Data in Motion - Scalable Analytics Workflows for Insight-Driven Teams
Join Adobe Champion Ioana Maftei, Senior Analytics Manager at IBM, as she shares how to build scalable, governed analytics workflows that make the content supply chain measurable. Learn how to deliver the right insights to the right people at the right time, enable cross-team collaboration, and tailor Analysis Workspace projects to fit different roles and needs.
Hello everyone and welcome to Data in Motion. I’m Ioana Maftei and today I’ll be focusing on analytics workflows and the role they play in the bigger picture and how they can help you in your day-to-day analytics world. To give you a bit of an overview of what to expect from the next 20 minutes, we’re going to focus of course on workflows and why it didn’t matter. We’re going to try to paint a bit of a picture around the end-to-end analytics lifecycle. Then we’re going to go into governance and we’re going to end with a couple of demos and see how Adobe Analytics can play a role into collaboration and going into the data activation. Then of course, we’re going to end with some takeaways. But before we start, I thought I would introduce myself. A couple of things about me, I’m a Senior Analytics Manager. I’ve been with IBM for over eight years now. I’ve moved through a couple of different parts of IBM. This is my second year as an Adobe Analytics Champion, and I also lead the Adobe Analytics EMEA office hours. If you’ve ever joined any of those, you might have seen my face there. If not, please come join us next month. In terms of what else I like outside of work, I am a fan of everything to do with creativity, arts, and crafts. You can see in the pictures, there is some of the baking I’ve done and some of my painting alongside some of the pictures from Adobe Summit.
Let’s get into it. What you’re looking at on the screen right now is a diagram of the content supply chain. Now, you might be familiar with this and it just shows the different steps you have to take to start by having an idea, starting from a particular idea, going into planning and design. This looks at the steps you want to take when you start building your ideas about content and how it should be created and delivered.
Now, as you can see here, the last step there is analysis, and of course, this is what I will be focusing on. I wanted to take a step back and look at analysis in the bigger picture and how important it is in this content supply chain process because analysis is the one that closes the loop and gives all the feedback needed to go back into ideation, into testing your ideas, and coming up with new concepts, anything you would like to test and see how it works for your customers.
That being said, if we think about content as a predictive output and we’ve got things like Workfront or AM that streamline the delivery of that content, then we also need to think about the data that powers those decisions. In that context, we can see analytics as a nervous system that informs everything and if the insights and the key stats do not flow seamlessly in this process, then decisions will be the ones suffering. There are some common bottlenecks when it comes to analytics, and all of us have been talking about siloed roles, siloed data, teams not speaking to each other and not having the right data validation processes in place in order to make sure that the data they’re receiving is what they’re expecting.
You can also think about confusing reports or relevant dashboards and how those are not designed to cater to specific roles and of course, governance gaps. There’s always roles, different types of roles. People don’t know exactly what they should be doing, particularly in terms of data and analytics because analytics is not just for web analysts. Having definitions around those should help. Let’s see how we can bring a bit of structure in this space. What I tried to do is create a bit of an Adobe Analytics workflow. If you’re new to Adobe Analytics, this should be a great starting point for you to understand the different areas you should be looking after. If you’re already familiar with this, I’m sure you’ve had all questions or maybe blockers at some of these stages. I’ve started of course from technical planning. That’s your step where you are looking at translating business requirements into technical requirements. Then you go into tagging implementation, deploying your implementation, and having the right validation in place to ensure everything works as expected, the data is consistent, is validated. Then of course, the fun part where you are to do the analysis and curate reports or dashboards to specific roles, which leads in the end at collaboration and hopefully activation, where you can start using data to empower people.
