Karen Bellin, Mirum Agency - Data & Analytics Ecosystem and the Marketing Cloud

October 18, 2018

Data & Analytics Ecosystem and the Marketing Cloud

By Karen Bellin of Mirum Agency

Slide 1:

Thank you so much Brian. Thank you so much for having me. I hope everyone is enjoying the virtual summit today. So, as Brian explained, I’ve been in this industry for a while. Like many others in this industry, I’ve had to broaden the way I approach data and analytics, consultation, and implementation as the tools that we’ve all started with ten years ago, or for some even longer. Despite, analytics tools are maturing into marketing clouds.


For instance, when I work on an Adobe implementation now, I’m working across AM, target, audience launch, and analytics to create a serviceable analytics ecosystem. That’s at least five different systems, and they’re all changing at different rates. AM versions keep increasing. Launch is replacing the Adobe dynamic tag manager. Every few weeks when I log into Adobe Analytics, I see an update. And that’s just the Adobe Marketing Cloud. So, it’s a moving target and each part of the marketing cloud matures at its own pace providing new opportunities to collect, integrate, share, and activate data. Now that’s what’s kept this area so interesting for me over the years that I’ve worked in it. Today, I’m going to share a chart or today I’m going to share how I chart a path through this marketing cloud to establish data and analytics capabilities.

Slide 2:

So, if you have any questions as we go through this, feel free to save them for afterward and even to follow up over email.


Slide 3:


So, looking back as I’ve been in this industry to pre-marketing cloud days. It wasn’t long ago that the digital analytics association, the professional organization in this field or career development was called the web analytics association. Maybe in the future it will be called the marketing cloud analytics association because the marketing ecosystem is becoming much less fractured thanks to the advent of this idea of a marketing cloud that’s been driven by demand for marketers and championed by the digital analytics association to connect each consumer touchpoint using data.

Slide 4:

Even so, I still address the challenge of these fractured reporting systems in a methodical way as I would any data challenge, but understanding the consumer touch points in the marketing cloud in context with defined measurement pillars.

Slide 5:


At least now with the marketing cloud, there’s a pretense that all of the different systems should work together to create a seamless user experience. I mentioned measurement pillars. A measurement pillar is a key area that a business wants to focus on over the long term. It answers the question, “What does long term success look like?”


An example of a measurement pillar for an insurance company might be titled “Growth and Acquisition.” With long term success described as increasing the amount of plans offered and the number of people who sign up for them.

Slide 6:


In this session, I am going to share some ideas for how to approach getting your arms under your current data and analytics ecosystem and steps to take to effectively mature the ecosystem to better enable answering those not so easy to answer business questions all in the context of the marketing cloud.


The ideas I’m going to share are ones that I use to save myself time to better be equipped to answer questions and to achieve outcomes. Working in agencies over the past twenty years, I’m often juggling lots of different projects. Having a way to keep straight what I’m trying to do and having a bigger goal to achieve with my clients, actually saves myself time.


So, moving target because each solution in the marketing cloud will be maturing at it’s own pace, as I mentioned. It’s happening at the same time that you or I am working to try and mature a data analytics ecosystem across those same solutions. Most organizations that I work with, at least, don’t put a pause on redesign, marketing experimentation, and replatforms. All which have data and analytics requirements of their own and often happen in parallel of trying to mature the data and analytics ecosystem overall.


By prioritizing, making progress towards a serviceable data and analytics ecosystem, aligned to measurement pillars, should result in time saving, more satisfactory, answers to questions, and better achievement of outcomes for you.


Slide 7:


So, by saying that we as marketers and data and analytics practitioners are better equipped to deal with this challenge than we were before.  


Slide 8:

I mean we can address whole use cases, not just simple questions. Business questions are common. How did that do? What worked? What didn’t work? What did we learn? What should we do next? Is this good?


Slide 9:


Satisfactory answers are not as common for the reasons I mentioned earlier. Primarily that reports for different facets of marketing are found in different places that the person trying to answer the question may or may not have access to. In cases where all of the data is in one place, there are still issues with understanding context as well as issues with understanding how the data fit together with answering the question.

Slide 10:

So, even having all of the data aggregated in one place doesn’t fully solve the issues. It just makes it faster to figure out that you don’t have the context needed to understand how the data fits together to best answer questions. It truly depends.

