<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Eric Matisoff on Medium]]></title>
        <description><![CDATA[Stories by Eric Matisoff on Medium]]></description>
        <link>https://medium.com/@ericmatisoff?source=rss-7511e96fa2------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*mwLLvYHC3LIJZXT22gul0w.jpeg</url>
            <title>Stories by Eric Matisoff on Medium</title>
            <link>https://medium.com/@ericmatisoff?source=rss-7511e96fa2------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 20:14:12 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ericmatisoff/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Assessing ITP’s Effect On Adobe Analytics Data]]></title>
            <link>https://medium.com/@ericmatisoff/assessing-itps-effect-on-adobe-analytics-data-469788e9733d?source=rss-7511e96fa2------2</link>
            <guid isPermaLink="false">https://medium.com/p/469788e9733d</guid>
            <category><![CDATA[adobe]]></category>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[safari]]></category>
            <category><![CDATA[itp]]></category>
            <dc:creator><![CDATA[Eric Matisoff]]></dc:creator>
            <pubDate>Mon, 22 Nov 2021 19:34:28 GMT</pubDate>
            <atom:updated>2021-11-22T19:34:28.349Z</atom:updated>
            <content:encoded><![CDATA[<p>Apple’s <a href="https://webkit.org/blog/9521/intelligent-tracking-prevention-2-3/">Intelligent Tracking Prevention</a> (ITP) has been around for several years now. The goal, to intelligently prevent tracking, has sent ripple effects through the adtech and martech industries, affecting nearly every company with a digital experience whether they know it or not.</p><p>There has, until recently, been a lack of clarity on how to properly assess the effects of these browser-driven cookie changes driven by one company’s need to prioritize privacy for their customers.</p><p>3rd party cookies were killed by Safari, then on all browsers on iOS. Firefox and Edge and Chrome (eventually) are going 💀 to 3rd party cookies.</p><p>1st party cookies set by JavaScript were then affected with shorter expirations, maxing out at 7 days in Safari and iOS.</p><p>CNAME and Adtech-driven cookies were yet another nail in the coffin, with 1st party cookies with expirations as short as 24 hours.</p><blockquote>You’re probably wondering: How have all of these changes affected my business’s data?</blockquote><p>Fear not, Adobe has your back. When ITP came for JavaScript cookies in February of 2019, we released <a href="http://adobe.ly/itp">adobe.ly/itp</a> to help customers begin to migrate to a CNAME, server-side set first party cookie. When ITP then limited the length (and usefulness of cookies set via CNAME) at the end of 2020, Adobe began developing new technologies and strategies to mitigate the effects.</p><p>I’m excited to share that you now have a way to <strong>assess</strong> the impact of these ITP-driven changes. Every Adobe Analytics customer now has a new Workspace Template in their arsenal. Let me show you how to access it and touch upon a few of the highlights:</p><p>1 Open the <em>ITP Impact</em> template in Analysis Workspace. If you haven’t turned on the New Landing Page for Analysis Workspace, click Project &gt; New, then select the ITP Impact template.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c_Y6bSavSUhZYc6amRhYkA.png" /><figcaption>Ah, there it is.</figcaption></figure><p>If you’ve turned on the New Landing Page for Analysis Workspace:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/448/1*jeWF3BVEY7eLIvU2ure8xA.png" /><figcaption>Turn this on, you’ll thank me later</figcaption></figure><p>Then click Reports from the left rail, and search for ITP Impact (it’s listed under Engagement):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*N8Kx8ZdN5YoKMz-7YfwX7Q.png" /><figcaption>Did you know Workspace templates are listed under the Reports area now?</figcaption></figure><p>2 Once you’ve got the ITP Impact template open, you’ll see a number of Panels with data accessible to be applied to any of your report suites. <br>Hint: make sure you have selected a production report suite :)</p><p>3 The <em>Introduction</em> Panel will provide an even more thorough history of ITP than what I’ve written above.</p><p>4 The <em>Understand: How many users are using an ITP-impacted browser?