Because we talked about governance, I thought it would be good to try to map all the roles that should be involved at each stage. Because as I said earlier, analytics is not just for analysts. You want to make sure you’re involving the right people at each stage in order to have a customized Adobe setup, Adobe Analytics setup that works for you and your business, and it’s helping you use the data to answer the right questions. If we look at technical planning, you don’t just need an architect or a tagging specialist. You also want to make sure you involved the product owner that looks after a specific feature or product, because they will know exactly what they need data for. Or you can look at the marketing team, the marketing lead, and how they would want to use that data in the future. This logic applies to each step of the process. If we’re talking tagging and implementation, you don’t want to have just a developer, you want to have your tagging specialist so they can work together to find the best solution possible. Same goes for validation. Maybe you want to involve your web analysts so they understand exactly what data is coming to them, how it’s enabled if it works the way they’re expecting. Then of course, in the last two steps, which I will be focusing on in this session, you want to make sure when you talk about analysis and curation that you involve, as I said, your product owner, your marketing analyst, and work together to report or dashboards that work for a variety of roles that can then go into collaboration and activation.
Let’s take these steps then one by one. Analysis and curation, and by curation I mean, making sure you help people looking at the right data. You curate for specific roles, you help them understand and have a quick look at exactly what they need, rather than them going to dig into large datasets and not knowing exactly how to explore your setup.
From this step, focusing on, I tried to give a couple of examples of how this would look like for different types of roles.
Having a look at this table, a couple of very common examples, a marketing manager, this is something probably everyone does. You have to build some campaign performance dashboards that help them have a quick understanding of how a campaign is performing, how the high-level KPIs are looking. It’s clear, you’ve got visualizations, summaries, this is quite straightforward. But then you might want to cater differently to a product owner that is looking at specific features and how they perform, if users are engaging with them, if there’s anomalies, if the journey breaks for some technical unknown reason. With web analysts, of course, that’s where you’ve got full data exploration, you’ve got them building those catered reports, so they do deep analysis, they’re involved in everything.
Something that’s not as common, I’ve included here an example of technical teams using Adobe Analytics. While they should not rely solely on coming to them to inform them that something’s broken, you can also use it to support their work and help them out. We’re going to focus on this in a couple of minutes, but basically is adding an additional layer to make sure it’s your team that identifies something that’s going wrong before the customer does. Then last but not least, I’ve included there an example of executive leadership teams using some of the feature Adobe Analytics has to be informed as well. Let’s look at some of these examples a bit more into detail.
I’ve mentioned the marketing manager one. This is the most common one where you want to make sure you’re building campaign reports that are easy to read, easy to access, something they can open on a Monday morning, and they can clearly see if the campaign is going better, a comparison, let’s say week on week, how it performed compared to the previous week. They want to have clear data summaries. Key summary metric is one of the features that I’ve used in this example. I think what you can do is try to build templates for your campaign reports and have your users, your analytics users be familiar with the setup and know what to expect when they get the campaign report. If you have templates, which Adobe Analytics has some nice templates to start from, a lot of inspiration there. If you don’t know exactly how to start building those. But if you start having templates, then it’s very easy for someone using those to do a comparison of something they have been running last year maybe or two years ago. If this is a consistent system, it’s very helpful for them. As I said, make it as visual as possible. There’s so many types of visualizations you can use. I’ve only included a few here, some graph with trend lines. This is looking at a comparison week on week of main KPIs in terms of a campaign, sources of traffic. I’m not going to spend too much time on this, but I think consistency templates, these are things that can really help having that step in your workflow a bit more structured.
Now, let’s go into the next example, which is the one that I mentioned earlier about executive leadership using Adobe Analytics features as well. In my past experience, I’ve had people, executives who were going to their own leadership team meetings, and wanting to be informed, which is a very empowering thing to have, to know the numbers, to be familiar to them, to be able to answer any question you’re being asked. It’s really nice. Something you could use in this scenario is mobile scorecards, which have been on for some time, but I know not everyone is using them. We’ve got a bit of a demo of how this works. But what I will say is not everyone logs into Adobe Analytics. This is something really helpful to have. Just getting your phone, everyone has their phone on themselves all the time, and checking some numbers. It’s something quite efficient. It works the same way as building a workspace. You go into Adobe Analytics, you just select mobile scorecard. As you can see here, it’s the same drag and drop feature where you can select any metrics you would like, and you have a lot of options in terms of your date ranges. All the rolling date ranges you are used to in workspace are present here as well. You can customize that based on your needs. Perhaps you want to look at last seven full days as I’ve selected here or want a longer period of time. You can just customize it as it works for you. The widgets you have in your mobile scorecards are clickable. If you want to get into a bit more detail and understand how things are performing on a lower granularity, in this case, day-by-day. You can add different types of visualizations here as well, which is quite handy. You don’t need to have just summary numbers. You can have a bar chart, a graph line, depending on what you’re trying to illustrate, but I would keep it as straightforward and straight to the point as possible.