Slide 11:


So, that’s why it’s not unusual to get different answers depending on who you ask. And for the answers to be caged in all sorts of caveats. First thing I do to get out of this situation which is very uncomfortable of trying of answer a question without having all of the information you need is to reframe the question. Understand the context for the question is how I start to do that. Understand it in the context of a use case.

Slide 12:


The data and analytics ecosystem overall can be matured around use cases. The use cases can be used to drive alignment in maturing the data and analytics ecosystem. Use cases replace question and answer pairs. They help you reframe the question and the answer. So even seemingly innocuous questions can open up Pandora’s box of possible answers none of which may even answer the question.


Use cases can frame questions in ways that can help guide alignment towards a more definitive answer. Use cases can be defined well ahead of the question being answered and use cases can often be completed without the question ever having to be asked.


So, here’s an example of some common business questions on the left and some use cases that could be put up that can avoid having that awkward feeling of trying to answer a question without having all of the information you need.


So, let’s take the first one here that’s listed as Target Achievement. This use case replaces the question, “Are the results good?” So, I’ve mentioned the measurement pillar earlier. That long term success goal. Within each measurement pillar, something like growth and acquisition is the example I gave earlier, you’ll have objectives. Different things that you need to do to achieve that larger long term goal. Within each objective, you’ll have tactics that you’re running to achieve those short-term goals. Each of those tactics will have a stream of data associated with it that you can use to measure the effectiveness and so forth. So, by setting a target for your bigger goal, your larger KPI metrics, and understand the budget and baselines for those KPI, you can start to set a target and monitor towards the target and adjust the target as needed. So target setting is something that’s important for your KPI metrics and those other metrics that feed into the KPI can also have targets. So, if your KPI is leads, you may be able to set a target for leads, and based on your lead target, you can set a target for visits to your lead gen form. So visits to your lead gen form become one of your effectiveness metrics. So I categorize metrics according to each use case.


So your KPI are associated with your measurement pillars and help you understand your long term success. But in the meanwhile there are things that you’re trying to achieve, your objectives, that you can measure according to effectiveness metrics. For instance if you set out to reach a certain audience in this use case, you would say, “What do we need to have in place from a data and analytics standpoint to know that we’ve reached that audience?”


Efficiency metrics or that use case around overall efficiency would say, “What do we need to have in place to know our tactics are executing efficiently as possible?”


In the diagnostic use case, we’re looking within each channel and tactic to see what can be done to improve and optimize in that tactic itself. You have a lot of metrics in that diagnostics bucket.


Finally, in the last use case that I show here, the journey map, is where the integration points are. So all of those metrics that we’re using to measure different aspects of the campaign might be better understood or better put in context by understanding the journey in where those metrics are generated and where they can be joined to tell a more holistic story.


Slide 13:


I’m going to talk more about these different metric types as well as the journey map throughout the presentation.


Slide 14:


So, marketing clouds give us a good start to understanding what the data and analytics ecosystem is expected to do. It allows for broader use cases than just looking at analytics as a standalone capability. It lets us look at analytics at something that addresses use cases not simply a place to get our answers. A mature data and analytics ecosystem is one that ties neatly to the measurement pillars. An effective data and analytics ecosystem will reflect what the data is intended to be used for. From implementation, data integration through the recording, modeling, and activation.


Slide 15:


It’s important to remember the measurement pillars. The measurement pillars reflect the long term vision of success for a theme such as growth and acquisition. Framing the data and analytics ecosystem in this way helps to ensure that it’s going to really be able to influence an organization, especially in terms of data driven decision making.


The first step of making this connection between the marketing cloud and the measurement pillars is to map out what the measurement pillars are.


And I’ll provide these slides so that you can work through this process. It’s ten steps that I’ve laid out here. This is the first step. Just defining what your measurement pillars are.


I’ve found success in bringing different groups into a room to talk about this and understanding this in context of individual goals and responsibilities to coalesce around what the measurement pillars are.


Slide 16:


Then, you can assign objectives to each measurement pillar. Different pillars should have different objectives. For instance, I gave an example of a long term goal in growth and acquisition to generate leads. So under that we may have several objectives. One of them might be to educate the sales force about what’s being sold and the appeal of that to the audience. So that would be a specific objective underneath a longer term lead goal.


