Most of your problems are due to (a lack of) Marketing Data Governance

Edward:
All right. We're back for another episode of the Marketing Intelligence Show, and we're back with Evan Kaeding, Lead Solutions Engineer at Supermetrics. So Evan, great to have you back on the podcast.

Evan Kaeding:
Great to be here, Edward. Thanks for having me.

Edward:
Yeah, always a pleasure to have you on the show. Now I want to start with something you said when we were in a meeting a couple of weeks ago. It was a super interesting thing you said: "Most marketing problems are marketing data governance problems." So to kick things off, can you just dig into that and explain what you mean by that?

Evan Kaeding:
Yeah. So when we talk about marketing problems, what I often see is that marketing problems are often actually realistically masquerading as marketing and data governance problems. And what exactly do I mean by that?
I mean that if you're faced with problems in your marketing, in terms of you don't know which campaigns are running, you don't know if you're getting close to budget, if you're underspending, if you're overspending, if you're not sure if your attribution models are working correctly, you're not sure how much credit you should assign to campaigns, or you're not sure if your pipelines are running correctly, you're activating that data for the right personas, for the right audiences. I would argue that a lot of these, in fact, can actually be traced back to marketing data governance problems.

If you have good marketing data governance, your campaign reporting and monitoring can actually be significantly improved so that each one of your campaigns has a very specific set of objectives, a set of metrics associated with them, and dashboards created to monitor and proactively maintain those.

If you're looking to understand which campaigns are the most effective, well, again, that's going to come down to making sure that you are tagging a set-up, your campaign naming conventions are broadly aligned, and that you've got the tools that you need in order to do the analysis that you might need, whether that's MMM, MTA, last-click attribution or anything like that.

And similarly, if you're activating on data, if you're pushing out those custom audiences, or are sending emails to customers, you want to make sure that your segmentation is on point. Sending out targeted messages to individual customers based on your own internal data is going to be, of course, most effective if you have proper governance and proper organization of that data.

So when marketers talk about, "We want to make our marketing more effective, or we want to measure the effect of our marketing more," a lot of that comes down to what kind of data we are generating and what quality control over that data looks like. And ultimately, those fall well within the domain of marketing data governance.

Edward:
Yeah, absolutely. And we'll dig into the model and what marketing data governance is in more detail later on. But another thing that I wanted to pull out at the start is that we were doing some research recently, which showed that over 50% of marketing leaders struggle with unreliable, siloed, and scattered data. And essentially, they don't trust their data. So I mean, how can you operate effectively if marketing and data leaders don't trust their data?

Evan Kaeding:
Yeah, it's really difficult, is what it is, and you end up doing a lot of the same work repetitively, which is unfortunate. There's enough work in marketing to go around, as I'm sure we all know. And if you want to ensure that your team has time to focus on what's actually going to matter, what's going to make your brand differentiated. Whether that's by putting the right ad in front of the right person at the right time or whether that's crafting a message that's truly unique and showcases your USP to your customers. You want to be focusing on automating everything that is automatable.

And with that, part of it becomes, let's make sure that our campaign monitoring and reporting are as automated as possible. Let's make sure that our insights and analytics are as streamlined as possible so that we don't have to recreate the wheel every single time. So once you've invested in a taxonomy that shows your marketing campaign naming conventions, you can proactively monitor your dashboards and build those reports. Whether you're building those reports for in-house stakeholders or clients, once all of those pieces are automated, you can pick up the slack with the time that you would otherwise save or otherwise be spending on some of these tasks that can be automated and invest it in higher ROI activities. As I said, investing in messaging and differentiated capabilities.

So, of course, if you can't trust that data, it's going to require you or someone on your team most likely to go in manually, look at that data, scrub it, make sure that it's correct, fact-check it, give it an eyeball check. And we often see at Supermetrics that reports that are not fully automated can take anywhere from five to 10 times as long to produce. That's five to 10 times as long as, realistically you should be working on hand curating that data. So that's really what the trade-off ends up becoming if you can't trust your data, someone's going to have to go in and do that work manually. And that manual work takes up a lot of time and energy and detracts from what are ultimately very strategic products and projects your team might otherwise be producing.