What’s quite nice with this mobile scorecard is that you can customize your template as well. As you can see here, you can add or change the structure of it, how things are placed, and you can add context to your data. You can have a text panel where you include anything that the user should know. Let’s say you had an outage in your data, you want them to know that something impacted the numbers, so that the data makes sense to them. Make sure you always add context to the data and help people be as comfortable with the numbers as possible.
But yeah, this is with mobile scorecards. I think it’s a feature that should definitely be used more often.
Now, you also have a feature to preview how this would look like on a mobile phone, which is quite handy.
You’ll see here, as soon as this loads, it’s quite user-friendly and minimalist, if I could say. Moving on, I wanted to point out that you share your data. You need to make it not just available, but actionable and make sure people are ready to be proactive and take actions based on the data. I think what’s important to understand is that you need to meet people in the middle. As I said earlier, not everyone will be using Adobe Analytics on the regular, on every morning, every day to see, to check dashboards. You have a couple of different features you can use to bring that data to them. It could be sending an e-mail, people check their e-mails, or it could be a Slack alert, or maybe you have alerts set up to notify them when something significant happen. Let’s have a look at these couple of different options, starting with sharing, which is the most common feature and something that makes users love Workspace. That’s what made me be really keen on using Adobe Analytics because it’s so easy to share data with people. As you know, perhaps if you’re just beginning with Adobe Analytics, there’s a couple of different types of sharing you can access. You create the workspace, you have a couple of different levels of access you can give people, depending on whether you want them to collaborate with you or just have read-only access or edit a copy only. That’s quite nice, depending on, again, if you want collaboration with your team, if you have people that perhaps are not as familiar with Adobe Analytics and you just want them to be able to change the date ranges perhaps. Then additionally, there is another feature where you can actually create a shareable link for people to access the workspace.
However, I think this is something to be very careful because of course, every company has privacy regulations in place. You don’t want this to be accessible to everyone. There’s a nice feature there to enable this link to only work for people that are authenticated already. Just be mindful of these things. But again, something super helpful if you just want to curate something for people to access, only change the date range perhaps, and not interact with the wrong metrics or segments or anything like that. That’s on sharing. Then we’ve got, of course, scheduling option as well, which is very customizable. Again, something that I think it’s a big plus when using Adobe Analytics workspace. You’ve got two options to share either CSV files or PDF files. You can add a bit of a description so that when this lands in people’s inboxes, they know what they’re looking at, they know maybe where it’s coming from, who sent this to them. You can customize the frequency to such granular detail. This is something I like as well. Perhaps you want everyone to check campaign performance dashboard every Monday morning or just before a particular meeting you have in your calendar, or perhaps you have a technical team checking things every morning, making sure nothing broke down over the night.
One tip I would give here though is make sure you don’t schedule all your reports on a Monday morning at 9 AM because your inbox will be flooded with emails, and people will get a bit overwhelmed. They might not check it when it happens. Maybe put them with breaks and a couple of hours apart for them to get the attention they need.
Then the third one, it’s about alerting people. Then I keep mentioning working with your technical teams and having analytics work for them as well. Perhaps if you have a feature that you want to make sure it’s not a form and you have an error, and you want to make sure you’re keeping an eye on error levels and those not spiking, as the example I’m showing in the demo here. This is something that in past projects helped out a lot because we were able to see something going wrong before the customers reached that step. We could easily put redirect in place until the issue was fixed, rather than losing on that customer information perhaps.
