{"id":2621,"date":"2024-07-22T13:35:04","date_gmt":"2024-07-22T13:35:04","guid":{"rendered":"https:\/\/ewebtoolz.com\/blog\/can-you-spot-google-updates-with-xmr-charts\/"},"modified":"2024-07-22T13:35:04","modified_gmt":"2024-07-22T13:35:04","slug":"can-you-spot-google-updates-with-xmr-charts","status":"publish","type":"post","link":"https:\/\/ewebtoolz.com\/blog\/can-you-spot-google-updates-with-xmr-charts\/","title":{"rendered":"Can You Spot Google Updates with XmR Charts?"},"content":{"rendered":"<p> <br \/>\n<\/p>\n<div id=\"\">\n<p>Website traffic data normally looks like\u00a0this:<\/p>\n<figure class=\"wp-block-image size-full\"><img fetchpriority=\"high\" decoding=\"async\" width=\"1647\" height=\"1014\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_14-05-58.png\" alt=\"\" class=\"wp-image-177422\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_14-05-58.png 1647w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_14-05-58-680x419.png 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_14-05-58-768x473.png 768w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_14-05-58-1536x946.png 1536w\" sizes=\"(max-width: 1647px) 100vw, 1647px\"\/><\/figure>\n<p>Ups and downs, peaks and troughs.<\/p>\n<p>If we\u2019re doing our job properly, we generally expect traffic to trend upwards over time, but in any given month, it\u2019s difficult to say whether a peak or a trough is worth paying attention to.<\/p>\n<p>Did we do something great and trigger a new phase of growth? Did we benefit from a new Google update? Or is it just normal variation, part of the natural ebb and flow of people finding our website?<\/p>\n<p>Or suppose you make a change to your content process\u2014you pruned and redirected a bunch of old content\u2014and then traffic dropped the next month. Was that drop <em>caused <\/em>by the change, or was it just a coincidence?<\/p>\n<p>I\u2019ve been experimenting with a simple statistical tool designed to help answer these questions: <strong>XmR charts<\/strong>, also known as process control charts.<\/p>\n<div class=\"post-nav-link clearfix\" id=\"section1\"><a class=\"subhead-anchor\" data-tip=\"tooltip__copielink\" rel=\"#section1\"><svg width=\"19\" height=\"19\" viewbox=\"0 0 14 14\" style=\"\"><g fill=\"none\" fill-rule=\"evenodd\"><path d=\"M0 0h14v14H0z\"\/><path d=\"M7.45 9.887l-1.62 1.621c-.92.92-2.418.92-3.338 0a2.364 2.364 0 0 1 0-3.339l1.62-1.62-1.273-1.272-1.62 1.62a4.161 4.161 0 1 0 5.885 5.884l1.62-1.62L7.45 9.886zM5.527 5.135L7.17 3.492c.92-.92 2.418-.92 3.339 0 .92.92.92 2.418 0 3.339L8.866 8.473l1.272 1.273 1.644-1.643A4.161 4.161 0 1 0 5.897 2.22L4.254 3.863l1.272 1.272zm-.66 3.998a.749.749 0 0 1 0-1.06l2.208-2.206a.749.749 0 1 1 1.06 1.06L5.928 9.133a.75.75 0 0 1-1.061 0z\" style=\"\"\/><\/g><\/svg><\/a><\/p>\n<div class=\"link-text\" data-anchor=\"XmR charts (briefly) explained\" data-section=\"xmr-charts-explained\">\n<h2 class=\"wp-block-heading\"><a id=\"post-177378-_pqyhh0dts8lk\"\/>XmR charts (briefly) explained<\/h2>\n<\/div>\n<\/div>\n<p>Here\u2019s an XmR\u00a0chart:<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"2048\" height=\"696\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2.jpg\" alt=\"\" class=\"wp-image-177380\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2.jpg 2048w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-680x231.jpg 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-768x261.jpg 768w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-1536x522.jpg 1536w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-650x220.jpg 650w\" sizes=\"(max-width: 2048px) 100vw, 2048px\"\/><\/noscript><img decoding=\"async\" width=\"2048\" height=\"696\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2.jpg\" alt=\"\" class=\"lazyload wp-image-177380\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2.jpg 2048w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-680x231.jpg 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-768x261.jpg 768w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-1536x522.jpg 1536w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-2-650x220.jpg 650w\" data-sizes=\"(max-width: 2048px) 100vw, 2048px\"\/><\/figure>\n<p>XmR charts are designed to tell you whether any single data point from a time series is likely to be caused by normal fluctuation (\u201croutine variation\u201d) or a sign that something happened and needs to be investigated (\u201cexceptional variation\u201d).