Edward:
Yeah, absolutely. So let's take a few steps back, and get into the solution marketing data governance. So I mean, it doesn't sound super jazzy, but it is critical, as you mentioned. So two parts here, what is marketing data governance, and why do marketing teams need to invest in strong marketing data governance?

Evan Kaeding:
Yeah. So first, what is the model, and how do we discuss it? And then the second is why this is important. So the three pillars of marketing data governance we talk about here at Supermetrics are data access, data quality, and data security. Let's dive into each one.

Data access is making sure that the right people have access to the right data to do the jobs that they need to do. This is really just fundamental. So make sure that you have your team given access to the paid ad accounts, the analytics accounts, and the websites; you have access to all of the tools that you need in order to ensure that you can do the analysis that you need on your campaigns. That's a big part of just making sure that things work because I can't tell you how much time is wasted in credential sharing, especially on the agency side between clients and between media planners on the agency side.

I've been through the onboarding of many different customers, and I've seen in the agency world, and I've seen all kinds of password sharing, I've seen all kinds of insecure practices. I've seen weeks and weeks, sometimes months, that it takes just to get access to the right data in order to do the analysis that's actually necessary. So data access, although if you have your systems in place, this can end up being done in a matter of hours, potentially days. For many customers, this isn't solved yet, and it can take weeks or months to get access to the right data to do the analysis you need to do.

Data quality is what we refer to when we talk about things like naming conventions, media taxonomy, and tagging infrastructure. Those kinds of things. So we put marketing out in the world, which generates a lot of data. It will generate cost, clicks, impressions, and some conversions if we're doing things right. But when that data comes back, it's important to analyze it and see what campaigns are working most effectively. And if you don't have an organized system around that, that job could be very difficult. Same thing if you're not checking to see if you're overspending and underspending. Of course, you could use cost caps, for example, in your media trafficking tool, whatever that happens to be, to ensure you're not hitting those. But are you necessarily getting the optimal allocation there? Real-time or near real-time data monitoring associated with that will be the best way to ensure that you can spend close to or exceeding your targets and monitor that for the different ways you look at it. So that's data quality.

Data security is what we talk about when we talk about it's a sub-component of data access where we say, "Okay, data access is unblocking the barriers to make sure that people have access to the right data." Data security is a couple of different components. Number one, we need to make sure that nobody has access to data that they shouldn't have access to. So that's the basic information security component of marketing data governance. So making sure that hey, these vendors or these external contractors that we're working with, of course, marketers work with website developers, agencies, and often freelance writers, for example. And too many times, I've seen that access has been granted to someone who no longer works at the company or a freelancer who hasn't been hired in years and years. So this is part of who has access to the data; it's also part of data sovereignty.

As a company, where is your data stored? Is it GDPR-compliant? Are you complying with local laws and regulations? Is it in the country in which your operations are taking place? And then further, are you respecting the wishes of the customers you're working with, obtaining consent during that consent? And do you have frameworks to ensure you're remaining compliant? Data security is the, I'd call it, the hard piece of data governance. Data access is the accelerator that ensures you can get things done and data quality, making sure that when the data comes in, you can start using it to understand the effectiveness of your campaigns and, ultimately, the success of your marketing initiatives.

Edward:
Yeah. Awesome. So that was the what of marketing data governance, and I think we could go into those three pillars in more detail. So we had data access, we had data quality, and we had data security. So let's go back to data access at the top. So if we go into a bit more detail here, who should have access to what kinds of data within an organization? What are the typical data access models? And we hear terms like — people are data owners or data guardians, or data stewards is another term I've heard. Could you open up the topic of data access more and talk about who should have access within an organization?

Evan Kaeding:
Yeah. And we've thought about this quite a lot at Supermetrics, and it's one of the things that a lot of marketers like about our products. So we've invested in a set of features called Shared Connections. And this is a tool where marketers with access to the data sources that they want to use for analysis can, with their own credentials, log in to these platforms using Supermetrics and basically use all of the tools of Supermetrics with their data source access — for example, to analyze data in Google Sheets and Looker Studio in a data warehouse environment, use our RAPI, et cetera, et cetera.

