Skill Exchange Event Aug 2023 - Grow Track - Putting it all together

Transcript

Hey, everyone. My name is Eddie Ahn. I’m a field engineering technical consultant here at Adobe, and I am very excited to be here to present on the last part of our session, putting it all together. Now in the first two parts of our session, we learn how to first get started in Analysis Workspace and how to analyze the data. Now in the first two parts of our session, we learn how to get started in Adobe Analytics, as well as how to analyze the data. Now in this third and final session, we will put it all together and dive more deeply into how we can gain better insight from our customer data through the use of visualizations as well as anomaly detection. And from there, we’ll also learn how to curate and share those projects as well with others. So let’s get right into it. So the first topic we’ll be getting into is how to build visualizations. Now within Workspace, there are a number of visualizations that lets you generate visual representations of your data. And you can access these visualizations from the top left icon in Workspace as seen here to the screenshot to the right, or through a blank panel, or even through the right-click menu in your workflow. Now in this next slide here, these are also two other ways where you can quickly be able to visualize your data. Now here on the first part to the screenshot to the right here, if you right-click on a free form table, for example, there will be a dropdown that will appear and you can click visualize to be able to quickly create visualizations based on that data. And even to the right here, if you go to a free form table and you hover over the table row, a visualize icon will appear. And if you click it, it will prompt Analysis Workspace to take an educated guess at which visualization would best fit your data as well. So now that we’ve covered how we can create visualizations, let’s get into what kind of visualizations Workspace has to offer. So these next two slides will be going over all the different visualizations that there are. And as you quickly skim through all of these, you’ll notice that some of them you might be already familiar with, such as a bar, a donut, or even an area. However, you might also notice that there might be some that you might not be as familiar with, such as a free form table, a cohort table, a fallout, or flow, which are the first floor listed here. If I move on to the next slide here, notice the same thing. These are more additional visualizations that Workspace has to offer. And again, you might be already familiar with some, but might not be as familiar with others. So what we’re going to do in these next following slides, we’ll be picking some of these visualizations that are not as straightforward and are very useful in Workspace so that you have a better idea on the advantages that Workspace has to offer. So with that, let’s get into our first visualization, which is the flow chart. And with the flow chart, you’re able to visualize the customer journey through your website or application. And you’re able to analyze where customers go before and after specified checkpoints, such as an entry page, a specific dimension, or even exit. Now if we look at the screenshot to the right here, here’s an example of a entry page. And we can see these are all the pages at which people first entered our website. And if we look here in the middle under page, we can see that home is listed up on top, which means that the home page is our most popular page at which people first are entering our website. Now, another new feature is you’re able to mix and match different dimensions and events. Now if we look at the third column here, again to the right, we’re able to see that from the page, we’re looking at internal search term here. So not only can we create a flow chart just looking at simply pages, we’re able to mix and match pages as well as internal search term or any other dimensions to see how they are related and what flow people are taking. So in this case, from looking at our home page, we’re seeing all the other internal search terms people are using on our website that have visited the home page from there. And if we want, we can expand this flow further in order to have further analysis on our path that people are taking in our website. Now with the flow chart, you’re also able to create segments by designating a specific point in a chosen path. In other words, if you have identified a specific path that you would like to further analyze, you can right click at any of these paths and a dropdown will appear where you’re able to create a segment. And by creating a segment, you can use that segment for further analysis and other portions of your report. Now the second visualization is the fallout table. And similar to the name fallout, we’ll go over the definition of exactly what a fallout table is. It shows you where visitors left or fell out and continue through or fell through a predefined sequence of pages. And you’re able to drag and drop and rearrange different funnel steps, otherwise known as touch points. If we look at the screenshot to the right here, we’ve dragged and dropped the home page, the gear page, and some shopping cart pages as well. So this is the predefined sequence of pages that we want to look at. And you can see here that it shows us the percentage of people and the percentage of people as well that have not followed this sequence of pages. In addition, similar to the flow chart, here we are also able to mix and match values from different dimensions and metrics. If we look at the screenshot to the right on the bottom here, we can see that not only are we looking at pages, we’re also looking at action names. In this case, we’re seeing after a person came to the item info page, what action names did they take. So we want to see whether they added a certain item or not. And from here, we’re able to mix and match different dimensions for maybe a multidimensional analysis and report using this fallout table. Now an interesting feature with the fallout table is you’re even able to identify where customers go immediately after falling out. If we look at the screenshot to the right on the top