<\/p>\n<p>XmR charts consist of an X plot (named after the <em>x-value<\/em>, the \u201cthing\u201d we care about\u2014like widgets produced or sales closed)\u2026<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1036\" height=\"708\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3.png\" alt=\"\" class=\"wp-image-177381\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3.png 1036w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3-622x425.png 622w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3-768x525.png 768w\" sizes=\"(max-width: 1036px) 100vw, 1036px\"\/><\/noscript><img decoding=\"async\" width=\"1036\" height=\"708\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3.png\" alt=\"\" class=\"lazyload wp-image-177381\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3.png 1036w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3-622x425.png 622w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-3-768x525.png 768w\" data-sizes=\"(max-width: 1036px) 100vw, 1036px\"\/><\/figure>\n<p>\u2026and an MR plot (named after the <em>moving range<\/em>, basically the \u201cgap\u201d between each data\u00a0point):<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1026\" height=\"698\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4.png\" alt=\"\" class=\"wp-image-177382\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4.png 1026w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4-625x425.png 625w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4-768x522.png 768w\" sizes=\"(max-width: 1026px) 100vw, 1026px\"\/><\/noscript><img decoding=\"async\" width=\"1026\" height=\"698\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4.png\" alt=\"\" class=\"lazyload wp-image-177382\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4.png 1026w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4-625x425.png 625w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-4-768x522.png 768w\" data-sizes=\"(max-width: 1026px) 100vw, 1026px\"\/><\/figure>\n<p>In its simplest use, if you plot your data on the chart and it wiggles up and down around the central line, without crossing the upper and lower bounds\u2014no problem! These ups and downs likely represent normal variation.<\/p>\n<p>But any points that appear outside the upper or lower bounds (shown in <strong>red<\/strong>) should be treated as anomalies that need to be investigated.<\/p>\n<p>In the X plot above, the time series seems to show routine variation until January 16th, when the first red out-of-bounds point appears.<\/p>\n<figure class=\"wp-block-image size-full\"><noscript><img decoding=\"async\" width=\"1032\" height=\"695\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09.png\" alt=\"\" class=\"wp-image-177418\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09.png 1032w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09-631x425.png 631w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09-768x517.png 768w\" sizes=\"(max-width: 1032px) 100vw, 1032px\"\/><\/noscript><img decoding=\"async\" width=\"1032\" height=\"695\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09.png\" alt=\"\" class=\"lazyload wp-image-177418\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09.png 1032w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09-631x425.png 631w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-40-09-768x517.png 768w\" data-sizes=\"(max-width: 1032px) 100vw, 1032px\"\/><\/figure>\n<p>The XmR chart suggests that <em>something <\/em>happened on the 16th to mess with our production process (for better or for worse). Our job is to investigate why.<\/p>\n<div class=\"sidenote\">\n<p>Sidenote.<\/p>\n<p>The line in the middle is the average value of the dataset; the upper and lower bounds represent 3-standard deviations away from the average (known as <em>three-sigma<\/em>). Any point that falls outside of these upper and lower bounds is very likely to be an anomaly, and not part of the original probability distribution.<\/p>\n<\/div>\n<p>There are other \u201csignals\u201d that the XmR chart can show you (like eight consecutive points on one side of the average line representing another type of exceptional variation)\u2014but I will leave you to investigate those on your own\u00a0time.<\/p>\n<div class=\"post-nav-link clearfix\" id=\"section1\"><a class=\"subhead-anchor\" data-tip=\"tooltip__copielink\" rel=\"#section1\"><svg width=\"19\" height=\"19\" viewbox=\"0 0 14 14\" style=\"\"><g fill=\"none\" fill-rule=\"evenodd\"><path d=\"M0 0h14v14H0z\"\/><path d=\"M7.45 9.887l-1.62 1.621c-.92.92-2.418.92-3.338 0a2.364 2.364 0 0 1 0-3.339l1.62-1.62-1.273-1.272-1.62 1.62a4.161 4.161 0 1 0 5.885 5.884l1.62-1.62L7.45 9.886zM5.527 5.135L7.17 3.492c.92-.92 2.418-.92 3.339 0 .92.92.92 2.418 0 3.339L8.866 8.473l1.272 1.273 1.644-1.643A4.161 4.161 0 1 0 5.897 2.22L4.254 3.863l1.272 1.272zm-.66 3.998a.749.749 0 0 1 0-1.06l2.208-2.206a.749.749 0 1 1 1.06 1.06L5.928 9.133a.75.75 0 0 1-1.061 0z\" style=\"\"\/><\/g><\/svg><\/a><\/p>\n<div class=\"link-text\" data-anchor=\"Can you use XmR charts on traffic data?\" data-section=\"xmr-and-traffic-data\">\n<h2 class=\"wp-block-heading\"><a id=\"post-177378-_bh674mptfrps\"\/>Can you use XmR charts on traffic data?