As I said, this is not something that your team should rely on. They should have other measures in place to identify technical issues, but why not use this as a helper to have as much checks in place to ensure nothing is going wrong? I think discussing all these steps and going back to the first idea we started with, with the content supply chain and the role analysis has in it, I think it’s pretty clear to see how it closes down that loop and how having a bit of structure in place makes sure this step of the process is airtight. It does what it’s supposed to do, to empower people to go into ideation again, using those insights, using the data to power their decisions, and then start experimenting, starting testing with some hypothesis that they had that they maybe gathered from the insights they got. And then they can start including other tools to make those things happen. And if there were a couple of things I would like people to take away from my session, it would be that, again, Adobe Analytics is not just for analysts. They empower a broad range of roles. You just need to make sure you build your analytics for your people to work for them and empower their roles. Then we’ve talked a lot about workspace and dashboards. If only an analyst understands your dashboard, then it’s wrong. It’s not doing what it’s supposed to do.
And last but not least, I think it’s important to make sure your insights are reaching your decision-makers, you’re activating, you’re closing the loop. And if you’re not doing that, you’re just measuring for fun. That was everything from me today. Thank you so much for listening, for being here. And I believe we’re gonna go into a bit of Q&A now. Thank you. Thank you. Thank you, Ayana. Wow, that was packed with so many great insights into scalable analytics workflows. Let’s take some questions. Drop them in the Q&A chat if you haven’t already. Ayana’s experience running scalable workflows is something many of us, including myself, I can learn from. Okay, we’ve got, let’s see, our first audience question here. Keep those coming, guys. We’ll start with this first one though. All right, so I don’t know. This is what the person asked. I work on the implementation side at my company. These metrics she is showing, are these based on out-of-the-box metrics within Workspace? Or are these based off of custom tracking? Good question. That’s a great question, thank you. I think most of the stuff I’ve been showing in the demos was actually based on out-of-the-box metrics. So that would, of course, require you creating your integration, dropping the actual script on your website or app to make sure the integration is connected to Adobe Analytics. But you wouldn’t have to create specific metrics yourself to be able to access things like page views, visits, visitors. These are available as a kind of like standard set of metrics in the beginning. And then you can build on top of that, of course, depending on your platform or application, making it as customized as possible. But yeah, you can benefit from a great set of standard things from the beginning. And I try to use these when I have demos because many people will have different implementations. And I think these are relatable to everyone seeing kind of a webinar or something similar.
I appreciate that. I know a lot of folks on the line come from a bunch of different industries. And so being able to take something like that and maybe apply it to their own unique instances, that’s definitely helpful. Thank you. Okay, here’s one.
From your experience, what is the most challenging part in the analytics life cycle? Oh, that’s a good question.
I’m very curious to hear your answer. Maybe someone from the product team. I’m like, oh, taking notes. No, what’s, yeah, what is the most challenging part of that analytics life cycle? I think everyone will relate to the fact that there’s many challenging parts depending on where you are at in the process. And I can think of many examples, but I think just going back to the life cycle kind of overview I tried to give in terms of the different steps in your implementation and your maturity in analytics, I think a bit of a challenge can be to bring everyone in the same room and align, especially when you work with a lot of teams. If you have technical teams, marketing teams, different people will have different expectations and kind of ideas maybe of what analytics means and how it should look like. So kind of being that translator between technical and business requirements and making sure everyone understands the benefit of it and the kind of power of analytics and what they can get out of it, it’s very important. And once you nail that down, I think you’re there in a great step towards success.
Another thing I can think about just right on the spot is validating data. So you’re integrating your Adobe Analytics, everything works correctly, you’re kind of confident in your implementation, but then you’ll have other data sources as well that look at different marketing metrics. And everyone will want to tell a story, to be able to correlate all these numbers. And at times it can be very tricky because when you have different data systems, there will be data delta in your numbers, they won’t match exactly like for like. So bridging that gap and telling back to your stakeholders what everything means, how a campaign performs can be also be tricky. So just taking the right steps throughout your implementation and perhaps having some sort of learning for everyone to understand that it’s normal to have a bit of a delta. I’d like to say a 10% delta, not more, and to kind of have great definitions and kind of a great ecosystem in place for them to feel confident about the data and understand what everything means and correlate things nicely in the end. Great point.