The key here is we optionally give that data source owner the ability to make that connection either private or shared among their team members. And that's important because, for example, something like Google Analytics that's not going to contain personally identifiable information, or at least it shouldn't. If you have proper data security and governance in place, you'll likely be in a case where that is okay to be shared among your immediate team members. You probably want only some of your freelancers or all of your contractors, or all of your writers to see access to it. You may do in some cases, but overall, the relative risk of that data being used for malicious purposes is relatively low in most cases. Exceptions do happen, but that's a reasonable analysis for analytics data.

Now, if we're talking about CRM data, that's a different story, especially depending on the sector that you operate in. If you're talking about connecting to HubSpot or Salesforce or something that's going to contain personally identifiable information, again, first names, last names, email addresses, phone numbers, home addresses, those kinds of things, that's going to have a pretty important level of security associated with it.

You may only want the actual owner of that platform to use it. That's okay, and we understand that. At Supermetrics, you can connect that to Supermetrics, and that person can mark that connection as a private connection. That way, only that person can access it. So we want to make sure that you can use Supermetrics in the line of products with as many data sources as you want but also maintain the security and convenience associated with those if there are cases where you need to restrict the use of that data for certain pieces. That's the data access piece. Should we jump into data quality and security as well, Edward?

Edward:
Yeah, that would be good. Moving on to the second one, data quality, you touched on that earlier, but digging into that in more detail. And this is something we spoke about when you were last on the podcast––talking about decentralized and centralized data is the classic crap-in-crap-out. A piece that your data will only be as good as what comes in, and the actions you take will only be as good as what you see from the data. So let's get into the second pillar. And this is, in many ways, almost the million-dollar question for many marketing leaders, CMOs, and anyone who works with marketing data. How do you then ensure that you have good data quality within your organization?

Evan Kaeding:
Yeah. And I'll first talk about why this is so important. Because what we're seeing now at Supermetrics is that we have this global phenomenon, which you can call the loss of signal phenomenon. We're starting to see cookie acceptance rates on websites across the developed world drop, which is reasonable. People are clicking, and people are required to opt-in or opt-out of web tracking on the websites they use, and as a result, these cookie consent rates are dropping. As a result, we're not seeing nearly as much coverage from the analytical tools that we use to be able to see.

Similarly, if you're used to relying on paid acquisition channels to tell you how many conversions they reported using something like Facebook audience conversions or Floodlight Tags or something like that, much of that's no longer able to be used or trusted reliably because it's being used with a variety of different signal based inputs. Things like machine learning, generated conversions, things like sampled audience insights, those kinds of things. And so what we see for customers is they end up using a variety of different conversion reporting tools, whether that's in the platform, whether that's analytically based or whether that's looking then at the transaction system of record, whether that's a CRM or ecommerce platform or whatever that happens to be.

And so we're seeing that, as a result, companies need to move to more statistical or inferential steps to understand the effectiveness of their campaigns. Statistics is fundamentally a science, and if we're doing science, we need a fairly structured approach. We need to go into it with a hypothesis; we need to go into it with a set of methods and generate some conclusions. And ultimately, that needs to be a fairly rigorous process, and it's very difficult to do science unless you have that process.

Similarly, it's very difficult to measure the outcomes of campaigns unless you have that process built out. So when we're talking about doing science and doing statistical measurement, let's put it in practice here, a lot of companies are now, for the first time, investigating, "Hey, can MMM help us out in terms of measuring the effectiveness of our campaigns?"
The answer is usually yes. But, and there's a big but, you need to ensure that your company's data quality processes are robust enough to ensure you're prepared for MMM. Because MMM's not going to be very useful if you don't have the data quality to back it. Let's talk about why. So you might take all of your spend, look at all of your conversions, and see some correlated relationship.

But keep in mind all of your spend is likely advertising for different products, different stages of the funnel, and might have entirely different outcomes or different audiences that they're using. And it's not going to be very easy to differentiate between those if you don't have good data quality in place. Companies looking to leverage these statistical techniques need to meet a minimum threshold of data quality for them to bear fruit in many cases.