again, I pointed out before that you’re able to see the percentages at which people are following through the sequence or falling out of the sequence. So if we hone in specifically where they are going from the homepage to the gear page, we can see that 7.2% of people, so only 7.2% people are going from the homepage to the gear page, while 71% of people have left and are not going to the gear page after going to the homepage. So from here, if we right click this touch point here, dropdown will appear and it will give us an option to be able to identify where customers are going immediately after the homepage instead of the gear page. So give us a list of all the pages at which people are going to instead of the gear page. So a neat feature with the fallout table. Now the third and final visualization that we will be going over in more detail is the cohort table. And to get a better understanding of what a cohort table is, let’s first go over the definition of what a cohort is, which is basically a group of people sharing common characteristics over a specified period. And with the core table, you’re able to identify trends or patterns of how a cohort is engaging. Now there are two different types of cohort. The first one is a retention table and the second one is a churn cohort. So for the retention, the retention shows the percentage of visitors who completed an action over time. If we look at the screenshot to the right on the top here, here’s an example of a retention cohort table. And we can see under the title retention, there are two parameters, an inclusion criteria and a return criteria. So for the inclusion criteria is looking at visits, at least one visit. And then the return criteria is looking at at least one online order. So our cohort here in this case is the inclusion criteria. So we’re looking at only people with at least one visit. And then out of all those people with at least one visit, we’re trying to see how many people have placed at least one online order. And what’s cool about the core table is that it breaks it down by day by day basis. So we can see here, let’s choose January 1st, for example. We can see that under the included column, 7,077 people have at least visited the website at least once. Now it separates this by day. So we can see after one day, we can see that 88 people have placed an online order. After two days, only five people have placed an online order. After three days, only one person has placed an online order. So from here, you’re able to really hone in on how long it takes for maybe a certain conversion. In this case, it’s an online order, which could be also useful for maybe some certain campaigns. So if you find a pattern and you observe that after two days, there’s a significant drop, you can maybe send out a campaign to remind the person to make that online purchase. Now the second core type is the churn table, which is just an inverse of retention and shows the visitors who fell out or never met the return criteria over time. So if you look at the screenshot to the right on the bottom here, we’re still looking at all the people with inclusion criteria, which is at least one visit. However, we’re now looking at people who have not placed at least one online order. So just the same thing again as a retention, except we’re looking at the opposite. That concludes the end of our visualization. So we will be going into our next topic, which is detecting anomalies and identify correlations, taking advantage of the anomaly detection feature in Workspace. So what is anomaly detection? So anomaly detection provides statistical methods to determine how a given metric has changed in relation to previous data and allows for a separation of true signals from noise and identifies any contributing factors to those anomalies. And it makes sure to consider any seasonality, such as Black Friday and any holidays, just in case sometimes there is a significant increase in a certain metric because of a holiday and might not be an anomaly. It might just be there’s might be a significant increase of a certain date because it was Black Friday, but it makes sure to consider all of that in the analysis when is running the anomaly detection. And you can view the anomalies in a free form table or in a line chart. And I’ve included what the free form table and a line chart here looks like in Workspace to the right. So let’s get into those specific free form table and line chart and how to identify anomalies in each. So first, the line graph. Now if we look at the screenshot to the right here, we can see that there are many different components that help us identify anomalies and we’ll be going over them one by one. So the first component is a solid line, which indicates the actual performance or the actual data. Next is the circles on the actual line. And these are where the anomalies were detected. And then the dotted line is the expected line. And then the light blue region above and below the dotted line is the upper and lower expected range. Now if we hover over and click on any of these circles, which again indicate that there is an anomaly detected then, it will show us the metrics. So in this case, we’re looking at visits. So on Friday, August 14th, 2020, there were 32,057 visits. However, this was an anomaly and anomaly was detected and we can see exactly the percentage above expected. So in this case, it was 468% above expected. Now below you can see this blue button here that also tells us to analyze. And if we click analyze, this will bring up something called contribution analysis, which we will also be going over within the next couple of slides. Now the second way to detect anomaly detection is using a freeform table. And whenever you drag in a freeform table, by default, it will automatically look for anomalies in the table using several built-in models that analytics has to offer. And wherever an anomaly is detected, a black triangle, as you can see here on the right, will appear on the right corner of the row. And now if we mouse over or click on any of these black triangles that appears, we can see similar to the line chart before, we can see the exact actual value that is above or below the expected value. So in this case, we can see on