_videoio.so.4.5.5 : référence indéfinie vers « av_packet_alloc »
collect2: error: ld returned 1 exit status
make[2]: *** [apps/visualisation/CMakeFiles/opencv_visualisation.dir/build.make:92 : bin/opencv_visualisation] Erreur 1
make[1]: *** [CMakeFiles/Makefile2:11412 : apps/visualisation/CMakeFiles/opencv_visualisation.dir/all] Erreur 2
make[1]: *** Attente des tâches non terminées....
[ 39%] Building CXX object apps/interactive-calibration/CMakeFiles/opencv_interactive-calibration.dir/frameProcessor.cpp.o


  


  So what can I do to compile opencv with cuda and ffmpeg ?
  
Thank you in advance for your answers.

  


  I use Ubuntu 20.04 with g++8

  


• How to Use Web Analytics to Improve SEO

  5 janvier 2022, par erin — Analytics Tips

  Everyone wants their website to rank highly in Google — and that’s exactly why the world of SEO is so competitive.

  In order to succeed in such a crowded space, it’s essential to equip yourself with the right tools and processes to ensure your website is maximally optimised for search engines.

  If you’d like to improve your website’s SEO rankings, leveraging web analytics is one of the best places to start. Web analytics provides valuable insights to help you assess performance, user behaviour and optimisation opportunities.

  In this blog, we’ll cover :

  The basics of SEO and web analytics

  Before we discuss how to use web analytics for SEO, let’s start with a quick explanation of both.

  SEO (Search Engine Optimisation) encompasses a broad set of activities aimed at increasing a website’s position in search engine results pages (SERPs). When a user enters a query (e.g. ‘marketing agencies in Dallas’) in a search engine, the websites that appear near the top of the page are optimised for search engines and therefore ranking for that particular term. 

  Web analytics refers to the monitoring/assessment of metrics that track traffic sources and user behaviour on a website. This involves the use of a web analytics tool to collect, aggregate, organise and visualise website data so that meaningful conclusions can be drawn.

  The importance of website analytics for SEO

  SEO revolves around search engine algorithms – a set of rules that dictates a website’s ranking for a given search query (i.e. keyword). The algorithm takes numerous factors into account to determine a particular site’s SERP ranking. So, to achieve strong SEO, your website needs to exhibit qualities that the algorithm deems important. That’s where web analytics comes into play.

  Web analytics allows you to track key metrics and data points that affect how the algorithm ranks your website. For example, how much time do users spend on your site ? Which external links are referring traffic to your site ? How do your site’s Core Web Vitals stack up ? 

  Understanding this data will supply you with the insights needed to make positive adjustments, ultimately improving your website’s SEO. 

  How do you analyse a website for SEO ?

  The SEO analysis of a website needs to be focused on relevant data that’s applicable to search engine rankings. When conducting your website SEO analysis, here are some notable metrics and data fields to pay attention to :

  1. Bounce rate and dwell time

  These metrics denote how much time users spend on your website. If users frequently exit your site after only a few seconds, Google may view this as a negative indicator. To reduce bounce rate and increase dwell time, you should work towards making your site’s content more captivating and ensuring that there aren’t any technical issues with your site (e.g. pages taking too long to load or not optimised for mobile).

  

  2. Broken/dead links

  Perform a technical analysis to scan your website for faulty links. If your site contains broken links that lead to 404 pages, this can detract from your website’s SEO rankings. Redirect those links to a related page or remove them.

  

  Matomo’s Crawling Errors report can give you instant access to this technical information so you can resolve it before it begins to impact your ranking.

  3. Scroll depth

  Measuring scroll depth (how far users scroll down the page) can help you gauge the quality of your content — and this goes hand-in-hand with bounce rate and dwell time. To assess scroll depth, you can use a Tag Manager to track the specific scroll percentage on your site’s pages.

  4. Transitions

  Studying how users transition from page to page within your site can help you understand their behaviour more holistically. Which pages do they tend to gravitate towards ? Are there CTAs on your blog that aren’t driving many click-throughs ? Optimising user journeys will, in turn, elevate the overall user experience on your site.

  

  5. Internal site search

  You can use site search tracking and reporting to learn what your audience is looking for. If you notice a trend (e.g. the majority of searches are for pricing because your pricing page isn’t in the navigation menu), this can inform both site architecture and content planning.

  

  Ecommerce sites in particular should be monitoring branded queries, especially in regards to brand misspellings that could be causing users to bounce off the site.

  6. Segments

  Separating your visitors into distinct segments can produce granular insights that paint a more accurate picture.

  For example, perhaps you notice that your bounce rate is far higher on mobile, or with users from the UK. In both cases, this knowledge will provide clarity on where to focus your optimisation efforts (e.g. mobile responsiveness, UK-specific content/landing pages, etc.).

  

  7. Acquisition channels

  It’s crucial to analyse where your website traffic is coming from. Among other things, reviewing your acquisition metrics will reveal which external websites are referring the most traffic to your website. 

  Links from external sites (also known as backlinks) are one of the most important ranking factors because this tells Google that your site is reputable and credible. So, you may choose to cultivate a relationship with these sites (or similar sites) by offering guest blogging and other link building initiatives.

  

  In addition to the above, you should also be monitoring your Core Web Vitals — which leads us to our next section.

  What are Core Web Vitals and why are they important ?

  Core Web Vitals are a set of 3 primary metrics that reflect the general user experience of a website. These metrics are load time, interactivity and stability. 

  1. Load time (LCP) refers to the amount of time it takes for your website’s text and images to load.
  2. Interactivity (FID) refers to the amount of time it takes for user input areas (buttons, form fields, etc.) to become functional.
  3. Stability (CLS) refers to the visual/spatial integrity of your website. If text, images, and other elements tend to suddenly shift position when a user is viewing the site, this will hurt your CLS score.

  

  So, why are these Core Web Vitals metrics important for SEO ? Generally speaking, Google prioritises user experience — and Core Web Vitals affect users’ satisfaction with a website. Furthermore, Google has confirmed that Core Web Vitals are, indeed, a ranking factor.

  Matomo enables you to track metrics for Core Web Vitals which we refer to as SEO Web Vitals.

  How to measure and track keyword performance

  We can’t talk about SEO and analytics without touching on keywords. Keywords (the words/phrases that users type in a search engine) are arguably the most cardinal component of SEO. So, outside of website performance, it’s also necessary to track the keywords your website is ranking for. 

  Recall from above that SEO is all about ranking highly on SERPs for certain search queries (i.e. keywords). To assess your Search Engine Keyword Performance, you can use an analytics tool to view Keyword reports for your website. These reports will show you which keywords your site ranks for, the average SERP position your site achieves for each keyword, the amount of traffic you receive from each keyword, and more.

  

  Digging into your keyword performance can help you identify valuable keyword opportunities and improvement goals.

  For example, upon reviewing your highest-traffic keywords, you may choose to create more blog content around those keywords to bolster your success. Or, perhaps you notice that your average position for a high-intent keyword is quite low. In that case, you could implement a targeted link building campaign to help boost your ranking for that keyword. 

  Final thoughts

  In this article, we’ve discussed the benefits of web analytics — particularly in regards to SEO. When it comes to selecting a web analytics tool, Google Analytics is by far the most popular choice. But that doesn’t make it the best.

  At Matomo, we’re committed to providing a superior alternative to Google Analytics. Matomo is a powerful, open-source web analytics platform that gives you 100% data ownership — protecting both your data and your customers’ privacy.

  Try our live demo or start a free 21-day trial now – no credit card required.