Edward:
Yeah. Absolutely. And I think just listening to your answer reminds us that marketing, I think, happens at the intersection of art and science. So you have the art side, the creative, the messaging, the empathy, the relating to your audience with how you speak about what it is you do in the world you operate. And then there's the science half, which is, as you spoke about there, the rigor, the data, the hypothesis, the structured approach, and in a way, the science allows the art to work.

And yeah, I think great to get insights in terms of how you can ensure that the data you work with is of high quality and that you can trust, which is what we spoke about at the start. So from here, let's dig more into the third part, which is the security piece. Super critical as well, particularly with the amounts of data we're generating. You spoke a bit about some things that we need to consider, but how do you ultimately ensure that your marketing data is safe and secure?

Evan Kaeding:
Yeah. This comes down to an organizational decision and will vary based on the industry in which you operate. The answer will be very different if you operate in the healthcare space or the financial services space vs. if you operate, for example, in the gaming space or the media and publishing space.
Of course, every organization's going to have some very bare minimum standards for data security, but these will change based on what kind of industry you're in, what kind of local regulations you have, and what kinds of rules you need to abide by as a company.

So it's hard to give a one-size-fits-all approach for this, but it ultimately becomes a fundamental pillar within data governance that needs to be worked with the larger portion of the business. As a marketing leader, you should have a wide purview of data access and quality. That's firmly going to be your responsibility. You'll share the responsibility for data security with your organization based on the industry in which you operate. So that data security responsibility will be shared between you, most likely your IT or your operations org or something like that.

And you should be looking at things like, where's our data stored? Do we have data sovereignty? What kind of data are we storing? What kind of data are we collecting? Do we have a requirement to purge data if notified about a particular customer? What is our notification process to our customers if some breach occurs? These kinds of things. As a marketer, do you need to have all of these answers by yourself? No, not necessarily. That's what internal company operations and security teams are for as well. So involving them early to ensure that things are secure and compliant is really important for many companies.

Edward:
Yeah. And this is a good segue into a follow-up. So you spoke about the different teams and stakeholders. Who owns marketing data governance? Would it be marketing, data, both, or someone else? Who owns it?

Evan Kaeding:
Ultimately marketing data governance should be owned by the marketers. That's my opinion because marketing data governance is going to be driven by your marketing strategy, and here's why. Because if my strategy is I want to rapidly experiment with different channels and activate against new audiences while simultaneously keeping the existing demand coming in with a set of bottom-of-funnel campaigns. That's great, and that's my strategy. And I'm going to design my data access models. I'm going to design my marketing data governance around both those pieces, all while maintaining my marketing data security that I've collaborated with or negotiated with my internal stakeholders. So marketing ultimately needs to be the owner of this.

Do I see this as a VP level, a CMO level, or a marketing manager level? For any company that is putting a sizable amount of money into digital media into offline media, the marketing data governance strategy needs to be a conversation that is probably led by VPs of marketing but is signed off on by CMOs. Because ultimately, the entire marketing strategy should influence marketing data governance. And if it doesn't, well, you're probably not necessarily going to get the buy-in that you need at all levels of the organization. Because if this isn't coming from the top down, who's going to bother with making sure that their campaign naming conventions comply with whatever taxonomy has been decided on?

Who's going to ensure that the right people have access to the tool they need to get done, and we've removed those accesses once they are done with these kinds of things? Who's going to make sure that data is stored safely and compliant if nobody's necessarily on the hook for it or, for example, it's not baked into the strategy coming from the top down?

So I truly see this as a responsibility for organizations large and small, whoever owns that strategy. Understanding that for smaller organizations, this might end up with the lead marketer, and understanding that for larger organizations, this might be, for example, someone in charge of regional marketing. Or we also see, for example, companies come to us who are global, where they say, "Hey, we need a global media taxonomy because when we roll up, and we do this big MMM exercise, all of our countries are reporting in completely different ways and we have no way to assess what's working and what's not." Globally, we might be spending €100 million across five different channels, but because everybody's using different conventions and buying things in different ways, it's impossible for us to get a sense of what's working and what's not.