August 14th, 2023, it was again 468% above expected. And now in the freeform table, if we look here in the middle, a shaded area will display, which represents the value. And the vertical line will also show you the expected value. So these are all the different components of a freeform table where you can kind of see how anomalies are detected. Now once you’ve detected anomaly, going back to the line chart and the freeform table, if you do run a contribution analysis on any of these anomalies, it will bring up something called contribution analysis. And what this does is it helps you evaluate your data immediately to answer why a certain anomaly happened. And it will run through many advanced algorithms and machine learning processes to evaluate associations that contributed to that significant spike or even dip. Now when you first run this contribution analysis, you also have the options to include some dimensions from the contribution analysis. So if you exclude a certain dimension, it won’t consider that when it’s considering all the contributing factors to that anomaly. Now after you press the blue button, run contribution analysis, and it will return a report for you. So all these calculations are then displayed in different visualizations designed to give you varying perspectives to help answer why an anomaly happened. So here on the screenshot to the right here on the bottom, I’ve included one example of those visualizations. So in this case, this is our freeform table. And this is still going off that same anomaly that was detected in the previous last two slides. So we’re again looking at visits on August 14, 2020. And there was an anomaly detected on this day with the number of visits. So once we run a contribution analysis, we can see here in this freeform table, it returns a list of all the different dimensions that this contribution analysis report thinks is contributing to that anomaly. So in this case, the top three are regions, Delhi, India, countries, India, persisting cookie support enabled. And then here in the middle, we can also see that there is a contribution score attributed to each of these dimensions. So what this means is we can see there’s one, 0.95, 0.65. So we can see that if a number is closer to one, that means there is a very direct correlation between that dimension and that anomaly. And as it goes closer to zero, it means there is no correlation at all. So one means very good and zero means very bad. And a note here that I’ve included here on the bottom here is that the number of runs per company is limited by monthly tokens that are determined by that Adobe Analytics product your company purchased. So you are unable to run contribution analysis unlimited times. There are certain numbers and limits to how many times you can run this contribution analysis. Now with that, let’s get into our final topic today, which is curating and sharing projects. So once you have created all your reports and all your findings within analysis workspace, maybe you would want to share those findings with other members within your team. So sharing makes a project available to other analysis workspace users in your team, and you can share a project with all the available default components or even with selected components. And you can do this through something called curation, which lets you limit the components before sharing the project. So if we look here on the bottom here on the first screenshot, if we click share, there are many different options to share. And we’re even here to the right. Again, this is going back to curation. We can curate different components to limit what certain dimensions, metrics, segments, and date ranges they can see when we share the project with them. So here are two screenshots of two different ways, main ways of how to share our project. So the first way here to the left here is just sharing a project, and we’re able to specify the specific permission at which we want to share the project. If we want to share the original project with them, we can put it under edit original. If we want to share a copy of them, we can share it under edit copy. And if we only want them to have read only access, we can do under read only. And you can even share a link with them. And here down below in fine print, you can see that shares projects as edit copy by default if you share by the link. Now moving over to the right here, we can also send the file with a CSV format or PDF format, and you can list the recipients below. And a cool feature is you can even schedule that. So you don’t have to constantly send the file to them. You can just let it be. And based on the frequency and the starting on and ending on date range at which you select, it will send it accordingly to them for all the recipients you have selected. Now for recipients who won’t necessarily log in to Analysis Workspace, you can even download a CSV or PDF. So you can download data to a CSV file and send them via email, and you can download up to 50,000 rows. And you can also download the dashboard itself to a PDF if you would like to keep all of the visuals, and you can also send this via an email. Now this next final bullet is to understand and a new feature on how to share a read only link to share with anyone, not just with Workspace users. So this is a new feature within Analysis Workspace and is right now only available in Limited Testing. But here on the right here, you can see there is an option with Share with Anyone. And if we share with anyone with this link, there is no login required in order to access our workspace. So that brings us to the end of the topics that we’ll be covering today. But I did want to point out that we do have an Experience League website, which documents all of the Adobe Analytics findings and different tools and tutorials that you can use to further enable yourself in Adobe Analytics and Workspace. And if you’re more of a video learner, visual learner, there is also a YouTube channel for Adobe Analytics where you’re able to view videos and learn more as well. So please feel free to check out either our Experience League documentation, YouTube channel, or there are many, many more documentations that you can use online to further enable yourself