<\/h2>\n<\/div>\n<\/div>\n<p>When I started reading about XmR charts, one obvious use came to mind: identifying the impact of Google algorithm updates.<\/p>\n<p>If a site\u2019s traffic tanks to zero, it\u2019s easy to say \u201cwe were hit by a manual penalty.\u201d But for smaller changes, like a few months\u2019 consecutive traffic decline, it\u2019s harder to work out the cause. Did we get caught out by a Google update? Is it seasonality? Or is it just a coincidence, with traffic likely to return to normal in the future?<\/p>\n<p>Here\u2019s two years of monthly organic traffic data for the Ahrefs blog, pulled from <a href=\"https:\/\/ahrefs.com\/site-explorer\">Site Explorer<\/a> and plotted on an XmR\u00a0chart:<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1200\" height=\"800\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5.png\" alt=\"\" class=\"wp-image-177383\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5.png 1200w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5-638x425.png 638w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5-768x512.png 768w\" sizes=\"(max-width: 1200px) 100vw, 1200px\"\/><\/noscript><img decoding=\"async\" width=\"1200\" height=\"800\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5.png\" alt=\"\" class=\"lazyload wp-image-177383\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5.png 1200w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5-638x425.png 638w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-5-768x512.png 768w\" data-sizes=\"(max-width: 1200px) 100vw, 1200px\"\/><\/figure>\n<p>Now\u2026 this is not particularly useful.<\/p>\n<p>There are tons of data points outside the expected range (red), with very few sitting nearer the center line than the quartile limits (orange).<\/p>\n<p>The XmR chart is supposed to show exceptional variation in a consistent process\u2014but in this image, almost all of the data points suggest exceptional variation. What\u00a0gives?<\/p>\n<p>Process charts were designed around simple manufacturing processes, and they work very well when the expected output of a process is constant.<\/p>\n<p>If your goal is manufacturing 10,000 widgets each and every week, an XmR chart will help you work out if that 5,600-widget month was a normal \u201cblip\u201d in routine operation, or caused by a real problem that needs to be investigated.<\/p>\n<p>Website traffic is more complicated. There are tons of variables that impact traffic:<\/p>\n<ul>\n<li><strong>the fluctuating search volume of each\u00a0topic,<\/strong><\/li>\n<li><strong>individual ranking positions<\/strong>,<\/li>\n<li><strong>new competing articles<\/strong>,<\/li>\n<li><strong>search features<\/strong>,<\/li>\n<li><strong>seasonality<\/strong>,<\/li>\n<li><strong>publishing frequency<\/strong>,<\/li>\n<li><strong>Google algorithm updates<\/strong>\u2026<\/li>\n<\/ul>\n<p>That means that running an XmR analysis on a long series of traffic data probably won\u2019t be very helpful. Your \u201cblogging process\u201d is not likely to remain stable for very\u00a0long.<\/p>\n<p>In my case, this particular two-year snapshot of data probably doesn\u2019t come from a single, stable process\u2014there may be multiple probability distributions hidden in\u00a0there.<\/p>\n<div class=\"post-nav-link clearfix\" id=\"section1\"><a class=\"subhead-anchor\" data-tip=\"tooltip__copielink\" rel=\"#section1\"><svg width=\"19\" height=\"19\" viewbox=\"0 0 14 14\" style=\"\"><g fill=\"none\" fill-rule=\"evenodd\"><path d=\"M0 0h14v14H0z\"\/><path d=\"M7.45 9.887l-1.62 1.621c-.92.92-2.418.92-3.338 0a2.364 2.364 0 0 1 0-3.339l1.62-1.62-1.273-1.272-1.62 1.62a4.161 4.161 0 1 0 5.885 5.884l1.62-1.62L7.45 9.886zM5.527 5.135L7.17 3.492c.92-.92 2.418-.92 3.339 0 .92.92.92 2.418 0 3.339L8.866 8.473l1.272 1.273 1.644-1.643A4.161 4.161 0 1 0 5.897 2.22L4.254 3.863l1.272 1.272zm-.66 3.998a.749.749 0 0 1 0-1.06l2.208-2.206a.749.749 0 1 1 1.06 1.06L5.928 9.133a.75.75 0 0 1-1.061 0z\" style=\"\"\/><\/g><\/svg><\/a><\/p>\n<div class=\"link-text\" data-anchor=\"Analyzing multiple probability distributions\" data-section=\"multiple-distributions\">\n<h3 class=\"wp-block-heading\"><a id=\"post-177378-_8nwx2i3hozxg\"\/>Analyzing multiple probability distributions<\/h3>\n<\/div>\n<\/div>\n<p>But we can make the analysis more useful.<\/p>\n<p>The best practice for XmR charts is to limit the analysis to a period of time when you know the process was relatively static, and recalculate it when you suspect something has changed.<\/p>\n<p>Looking at the Moving Range chart for this data below, large amounts of traffic variance happened in November and December. We should investigate possible causes.