</em> Panel will help you understand how many of your visitors are using browsers that have been affected by ITP, to the best of Adobe’s abilities. It’s a pretty thorough segment that is worth combing through to see how we’ve focused on iOS and Safari browsers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NkpRk20KMMj1tJpJextK8Q.png" /><figcaption>18% of Visitors are using an ITP-affected browser</figcaption></figure><p>5 The <em>Understand: How many visitors return after 7 days?</em> Panel shows you details about non-ITP-affected visitors (ie — Chrome browsers). This provides a feel for how often your visitors return to your site without the limits imposed by ITP.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wQ2uuWhEWCaYFoBSwypngA.png" /><figcaption>Hopefully your site is stickier than mine</figcaption></figure><p>6 The <em>Understand: Impact of ITP on Conversion Rate</em> Panel provides a framework for you to quickly assess how ITP has affected your conversion rate. Note that by default, we show Page Views / Unique Visitor, but you can simply drag and drop different Conversion Rate metrics to the table to see how the ITP-driven increase in Unique Visitors (the denominator) causes a decrease in Conversion Rate.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zkI1mlRFnANRFnWCbp_XAg.png" /><figcaption>The increase in UV’s has caused a decrease in Conversion Rate.</figcaption></figure><p>7 The<em> Understand: Estimate Unique Visitor Inflation on ITP browsers</em> Panel helps you to assess the amount of Visitor inflation due to the decreased expiration of first-party cookies in browsers affected by ITP. If you haven’t already performed this analysis, it may be interesting to see how this data compares to the same metrics prior to ITP 2.1 taking effect (February 2019). Keep in mind that the availability of this data depends on your company’s <a href="https://experienceleague.adobe.com/docs/analytics/technotes/data-retention.html?lang=en">data retention policy</a>. Most Adobe Analytics customers utilize the default 25 month retention policy, which unfortunately means that you may not have data from early 2019/late 2018 available in Analysis Workspace. If you’ve been backing this data up in internal systems and data lakes, now may be a time to use it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*W7iCLfDV909Y9OtQ0yXEfA.png" /><figcaption>The increase in UV’s is evident in the significantly decreased Visits / Visitors metric.</figcaption></figure><p>Consider adding Visits to this table and trending ITP vs Non-ITP Visits to feel confident that there haven’t been any significant, recent changes to one browser type to another.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FIwomBUfsA7YbrL_Cf0GKQ.png" /><figcaption>The trending of ITP and Non-ITP browsers is similar and consistent, helping ensure the Visits / Visitors metric continues to be reliable.</figcaption></figure><p>8 The<em> Understand: Days since last Visit</em> Panel provides you with a more granular view into how time data is affected by ITP. The <a href="https://experienceleague.adobe.com/docs/analytics/components/dimensions/days-since-last-visit.html?lang=en#dimensions">Days Since Last Visit</a> dimension identifies the time passed between the most recent and immediately previous visit. I find the default freeform table for this visualization to be helpful, but I love to expand it by adding Non-ITP Visitor and All Visits segments. You could even consider adding conversions as a metric here to again see how ITP is affecting conversion data from week to week.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2WWp4S0me4fKYYVfuM_o5w.png" /><figcaption>This customized Freeform Table shows how Visit and Conversion Retention appears to lean heavily on shorter days between visits, when in reality ITP is causing cookies to reset, decreasing the average days since last visit for ITP browsers.</figcaption></figure><p>9 The<em> Adjust: View only data unaffected by ITP</em> Panel provides you with a playground to analyze data in your report suite that has not been affected by ITP. We’ve seen some customers take this to the next level by creating a <a href="https://experienceleague.adobe.com/docs/analytics/components/virtual-report-suites/vrs-about.html?lang=en">Virtual Report Suite</a> that focuses on this data.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*z1CRZn3Cuyhu1qmdowAd2A.png" /><figcaption>Non-ITP data can be analyzed without any concern for unusually short 1st party cookie expirations.