in Adobe Analytics. That ends my presentation. Thank you. Eddie, it’s wonderful to have you here live with us. Thank you for being here. Thanks, Brad. Thanks for having me. Excited to be here. Awesome, awesome, awesome. Got some questions from the audience. So let’s get right into it. The first one we have here is from Lilia. Can you see flowchart from channel visitors came into a specific page, for example, like came from YouTube or Instagram? Certainly. So that is a powerful feature with the flowchart. So you’re not only limited to one single dimension. So you have the ability to mix and match different dimensions to your flowcharts. So in the case of seeing which certain marketing channels it came from, you can see and set a dimension and pull in a dimension to the flowchart with a channel dimension. And then from there, you can also pull in a page dimension and combine them together to see which channels led to different pages. Awesome. That makes sense. Next question. Is there a different visualization for the fall table, perhaps like a vertical bar style? Yeah, great question. So unfortunately, that’s the way the fallout chart looks like. It is all stacked on top of each other so you can drag and drop different touch points. So there really isn’t a way to really change the way it looks. However, there are different visualizations. One similar one is a flowchart where it is going vertically across or around with the fallout table. Since you have the ability to pick certain touch points instead of just looking at every single path, that’s when a fallout visualization will be more useful. Got it. Makes sense. Next question. It’s about customer journey analytics, the newest and hot topic in the Adobe analytics world. Is that a separate product or is it combined with Adobe analytics? How does that work? Yeah, great question. So that is actually a new tool that Adobe provides with AEP. So there’s something called Adobe Experience Platform. So CJA or customer journey analytics is built on top of AEP. So CJA is a separate product from Adobe analytics. However, it is built, it is basically analytics, but it is still using analysis workspace, but on top of AEP data. Got it. That makes sense. I know that CJA has been the hot topic coming around for a long time lately. So it’s good to hear that people are starting to have some questions about it and we’d love to see more people utilizing that tool. So any questions or more feature requests or anything like that on CJA, keep them coming for all of us at Adobe for sure. Next question is from Steven. Can you set up alerts for anomalies? Certainly. So there is a feature called intelligent alerts in Adobe analytics. So you have the ability to create any metrics inside of those intelligent alerts and be able to track and detect any anomaly. So if an anomaly is occurring, then you can receive an alert based on those anomalies that you have selected within the intelligent alerts portion of Adobe analytics. Got it. Got it. Next question is from Brayden. I might need more clarification on this one, but I’m going to ask it for you. Does the UI change at all for users that don’t have a login? So like, let’s say you just go to your version of Adobe analytics and you haven’t logged in. Is it a different view or how does the UI change? Yeah. So unfortunately you do need to have a login in order to access Adobe analytics. So without a login, if you’re not an analytics user, you won’t be able to access any of the projects that you are able to report on. Gotcha. Hopefully that answers your question Brayden, but if you have a follow-up, please type it into the chat for us. Next question. Can you share a little bit more about the upcoming share with anyone feature? Will it provide the viewer with a dashboard or view of visualizations and or panels that have been created or will it reflect current day views of the visualizations? Yeah, great question. So that is a new feature that has been rolling out. So personally I haven’t really dived too deeply into the new feature, but from my understanding it is just like how you can share any workspace project with any existing user. But now with the public link now, you can now access those dashboards with that public link that you can use. Got it. So bringing that back to us, his question was actually similar to what we were just talking about with that share with anyone feature. When we share a project with a non-user, does that sound the same thing? Can they, can I get into it or do they need to have a login? Yeah, so you can view, from my understanding, you can view the project. However, you’re not able to create any projects or edit anything within the project since you’re not an analytics user. Got it. Excellent. All right, Indipreet’s got a question here. Where can we check all the EVARs, props and details and experience space? So in order to check all those EVARs and props within your report manager under report suite settings, you have the ability to look at every single EVAR and prop that you have set up that’s specific to your report suite. So if you have multiple report suites, you can select that specific report suite within your report suite settings. And then from there, you can look at all the EVARs and the settings that you have set up for all of that as well. All right. Next question is from Will. If you apply a segment to the inclusion criteria of a cohort table, is it necessary to apply that same segment to the return criteria or is it automatically segmenting it to the user group? Yeah, so the inclusion criteria and the return criteria are two different criterias. So if you apply the segment to the inclusion criteria, it will only be applied to that inclusion criteria and then the same thing vice versa with the return criteria. So if you would like to apply the same segment to both criterias, the recommendation is to drag and drop both segments to both of those criterias. Got it. Okay. Next question. What is the best tool or feature in Adobe to show both online and offline channel mix or data in one visualization or workspace? Yeah, great question. So that kind of goes back to how CJA works. So the new tool with Adobe Analytics. So it’s using the AP data. So how AP works is you have