I wanna dig in a little bit more into kind of what you were talking about there. So one of our questions, bump up because I think it’s interesting. How do you get buy-in from these different parts of the business? So you work at IBM, that’s a incredibly large, very complicated company. What is that like, like getting folks that maybe aren’t as familiar with data, aren’t as familiar with what the tools are, what you’re using, how do you get them to realize the value in what you are bringing and be able to get their buy-in to kind of get them to act, right? And I know that’s a really big question, but I’d love to get your thoughts and your insights on that.
Absolutely. I think some stakeholders at times can be a bit skeptical, especially if you come in with a new technology and something they haven’t worked with in the past. So I think that is an important step in the beginning to explain to them the benefits of a new technology, of Adobe Analytics, of having a customized setup. And ultimately you need to prove to different types of roles the benefit of the data, how can the data, the insight empower them in their roles and make them successful and how easy it is to access it, to be in a meeting, as I was saying earlier in the presentation, for someone in an executive role, be in a meeting, be able to quickly open up a mobile scorecard and be familiar with the numbers, know what everything means. So I think it takes a bit of a learning curve for everyone to prove that to them, to make them feel safe. I also like a lot of times to do live demos. So even if it’s something a bit more technical, I’m implementing analytics, I’m making a demo so easy for them to understand exactly how the data flows. So I’m showing them, look, I’m clicking on this right now. This is how it looks like in a debugging tool. This is how it looks like in the workspace. And they feel a bit more confident because I think a lot of people that come maybe from a marketing background, they’re used to sometimes to different metrics and they like to rely on that because that’s what’s familiar and sometimes it’s normal. But once they see a very clear demo, they start feeling a bit more comfortable and they embrace that. And ultimately they will end up using the data as an objective source to say, this is how we’re performing. We know it’s working because we’ve seen the demo. We know it’s accurate. Let’s start from here and see how we’re performing with this campaign or with this new product we’ve launched. So it’s all kind of working together, coming together at the same table. As I said in the presentation, Adobe Analytics is not just for a web analyst to sit in a corner and kind of analyze data. It’s for everyone to be empowered and understand that a marketing campaign or a technical issue can benefit from insight.
Oh man, I should have, I’m gonna go back and copy what you just said and just use that for our next marketing campaign on why people should be using our tool. But no, that was amazing. Thank you so much. Here’s another question here. What marketing channel attribution are you using? What it will take people to move away from last touch as we more logically, as we more logical attribution models are available now? That’s a great question. I think I’ve been also using a lot the last such attribution, but there’s been occasions where I wanted to dig a bit deeper so I’ve been kind of combining last touch with first touch, depending of course of what I was trying to get out of the data and the purpose of the analysis as well. I think you are right. We are, you know, more data is available. We have access to more complex systems and configurations. So it’s definitely something that people should be looking at and spending more time, let’s say. But I think there are available tools in Adobe analytics to kind of configure things the way you prefer them, depending on your company’s needs, on your marketing channels, the way you’d like to attribute different events to the success of a campaign.
So, you know, you can definitely try to create a strategy that works for you.
I think last channel works very well in a lot of scenarios with a lot of the clients I’m working with, depending on their needs.