</figcaption></figure><p>10 The <em>Adjust: Use calculated metrics to adjust for the skew caused by ITP tracking limitations</em> Panel provides you with a framework for building calculated metrics that incorporate adjustments based on the previous analyses. Inspect the way in which the <em>Estimated Unique Visitors (ITP 2.1)</em> and <em>Page Views / Estimated Unique Visitors (ITP 2.1)</em> calculated metrics were created. Instead of using the default denominator of 2, utilize the data within this template to find a static number that aligns with your organization’s unique ITP effect.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cvNM5daQrjgCLJ_dX_VniA.png" /><figcaption>Note how my Page Views / Estimated UV of 5.6 is closely aligned with the Non-ITP Page Views / Non-ITP UV metric of 5.5 in the previous section.</figcaption></figure><p>11 <em>The Reference: ITP Segments</em> Panel is to be used exactly as it sounds, as a reference panel. Feel free to use the pre-built Freeform Table to compare and contract the different types of ITP-related Segments that have been built by Adobe.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R98KLVOX70lntVCgwvembA.png" /><figcaption>The four key ITP-related segments are provided for you.</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=469788e9733d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Rockstar Classifications use cases]]></title>
            <link>https://medium.com/@ericmatisoff/rockstar-classifications-use-cases-bc64212d7a31?source=rss-7511e96fa2------2</link>
            <guid isPermaLink="false">https://medium.com/p/bc64212d7a31</guid>
            <category><![CDATA[classification]]></category>
            <category><![CDATA[analytics-rockstar]]></category>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[adobe-analytics]]></category>
            <dc:creator><![CDATA[Eric Matisoff]]></dc:creator>
            <pubDate>Wed, 18 Aug 2021 19:05:01 GMT</pubDate>
            <atom:updated>2021-08-18T19:05:01.172Z</atom:updated>
            <content:encoded><![CDATA[<p>Several years ago, we had an Analytics Rockstar deliver a unique and powerful tip that helped her win the competition. The tip was presented at 2019’s SUMMIT in Las Vegas by <a href="https://www.linkedin.com/in/audrey-salerno/">Audrey Salerno</a>, of <a href="https://cars.com/">cars.com</a> fame. You can actually watch the entire session via this link which is chock full of previously shared <a href="https://adobe.ly/aarockstartips">Rockstar tips</a>.</p><p>The tip is ingenious (<a href="https://bit.ly/AudreyTips2019">Audrey’s Analytics Rockstar slides</a>) — prepend the current date to your Page Name value and throw it into an evar and/or prop.</p><h3>For example:</h3><p>a Page Name of:<strong><br>rockstar: discography: live album</strong> <br>is saved in an evar/prop as:<br><strong>17aug2021|rockstar: discography: live album</strong></p><p>Audrey then shared how easy it’d be to classify the concatenated variable for a great use case: annotations! If there was a tagging issue, or a new campaign that went live on the 17th, she’d classify that date’s dimensional values with this additional information. It’s so, SO slick!</p><p>So why am I telling you this? Because I think there are plenty more use cases for applying Classifications to a variable with a date prefix.</p><h3>Date-Enabled Classifications</h3><p>For those of you that have been around the Adobe Analytics world for a while, you may remember a feature called <a href="https://experienceleague.adobe.com/docs/analytics/components/classifications/c-classifications.html?lang=en">Date-Enabled Text Classifications</a>. The concept was great, apply a date range to your Classification values so they would only apply to a certain snapshot in time.</p><p><a href="https://www.youtube.com/watch?v=6laI5SBXY-I">Classifications without Dates</a> are incredibly valuable, but they are always set to overwrite and display the most recently uploaded value. So if you were to classify <em>product-abc</em> as a <em>shirt </em>today, then tomorrow you classify it as <em>clothing</em>, then Adobe Analytics will show the classification as <em>clothing</em> for your entire report suite’s history for <em>product-abc</em>.</p><p>As you can imagine, Classifications without Dates are useful for probably 90% of Classification use cases. However, Date Enabled Classifications, even though they no longer exist, still have several valid use cases — classifying values that change over time:</p><ol><li>Your organization likes to reorganize the hierarchy of your product SKUs and would like the data to match with the assigned hierarchy during that time.