The objectives are the goals we must meet that fall under larger measurement pillars, and they answer the question, “What are we trying to achieve in the near term?”


Slide 17:


The third step then is based on those objectives to define the full journey with the mix of tactics that will leave a trail of metrics that we can use to effect the near term goal for objectives and their contribution to the long term goal achievement of the measurement pillar. By tactics I mean all of the tools, processes, and efforts that support the objectives. It answers the questions, “How do we accomplish our objectives?” It could be, a tactic could be, an email campaign, a landing page, a TV commercial. If we look at a goal, a longer term measurement pillar around growth and acquisition, that we’re measuring in terms of lead growth, and our objectives include things like education of our sales force or education of our audience around how they can become a lead, those things are things that we can drive, that we will have tactics for, that we can measure specifically.


Slide 18:


The next step is to identify those metrics. Knowing the tactics, you can find your metrics and you can organize those as a full set. A full set of metrics being your KPI, your effectiveness metrics, your efficiency metrics, and your diagnostic metrics by channel and tactic. Some of these metrics may be your KPI, and those are the very critical metrics that give an indication of how you’re doing. They answer the question, “How do we know that we’ve chosen the right path?”


The KPI are important first at findings baselines and targets and mapping the journey across data and consumer touchpoints that will inform data integrations so that your data can be joined to tell its complete picture and not just used side by side.


Slide 19:


This is what it will ultimately look like as you’ve mapped out your pillars, objectives, tactics, and metrics. We’re not done yet. We’re just finished, that’s one through four. So in this way, you’re meeting the marketing cloud systems halfway. Where the technology is available and the opportunity to integrate the technology is available through data at different touchpoints. It’s important to have a goal in mind around your measurement pillar.


In this example, I’m showing a measurement pillar of acquisition. Then the associated metrics in the set that I explained where you have your KPI metrics which are your long term goal and where it’s being measured. In this example, I haven’t filled out the baseline and target, but if you have them you would be able to fill in your baseline and target for your KPI. Once you’ve done that, you can start to fill in baseline and target for the different metrics that bubble into the KPI where applicable.


I mentioned earlier, effectiveness metrics. These are things that are more tied to different objectives or things that you need to do to generate leads. If we want to generate a lead, we need to make sure we’re reaching our target audience. If we want to generate a lead we need to make sure that people are calling us. Those calls and reach on your target audience are the effectiveness metrics where leads still remains your KPI.


Efficiency metrics is how well we’re doing from a cost perspective. Based on the budget of the campaign and how much money is being spent, what are we spending to get that growth? Are we spending money as efficiently as possible? These efficiency metrics are always your cost per metrics.


Then we get into the diagnostic metrics, and this is where you have a very long tail of metrics and the ability to pull different levers within your tactics as you need to to optimize toward a bigger picture goal. So everything is bubbling up into your KPI. Here, I have a few tactics listed. Display advertisement, email, non branded search, landing page. You can see that each tactic has its own set of metrics and they come from a different source.


So here, by identifying the pillar, the associated metrics, where they come from, baselines, and targets, you have a launchpad for them to find the integration points for each touchpoint from a consumer as well as a data standpoint. Meaning that these metrics may happen sometimes together and sometimes apart, and that’s the next thing that we’ll do in step five is starting to map that journey.


Slide 20:


So mapping the data touch points in the journey.


Slide 21:


So here’s an example of what that data journey looks like. It’s a journey that’s across internal systems, consumer touch points, and marketing cloud systems. What I’m showing you here right now is a work in progress. All of this journey work, especially in this marketing cloud that’s continuing to evolve, there’s going to be more going on every day and so you really need to think of this work in the ecosystem as a living, breathing, initiative that can be updated as input is gained and new things are learned.


But essentially, the colored dots each represent one of the rows in the table that I showed earlier. Each one of those colored dots could be a data source. Where something is being measure from. And then each step in the campaign from internal systems to consumer touchpoints through to measurement of outcomes, there could be a different set of data sources that come into play.