But I do combine it with first touch and also taking into account other things, not just marketing channels. So the tracking codes, the different times of the year, we have campaigns running, we have a lot of information in captured in the kind of URL parameters that we track, and we combine all these to kind of try to paint a better picture in terms of where people are coming from, what these channels are and what’s performing best. But, you know, I think we all need to adapt our strategy as maybe you’ve been familiar with this. People, as Eric said, they’ve started using AI. A lot of your traffic will start coming from those AI tools as well. So that’s something you need to factor in when you do your analysis. And maybe that’s something we were not thinking about a couple of years ago, but it’s definitely something that needs to be taken into consideration and adapt a bit around it. So I would say that’s definitely a good topic for our next session to focus on. Yeah, there’s a lot there. We know things are moving fast. So it’s gonna be interesting how, you know, we as an analytics group kind of bring that new problem into the mix, for sure. We are, excuse me. So we’ve got a little bit of time. We’ve got plenty of questions. So I’m trying to kind of pick the best ones here. So here’s one. When getting metrics directly to people’s phones, while it is great for internal analytics teams, as they would be helping monitor, how do you handle when there are data issues? Wouldn’t it be a problem if leadership finds too soon? That is a great question. And I’ve actually had some of these scenarios in the past. What I would say is you always need to add context to your data. And that’s why I’ve talked about alerts and like being the first one that finds out when something’s broken, because you can actually avoid a scenario like that. I’ve worked with teams where we would, you know, we would be working on our implementation. We weren’t yet in the place where everything was proactive. To identify an issue quickly by having alerts set up. So we would get some of those messages saying, wait, something’s wrong on this page. Or there’s, you know, there’s a broken link on the CTA. Why hasn’t been captured by the team? So I think that’s why you need to invest time in your implementation and get it to a point where analytics becomes proactive. So it tells you exactly what’s wrong before a customer finds out. But to kind of go back to the question, I don’t think it’s an issue. I think, you know, you’re being transparent. Things can break down and it’s great when they’re captured immediately. They’re notified or kind of alerted to a technical team. Everyone is on it and can identify an issue.
And, you know, add the context, as I said, like make sure something breaks down and let’s say yesterday data is not showing up because your tags just fell out of a sudden. Then add the context to your workspace or a mobile scorecard and say, please, you know, be aware there’s been a technical issue. This is being investigated. We’re making, we’re taking all the measures. We’re looking into it. But you need to be aware that that happens. So there’s nothing’s wrong with the website. It’s just a technical issue, for example.
But, you know, there’s also ways then you can rectify that afterwards. So I would say in my experience, just being transparent about it is the best thing ever. You know, technical issues will always happen. You’ve seen it happen with major big companies. It’s not something we can avoid. So be transparent and proactive about the data, I would say. It’s a long answer, but hopefully.
No, I think that’s a great summary. It’s better to be upfront rather than trying to sweep something under the rug, right? I can get, that’s where you get in trouble. All right, we just have a little bit of time. I’d love to get kind of your final thoughts that you’d love to share with folks on the call. Thank you. Well, first of all, thanks for having me. This is my second time doing Skill Exchange and I absolutely love it. There’s a fantastic team behind this whole process. So I hope everyone enjoys it. And I hope everyone kind of managed to take something away from this session. If it’s, I don’t know, from the demos or from the actual process, I think the moment we bring a bit of structure in the analytics world, which at times can be a bit forgotten because there’s other priorities. We’re building new tools, we’re developing websites. Sometimes analytics can be an afterthought. And I think if I want something to be taken from this session, besides the takeaways I’ve shared earlier is to try to always advocate for analytics in your company and make sure it’s at the forefront of people’s mind and they realize how important insights are for success and growth. But yeah, thank you. Thank you everyone for tuning into my session. Really appreciate it. Oh, it has been a pleasure. Thank you so, so much.
We’ll let you go reluctantly, but thank you for your time and appreciate all that you do for us.
Mapping Roles to Analytics Stages
Overcoming Analytics Challenges
- Bridging Silos Align technical and business teams early to ensure shared understanding and buy-in.
- Data Validation Expect minor discrepancies between systems; set clear definitions and educate stakeholders on acceptable deltas.
- Driving Adoption Use live demos and role-specific dashboards to build confidence and demonstrate value.
- Transparency Proactively communicate data issues and context to maintain trust, especially with leadership.
Addressing these challenges ensures analytics drive real business impact and foster a data-driven culture.