the ability to send offline and online data to AP and from there using customer journey analytics, you can analyze all of that data using Analysis Workspace but just using that AP data. So unfortunately there’s no real way for currently within just Adobe Analytics to look at that offline data but with the new features coming out like we were talking about before, there are certainly ways in the future to do so. Awesome. CJA, come back again to save the day. That’s how it happens. I like it. I like it. Is it possible to compare visualizations for a different date range? We have events that are in like August last year and October this year and you want to compare. Is that possible? Yeah, certainly. So there are many different ways how you can apply different date ranges so that you can compare your data. One couple immediate solutions are you can apply a date range using the components. So in the left hand side of Workspace, you have the ability to select any segment, your dimensions, your metrics, but also there’s ability, a section where you can see the different date ranges that Workspace has to provide. So you can drag and drop that into any of your panels and apply that either as a segment, apply them directly into a free form table or even in the top right corner of any panel that you’ve created, there’s also the ability to change the date range at which you want to see the data. So many, many different ways of being able to be creative and how to look at different data from different date ranges. Got it. I got it. Another question here from Inderpreet. Is it always necessary to have a tag manager to map EVAR for props or can we go directly into a Workspace? Great question. So there are many different ways to actually implement Adobe Analytics and to map those EVARs and props. So using a tag manager such as Adobe Tags is definitely a very easy way in order to do so. However, there are different ways such as you can even do something called processing rules. So within Adobe Analytics, what processing rules are is as the data comes in, for example, if you’re sending in any context data, you can send that data and map those data into certain EVARs and props that you would like to in Adobe Analytics. Got it. Got it. Got it. Okay. So we’re talking about Adobe’s offering of Adobe Analytics and its full suite of analytics packages. So is there a light version of Adobe Analytics for AEM assets or ACS, something like that? People are having some feedback that Adobe Analytics might be too much for them. So they’re just wondering what their other options are. Great question. So from my knowledge, there is no light version of Adobe Analytics. It is just Adobe Analytics, but maybe there is something out there that I am not aware of. Experience league, probably a great question to ask that one in all honesty. All right. Next question. What is the quickest and easiest way to show year to date percentage progress towards an annual goal or target? If your 12-month goal is X, how do you visualize the percentage achievement year to date? That is one of the visualizations that Analysis Workspace has to offer called a bullet graph. So with a bullet graph, you can specify the mid goal, a low goal, as well as an end goal. And with that, as you set those, it will create a visualization for you. So you’ll be able to easily identify where you are currently at and how far, let’s say, you have surpassed the low goal and how close you are to that end goal that you have all through that bullet graph. Awesome. Great name. I love the bullet graph. I never heard of that one before, but I like that. Did you come up with that yourself or who is that one? You would have to ask the product team. Is there a way to create a visualization that is looking at only specific set of people? In addition, is there a way to change the container to look at a visit or visitor level? Yeah, so to answer that first part of the question, that is where segments come into place. That is a very powerful tool within Analysis Workspace to apply segments anywhere you want. So if you set and create a segment of that specific set of visitors that you would like, you can apply that segment to any visualization or any panel. And from there, all that data will be filtered so that you’re only looking at those specific set of people that have satisfied on that segment. And then for that second portion of the question, for many of those visualizations, you do have the ability to change the container level from a visitor. So if you would like to look at a whole visitor sequence, not just one visit, then you can select the visitor container. And then if you would like to change it to look at only a single session or look at only a single visit instead of what a whole visitor did, you also have the ability to do so by selecting just to look at the visit as well. Got it. Okay. What is the best place for a customer or a prospective customer to come in and try out Adobe Analytics? Is there like a sandbox, like an open source place where they can practice? Great question. So that definitely you can maybe hopefully find that out from someone else, maybe experience it or a different source. But personally for me, I won’t have the ability or are aware of any way to give someone a test trial of Adobe Analytics, but I’m sure if you talk and you are interested, I’m sure there is a way for you to try out the solution before you actually purchase. Definitely a good question for like your customer success manager or account manager or something like that. Cause I definitely know there’s some ways to do it for customers for sure. Awesome. Awesome. Awesome. Awesome. Next question. What happens if two editors save a project at the same time? Good question. So with Workspace, unfortunately there is no live collaboration. So with that, if you two people are working on the project at the same time, whoever saves last that all those updates made with that person who has saved last will be what you see when you go onto the project next. Awesome. Awesome. Awesome. Well, Eddie, thank you so much for answering all of our questions. That’s going to wrap up the time that we have here today and I appreciate you being with us here. Thanks, Brad. Thanks for having me.