So on the left here I’m showing the planning stage where nothing is consumer facing at this point. A marketing planning a campaign. As they plan the campaign and come up with the tactics and channels that their going to use, they can assign an ID and phone number, I’m still using the lead generation example, to that campaign and that would hit internal systems as well as call analytics in order to achieve even just setting an ID and phone number to each campaign, they’ve already touched two systems. The campaign ID is then incorporated into every associated channel tactic so now we’re touching the different ad platforms, email platforms, and so forth. The green and blue dots there.


From there, that ID gets incorporated into a destination URL for a landing page. So when we move into the second column for activate, these campaigns are running, they’re driving to a landing page, they also have the telephone number in some cases. So, you can go to the landing page or you can call immediately for some tactics. Some tactics may force you to go to the landing page and then call. Whereas others may let you call without visiting the landing page. So here we have another point where things can get a little brittle if you’re not passing through these IDs and telephone numbers and all this connective tissue. Once the user hits the landing page, all things going well, destination URL, including the campaign ID, we tap into our web analytics, our call analytics and our ad platforms to capture and reflect what’s happening in terms of attribution for where this person came from and what they’re going to do next.


On the landing page, you can submit the form in this example, or you can call a phone number. In both of those cases we need to continue to pass that campaign ID into the new data sources to allow for integrated reporting and different analytics capabilities down the road. For instance, in order to do customer journey analytics, you need to keep passing this data from data source to data source in order to join the data and really understand in a tribute what happened in your campaign from the time it was planned through to the time someone became a lead and everything that someone did throughout that process. And this can be done completely anonymously.


So, like I said, this is a work in process. There’s even more going on here than meets the eye, and even as I described it, I know I mentioned additional details that aren’t reflected here. So, again, it really needs to be done as a work in process and it needs to be continually built upon. As the marketing cloud evolves there become new opportunities to better track and connect things and different ways to do that. So this is living, breathing, but in general, it shows the user flow, the data sources, and all the places something can go wrong if things aren’t set up right.


So, having this map reflects a shared understanding of what the data can tell us. What the use cases it can be used to address are. Again, it helps with the conversation to share this with people, bring in alignment, help people understand all the complexities and the bridging that is needed to really measure effectively within a marketing cloud.


Slide 22:


So, now we understand what our metrics are, we understand the journey and where the metrics come into play. So we can talk about the ecosystem and why it’s important to have a data and analytics ecosystem.


Slide 23:


So, the measurement pillars, the objectives, the tactics, and the metrics alone with the journey map are identifying our integration points and illustrating how the data and analytics ecosystem is set up today to let you address use cases related to each measurement pillar and what changes may be needed to the data and analytics ecosystem to make it better, more connected. So, maybe in your map that’s what you want to happen, but that’s not what is happening. So that can be a key to some of the steps you can take to mature your data and analytics ecosystem in a marketing cloud journey.


So, first in doing this, you want to make sure you’re crisp on what the benefit of doing this is and why you’re doing it. And for that, I’ll remind you, that a well coordinated data and analytics ecosystem that is used for effectively tracking, reporting, and optimizing user journeys across multiple touchpoints can be a foundation for broad marketing cloud and business alignment that is necessary for long term success in achieving outcomes for any given measurement pillar.


Slide 24:


So, when I look at building out the data and analytics ecosystem, this is the third exercise. The first one being the metrics table and the grouping and the set of metrics under each measurement pillar is the first exercise. Again, I feel like that can be achieved very well through a workshop where you’re bringing in people from different parts of the organization to map that out.


The second one is the journey which is something that, based on that initial workshop, can be drafted and shared and refined with different users and looked at as a work in process knowing that it’s a moving target and capabilities are evolving.


The third step is to say, “What are all of the tools and capabilities that we have in our ecosystem?” And mapping those as a currency and then defining what the future state should look like. I like to bucket the ecosystem along five areas. You can take some liberty with this if needed in terms of how you understand the data being used in your organization.


Just to explain how I see it, I look at the analytics capability in terms of how the data is going to be used from analysis, modeling, and activation standpoint. For some instances, maybe you have the data and if someone asks you a question you can kind of describe what happened in some way. So, I would call that descriptive ad-hoc capability. You may also have implemented A/B testing and be doing A/B testing on some section or all your pages or some pages. Maybe you don’t have the ability to do a full customer journey analytics because things aren’t, your touchpoints aren’t all measured or your data isn’t all in one place or you don’t understand how one data set relates to another data set. So, here for analytics capability, I listed some of the things that you may be wanting to do with your data and you create this ecosystem mapping. You can create a current state version and a potential future state version where your current state version lists the things you truly are able to do with your analytics and in the future state list the things that you’d like to do.