\u00a0<noscript><img decoding=\"async\" width=\"1023\" height=\"679\" class=\"wp-image-177384\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6.png\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6.png 1023w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6-640x425.png 640w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6-768x510.png 768w\" sizes=\"(max-width: 1023px) 100vw, 1023px\"\/><\/noscript><img decoding=\"async\" width=\"1023\" height=\"679\" class=\"lazyload wp-image-177384\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6.png\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6.png 1023w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6-640x425.png 640w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-6-768x510.png 768w\" data-sizes=\"(max-width: 1023px) 100vw, 1023px\"\/><\/p>\n<p>I know that our publishing frequency was fairly static (we definitely didn\u2019t double our content output). Seasonality would cause a traffic drop, not a spike (we\u2019re writing about SEO, not holiday gift guides).<\/p>\n<p>But there was a big Google update at the start of December:<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1682\" height=\"672\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7.png\" alt=\"\" class=\"wp-image-177385\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7.png 1682w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7-680x272.png 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7-768x307.png 768w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7-1536x614.png 1536w\" sizes=\"(max-width: 1682px) 100vw, 1682px\"\/><\/noscript><img decoding=\"async\" width=\"1682\" height=\"672\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7.png\" alt=\"\" class=\"lazyload wp-image-177385\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7.png 1682w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7-680x272.png 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7-768x307.png 768w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-7-1536x614.png 1536w\" data-sizes=\"(max-width: 1682px) 100vw, 1682px\"\/><figcaption class=\"wp-element-caption\"><a href=\"https:\/\/www.searchenginejournal.com\/google-algorithm-history\/\">Source<\/a><\/figcaption><\/figure>\n<p>If we work on the assumption that <em>something happened <\/em>to our blog process around this time\u2014likely a change to traffic caused by the Google update\u2014we can add a divider to our XmR\u00a0chart.<\/p>\n<p>Instead of trying to analyze our traffic as a single process, we can treat it as two processes, and calculate XmR charts separately:<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1036\" height=\"683\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8.png\" alt=\"\" class=\"wp-image-177386\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8.png 1036w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8-645x425.png 645w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8-768x506.png 768w\" sizes=\"(max-width: 1036px) 100vw, 1036px\"\/><\/noscript><img decoding=\"async\" width=\"1036\" height=\"683\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8.png\" alt=\"\" class=\"lazyload wp-image-177386\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8.png 1036w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8-645x425.png 645w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-8-768x506.png 768w\" data-sizes=\"(max-width: 1036px) 100vw, 1036px\"\/><\/figure>\n<p>Now the first process looks stable (all black dots). The second process shows less extreme variation (red) too, but there\u2019s still too much moderate variation (orange) to look stable. There may be another process lurking within.<\/p>\n<p>And per a <a href=\"https:\/\/www.intrafocus.com\/2014\/07\/xmr-chart\/\">rule of thumb<\/a> for analyzing XmR charts:<em> \u201cthe duration of an XmR chart needs to be revisited when a \u2018long-run\u2019 of data remains above or below the Average line.\u201d<\/em> This trend begins in late summer (which is also<em> <\/em>around the time that Google announced another core update):<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1520\" height=\"536\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9.png\" alt=\"\" class=\"wp-image-177387\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9.png 1520w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9-680x240.png 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9-768x271.png 768w\" sizes=\"(max-width: 1520px) 100vw, 1520px\"\/><\/noscript><img decoding=\"async\" width=\"1520\" height=\"536\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9.png\" alt=\"\" class=\"lazyload wp-image-177387\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9.png 1520w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9-680x240.png 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-9-768x271.png 768w\" data-sizes=\"(max-width: 