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ArkAgy16cKSYnxSrBMeDKQ.png" /><figcaption>A Date-Enabled Product SKU is classified to show different hierarchy values for the same value depending on the date</figcaption></figure><p>2. Tracking code values are re-used across different campaigns at different times, so the tracking codes need to map to different Classification values depending on the time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nqImjwUeW8eTTzJU35f8MQ.png" /><figcaption>A Date-Enabled Tracking Code is classified to show different classifications for the same value depending on the date</figcaption></figure><p>3. As a publisher you are constantly moving articles from one section of the site to another and you need the Classification of these articles to always properly align with the section that it was placed in on that day.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wRTy_G6TmOMcRIyCpl_lew.png" /><figcaption>A Date-Enabled Article ID is classified to show different article site sections fort he same value depending on the date</figcaption></figure><h3>So What Can We Do?</h3><p>Finally, we get to the point — you have the option to take Audrey’s trick above (classifying a dimensional value with the current date prepended) but apply it to Date Enabled Classification use cases!</p><p>Example 1 <em>an organization that likes to reorganize their product SKU hierarchies</em>: imagine if they had an evar that captured: <strong>date|SKU</strong>. The data could then be classified based on date! Voila — Date Enabled Classifications are back!</p><p>Example 2 <em>an organization reuses tracking codes that need to be classified differently depending on the time</em>: imagine if they had an evar that captured: <strong>date|trackingCode</strong>. The data could then be classified based on date! Bingo bango — Date Enabled Classifications are back!</p><p>Example 3 <em>a publisher that likes to move their content to different site sections and classify them based on their location at that time</em>: imagine if they had an evar that captured: <strong>date|articleID</strong>. The data could then be classified based on date! Easy peasy lemon squeezy — Date Enabled Classifications are back!</p><h3>OK, but what are the limitations?</h3><p>Yes, you guessed it, there are still some limitations. Let’s go through them one by one:</p><ol><li><strong>Plan ahead.</strong> You’ve got to create and set the evar in your Adobe Analytics tag before you Classify it. This means you’ll need to think ahead as you’re planning your implementation enhancements — think about the types of dimensions that Date Enabled Classifications could be applied to.</li><li><strong>Consider </strong><a href="https://experienceleague.adobe.com/docs/analytics/technotes/low-traffic.html?lang=en"><strong>cardinality</strong></a><strong> limitations.</strong> You’ll basically be multiplying the number of unique values in your new dimension by up to 31 every month. Cardinality limitations kick in around 500,000 unique values in a month, and if you divide 500,000 by 31, then you’ll get 16,129 unique combinations of values with date prefixes. If in a month you sell more than 16,129 products, or utilize more than 16,129 unique tracking codes, or publish more than 16,129 unique articles, etc, then you may run into our old friend <em>Low Traffic</em>. Keep this in mind as you’re planning which variables to apply this trick to. Perhaps you can deal with just have weekly classifications, rather than daily? You could then set your base dimension as <strong>week1aug2021|abc123</strong>. Then you’d max at just 5 weeks per month, and your unique values goes from 16,129 to 100,000 per month.</li><li><strong>Product SKU is tricky.</strong> Yeah, we all know this. The s.products variable is a beast, and you don’t want to pass date enabled SKUs into the <em>Product</em> variable in Workspace. So you would want to consider using a merchandising evar that you then populate using product syntax every time you set s.products in your tags. (If that didn’t make any sense, check out the documentation on <a href="https://experienceleague.adobe.com/docs/analytics/components/dimensions/evar-merchandising.html?lang=en">merchandising evars</a> to learn more).</li></ol><blockquote>Go forth and classify, you Rockstar you!</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bc64212d7a31" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>