The second area from the analytics ecosystem is for managing data. These are the tools that you have to store, protect, share, integrate, transform, and otherwise manage the data. Here, things that come to mind are a dashboard or even a self service reporting capability where people have access to certain data sets. Your site analytics likely has a place where the data is made available and reports and so do other systems that might be pulling data from your marketing, your call analytics, your marketing analytics and so forth, your media analytics has something to log in to. Even SEO, you can think of Google search console as a place where data is stored and made available.


The next area is the data collection, and these are the tools you use for gathering the data. Again, don’t take this as comprehensive, this is just an example of what your ecosystem might look like. Collecting the data with a tag manager or having different site analytics tools. Survey is another way that data is captured. Third party marketing tags and call tracking.


The fourth area are the connective systems, and this is some of the tools that you might use to bridge one touch point to another, one data source to another. One way is dynamic number insertion. You know when somebody calls a number that it’s associated with a certain campaign dynamically. That would be one linkage. Another one is to use that campaign ID in the URL or your destination landing pages, and then there’s other systems that people use to better bridge a one data source to another data source to tell a whole journey story.


And then finally, last but not least, and this is really the foundation of everything is what are even the touch points where the data is available to be collected or activated from.


So once the current state is mapped in the ecosystem, say these are the tools we have to do these five things. We have tools for using our analytics, tools for managing our data, collecting our data, tools for connecting our data across touchpoints and we have our touchpoints in general. You’ll want to have that for your current state as well as for potential future state. And then as progress is being made you can show incremental updates on that current state mapping. A potential future state is recommended because it's going to be a moving target. You’re going to periodically adjust this based on how other aspects of the marketing cloud are maturing and of course your priorities. The future state of the data and analytics ecosystem will also change if new channels come online.


Slide 25:


So, now, the rubber hits the road and we have to talk about developing a roadmap for maturing this data and analytics ecosystem. The road map is going to show the steps you need to get to a more mature state and is truly a conversation. According to a report in Forester, in 2017 only 45% of business decisions were made using quantitative analysis versus gut feeling or opinions. 45%. That actually was a decrease from 2016 when it was 49%. The primary cause of this gap was studied, and it was determined that there is a lack of alignment on business outcomes between lines of business, business insights, and technology teams. Maturing the data in analytics ecosystem will certainly be a conversation between those teams. So the roadmap that’s being developed is something that can be used to gain by-in for multiple parties and to use to get more information in certain areas to use to build a rapport around the initiative and to get input on things that you need to move forward with. For something as complex as maturing the data and analytics ecosystem, you will need continuous engagement with multiple groups in your business and a well documented grove map that can be managed to can be something that you can use to continually engage your stakeholders with.


So again, a well coordinated ecosystem for effectively tracking, reporting, and optimizing user journeys across multiple touchpoints can be a foundation for broad marketing cloud and business alignment necessary for long term success and achieving outcomes for any measurement pillar. Each step you take to achieve this level of coordination will have its own benefit.


Slide 26:


So, this is going into the roadmap. It’s the benefit. The overall benefit of the ecosystem, and then for each step that needs to be take to mature your ecosystem should have a benefit as well. So, example of a step that can be taken to mature data and analytics ecosystem is to integrate call analytics with your site analytics. And in that case the benefit would be that you’re closing the tracking and reporting gap between site visits and offline call conversion.


Slide 27:


So, for each step, such as that one that I just explained, you would identify the key milestones that need to be worked through to complete that step. This is very high level. The ideas that each step can be torn off as its own project to be run. Some could run in parallel, some could run waterfall one after the other, but they could be planned as its own project. You want in your roadmap to have some of the key milestones so that as you’re checking in progress and communicating progress, you can talk about where you are in the process not necessarily get into the nitty gritty details.