1520px) 100vw, 1520px\"\/><\/figure>\n<p>We can add another divider at the start of this \u201clong-run\u201d of data to create three separate XmR analyses:<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1041\" height=\"694\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10.png\" alt=\"\" class=\"wp-image-177388\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10.png 1041w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10-638x425.png 638w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10-768x512.png 768w\" sizes=\"(max-width: 1041px) 100vw, 1041px\"\/><\/noscript><img decoding=\"async\" width=\"1041\" height=\"694\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10.png\" alt=\"\" class=\"lazyload wp-image-177388\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10.png 1041w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10-638x425.png 638w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-10-768x512.png 768w\" data-sizes=\"(max-width: 1041px) 100vw, 1041px\"\/><\/figure>\n<p>In doing so, all three analyses seem stable, with no points of extreme variance. In other words, we seem to have done a good job at capturing <em>three distinct processes<\/em> happening within our traffic data.<\/p>\n<p>From this analysis, there seems to be a good chance that our traffic was impacted by external factors around the time of two major Google updates.<\/p>\n<div class=\"post-nav-link clearfix\" id=\"section1\"><a class=\"subhead-anchor\" data-tip=\"tooltip__copielink\" rel=\"#section1\"><svg width=\"19\" height=\"19\" viewbox=\"0 0 14 14\" style=\"\"><g fill=\"none\" fill-rule=\"evenodd\"><path d=\"M0 0h14v14H0z\"\/><path d=\"M7.45 9.887l-1.62 1.621c-.92.92-2.418.92-3.338 0a2.364 2.364 0 0 1 0-3.339l1.62-1.62-1.273-1.272-1.62 1.62a4.161 4.161 0 1 0 5.885 5.884l1.62-1.62L7.45 9.886zM5.527 5.135L7.17 3.492c.92-.92 2.418-.92 3.339 0 .92.92.92 2.418 0 3.339L8.866 8.473l1.272 1.273 1.644-1.643A4.161 4.161 0 1 0 5.897 2.22L4.254 3.863l1.272 1.272zm-.66 3.998a.749.749 0 0 1 0-1.06l2.208-2.206a.749.749 0 1 1 1.06 1.06L5.928 9.133a.75.75 0 0 1-1.061 0z\" style=\"\"\/><\/g><\/svg><\/a><\/p>\n<div class=\"link-text\" data-anchor=\"But is this actually\u2026 useful?\" data-section=\"forecasting-with-xmr\">\n<h2 class=\"wp-block-heading\"><a id=\"post-177378-_77m07rtzmzpi\"\/>But is this actually\u2026 useful?<\/h2>\n<\/div>\n<\/div>\n<p>Now\u2026 this is basically a post-hoc data torturing exercise. We can\u2019t infer any causation from this analysis, and it\u2019s entirely possible that other arbitrary divisions would yield similar results.<\/p>\n<p>But that\u2019s okay. These charts can\u2019t give you definitive, concrete reasons <em>why<\/em> your traffic changed, but they can tell you <em>where to look<\/em>, and help you work out whether troubleshooting a traffic dip or spike is a good use of your\u00a0time.<\/p>\n<p>The ultimate measure of a model\u2019s usefulness is its ability to help you <em>predict <\/em>things. Will XmR charts help me do a better job running the Ahrefs blog in the future?<\/p>\n<p>I think\u00a0yes.<\/p>\n<p>Assuming my \u201cblog process\u201d remains relatively stable\u2014I publish at the same frequency, target the same topics, compete with the same competitors\u2014I now have a set of \u201cstable\u201d data that I can use to provide extra context for future traffic numbers:<\/p>\n<figure class=\"wp-block-image size-full\"><noscript><img decoding=\"async\" width=\"2076\" height=\"698\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1.png\" alt=\"\" class=\"wp-image-177400\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1.png 2076w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-680x229.png 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-768x258.png 768w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-1536x516.png 1536w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-2048x689.png 2048w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-650x220.png 650w\" sizes=\"(max-width: 2076px) 100vw, 2076px\"\/><\/noscript><img decoding=\"async\" width=\"2076\" height=\"698\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1.png\" alt=\"\" class=\"lazyload wp-image-177400\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1.png 2076w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-680x229.png 680w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-768x258.png 768w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-1536x516.png 1536w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-2048x689.png 2048w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/2024-07-22_12-46-47-1-650x220.png 650w\" data-sizes=\"(max-width: 2076px) 100vw, 2076px\"\/><\/figure>\n<p>In the months that follow, I can work out whether dips or spikes in our traffic are likely the result of normal variance, or whether something has changed that requires my attention\u2014like a Google update.