For example, for the call insight analytics integration, the milestone steps might be:

  1. Document your requirements and have a checklist.
  2. Architect the solution.
  3. Configure and enable integrations and reporting.
  4. Deploy the solution.
  5. Validate the solution.
  6. Update reference documentation.
  7. Update your recording dashboards and other analytics capabilities.


In fact, the steps for each initiative may look very similar to that with some variation. The other thing you can do in identifying these milestones is sort of size them. One of the tricks to sizing them at this higher level without knowing all of the details is to gage the level of involvement needed from different teams. If it’s something one person can do independently, it’s a small task. If it involves multiple teams, it’s a large task. Everything else that falls in between is medium. You should be able to have your steps loosely sized and with key milestones as a start in this roadmap. That way you can kind of mix and then you can order them. What can we do in parallel? What’s the dependency on the other? What do we just have resources, we can knock out this one then that one? I like to mix it up taking on something big, and then following it up with a couple small wins while getting the next big thing moving.


Slide 28:


The next step is to socialize the roadmap with those different groups that I mentioned, and to continuously engage the stakeholders, the people who are going to be affected by this data and analytics ecosystem in the roadmaps to drive alignment and set expectations around that potential future state. The expectation should be, it is a conversation. That input will be provided as you’re doing this and you would update the roadmap as you do. That’s part of the socialization and the conversation. Because as I said, it’s a moving target.


Slide 29:


While you’re working to mature the ecosystem other initiatives in the ecosystem are also going to be worked on. Some may involve you, some may not. So, this is initiative like a section of a site getting redesigned, a new marketing experiment being launched, a new section of the sight being replatformed, a new microsite coming online, a new channel being introduced, new landing strategy being evaluated. All of these things will have demands on the current state data and analytics ecosystem. As you’re managing trying to mature that ecosystem, but having these more immediate needs to address, you can try to address them in a way that advances the roadmap. For instance, if a digital campaign is running and there’s a question about calls, answer it using the call analytics which will eventually be integrated with your site analytics even if that integration isn’t complete yet.


So, this is where the time savings that I mentioned comes in. You can kind of, once you kind of have a vision of where you want to go, you can save time by then just trying to align that vision even with your current state activities and the immediate demands on your time. That’s the precedence for how you’ll be acting in the future when the data and analytics ecosystem is actually mature.


Slide 30:


So the other way to continually engage the organization around the roadmap include: in general status meetings having some general high level status update on where you are with the roadmap. Some steps that may be in progress or completed already. Then to have regular meetings on roadmap items with the people who are directly involved in those steps. That’s another way to keep different people engaged.


Thirdly is to distribute the proof points from each step. So, for that call and site analytics integration example, to share those reports as they’re available, not waiting for the whole ecosystem to be mature to show the results. You’ll have results for each step. Finally to connect the dots between what you’re doing day to day with the vision for the future state. Again, keeping it alive and integrating updates as you get inputs.


Slide 31:


So, the data and analytics ecosystem is expanding, and it’s because of this marketing cloud. It’s requiring marketers to be well-versed in more than just site analytics implementation.


Slide 32:


I hope that based on the presentation today you are going to be able to have a way to take a methodical approach to socialize your work and show results as you mature the data and analytics capability that will address use cases rather than attempting to answer questions. Keep it flexible because it’s a moving target. As the data and analytics maturity is evolving, so are all of the capabilities around it. It’s a time investment in the beginning, but it will save you time down the road, making it a worthwhile endeavor.


Slide 33:


Thanks for being here today and for your commitment to putting in the hard work that’s needed to have the capability not for answering questions but for addressing use cases. Now I’ll hand it back to Brian. 

Previous Video
Lara Fisher, Blast Analytics & Marketing - ROI vs. ROAS Understand The Value of Your Marketing Efforts
Lara Fisher, Blast Analytics & Marketing - ROI vs. ROAS Understand The Value of Your Marketing Efforts

Learn how you can use digital analytics solutions to compare marketing campaigns, measure performance, and ...

Next Video
John Lovett, Search Discovery - Data, data, everywhere: Leveraging an actionable data strategy to understand your marketing ROI
John Lovett, Search Discovery - Data, data, everywhere: Leveraging an actionable data strategy to understand your marketing ROI

Is your data strategy (or lack thereof) causing you to lose sleep at night?