<\/p>\n<p>If, for example, my traffic does this next\u00a0month\u2026\u00a0<noscript><img decoding=\"async\" width=\"1030\" height=\"679\" class=\"wp-image-177390\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12.png\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12.png 1030w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12-645x425.png 645w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12-768x506.png 768w\" sizes=\"(max-width: 1030px) 100vw, 1030px\"\/><\/noscript><img decoding=\"async\" width=\"1030\" height=\"679\" class=\"lazyload wp-image-177390\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12.png\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12.png 1030w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12-645x425.png 645w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-12-768x506.png 768w\" data-sizes=\"(max-width: 1030px) 100vw, 1030px\"\/><\/p>\n<p>\u2026I know that\u2014given this distribution\u2014that traffic drop could well be normal, unexciting variance.<\/p>\n<p>But if it does\u00a0this\u2026<\/p>\n<figure class=\"wp-block-image\"><noscript><img decoding=\"async\" width=\"1037\" height=\"701\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13.png\" alt=\"\" class=\"wp-image-177391\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13.png 1037w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13-629x425.png 629w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13-768x519.png 768w\" sizes=\"(max-width: 1037px) 100vw, 1037px\"\/><\/noscript><img decoding=\"async\" width=\"1037\" height=\"701\" src=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13.png\" alt=\"\" class=\"lazyload wp-image-177391\" srcset=\"https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13.png 1037w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13-629x425.png 629w, https:\/\/ahrefs.com\/blog\/wp-content\/uploads\/2024\/07\/word-image-177378-13-768x519.png 768w\" data-sizes=\"(max-width: 1037px) 100vw, 1037px\"\/><\/figure>\n<p>\u2026there\u2019s probably something else at\u00a0work.<\/p>\n<p>With extreme traffic changes you can usually \u201ceyeball\u201d traffic charts and guess what happened. But XmR charts are useful for more subtle variations, and there\u2019s a chance I will be able to identify and act on just a single month\u2019s worth of data. That\u2019s pretty cool.<\/p>\n<h2 class=\"wp-block-heading\"><a id=\"post-177378-_kiyawssgwcb4\"\/>Final thoughts<\/h2>\n<p>Troubleshooting traffic changes is a big challenge for SEOs and content marketers (and we\u2019re working on a few ways to help you identify the signal amongst the noise of your traffic data).<\/p>\n<p>In the meantime, I have found XmR charts an interesting tool in my toolkit, useful for contextualizing my monthly reporting numbers and justifying when I should (or shouldn\u2019t) spend my energy troubleshooting a down\u00a0month.<\/p>\n<p>(At the very least, XmR charts might just give you the confidence necessary to say \u201cget off my back\u201d when that VP sends you a brusque 3AM email complaining about last month\u2019s 8% traffic dip.)<\/p>\n<div class=\"sidenote\">\n<p>Sidenote.<\/p>\n<p> Thanks to <a href=\"https:\/\/www.linkedin.com\/in\/benyamin-elias-5b1b0396\/\">Benyamin Elias<\/a>, VP of Marketing at Podia, for introducing me to XmR charts.<\/div>\n<\/p><\/div>\n<p><br \/>\n<br \/><a href=\"https:\/\/ahrefs.com\/blog\/can-you-spot-google-updates-with-xmr-charts\/\">Source link <\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Website traffic data normally looks like\u00a0this: Ups and downs, peaks and troughs. If we\u2019re doing our job properly, we generally expect traffic to trend upwards over time, but in any given month, it\u2019s difficult to say whether a peak or a trough is worth paying attention to. Did we do something great and trigger a [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":2622,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[22],"tags":[],"class_list":["post-2621","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-seo"],"_links":{"self":[{"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/posts\/2621","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/comments?post=2621"}],"version-history":[{"count":0,"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/posts\/2621\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/media\/2622"}],"wp:attachment":[{"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/media?parent=2621"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/categories?post=2621"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/ewebtoolz.com\/blog\/wp-json\/wp\/v2\/tags?post=2621"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}