So the degree to which your business is localized also is going to impact this as well. So if your business operates similarly across the different geographies in which you operate, then I'd say you could largely have a very top-down approach to marketing data governance. But if your business is wildly different in each of the industries or each of the countries in which you operate, it might make more sense to have a more federated marketing data governance model where each country or each region or maybe each brand has their own set of strategic objectives and has their marketing data governance plan associated with that. So it's hard to give a one-size-fits-all answer, again, depending on the industry, the size of the company.
But it's also good to understand that you can have two models. You can have a federated model where people who are closer to making decisions on the ground are making some of those decisions so long as it rolls up cleanly to the top. And you can also have a top-down strategy if everything is reasonably concentrated and reasonably repeatable across the different areas of operation.

Edward:
Yeah. So marketing would own it, but obviously, as you said, data is a key stakeholder here, and marketing and data collaboration is often insufficient or perhaps even overlooked. Historically, within marketing, we've spoken a lot about marketing and sales alignment; on the B2B side, we speak a lot about marketing and product alignment, but more recently, we're hearing about marketing and data alignment. So how should marketing and data teams work together to ensure they can ultimately trust their data?

Evan Kaeding:
Yeah. I've talked about organizational structures in which marketing analysts are embedded within marketing teams. And this starts to make sense at, I think, usually around if your paid media team is four or five people, I think right around there it starts to make sense to have an embedded marketing data analyst. Your paid media team is 50 people — you could probably have a dedicated team of marketing analysts associated with that. If you've got thousands of people, having that one-to-five ratio or so, I think, is pretty reasonable for embedding that analytical practice in the team.

And it's very difficult for teams to contextualize the data generated as a result of their campaigns because of someone who knows the data side of the business and understands that specific domain. So if you talk to a data engineer, that data engineer is most likely, unless they're very in tune with marketing, going to have a very different process for understanding what a paid acquisition campaign is and how that differs from, for example, an awareness campaign.

Of course, Edward, you and I know that you're going to look at different metrics for that; one will be a CPA, cost per acquisition. The other's going to be, am I getting good reach? Am I getting efficient reach out of CPM? Am I incrementally reaching new people? That's going to be challenging if you're trying to push all of those levels of intricacy down to a centralized data platforms team or data insights team. And so having embedded business analysts is a practice that we've seen in the BI industry coming out of operations and finance for, I'd say, probably for the last several decades, really.

Is having someone on the team who understands the operations and can help essentially the marketing team make sure that the data that's generated, as a result of those processes, is clean, it's categorized, it flows into the centralized data warehouse or whatever systems are being used to analyze it and can advocate on behalf of that team for resources that are needed. Whether those are technical resources or guideposts like comprehensive marketing data governance frameworks or ensuring that that team has access to the data or technical resources they need.

Having that business analyst function is helpful. Whether you decide that that person reports to a marketing person or a data person that probably depends on the nature of your company and your relationship with stakeholders, but that's what we've seen be most effective — if I talk to a data engineer for our customers and they truly understand the difference between paid acquisition campaigns and awareness campaigns, that's fantastic. But those people are super rare, and so most of the time, we see that embedded data stakeholders on those teams are usually the best way to go.

Edward:
Yeah. For sure. That makes a lot of sense. And obviously, you talk with many folks from marketing teams and data teams about their marketing data. So what are some of the common challenges you're seeing in the market when successfully implementing strong marketing data governance practices?

Evan Kaeding:
The biggest challenge I see is the lack of executive buy-in, which is probably the biggest piece. Because if it's not at the forefront of, not necessarily, every meeting, but if it hasn't come from the top down, it's going to be really difficult to get everybody to subscribe to this. It's also probably not going to be an all-encompassing framework. And it doesn't necessarily need to be useful, but it does need to be comprehensive enough so that everybody can look at it and say, "Hey, okay, based on this data collection framework, here's how I need to do my job." If it's been built or crafted within one team to serve one team's objectives, then guaranteed it won't work very well when we bring it over to the paid social team, the SEO team, or the organic social team.

It needs to be comprehensive enough to show all of the different areas of marketing within the business, online, offline, digital, and the other pieces I just mentioned. So not sitting down and taking the time to craft this at the executive level is one of the mistakes that I see making. Otherwise, you're just microscopically optimizing in a smaller section of the organization. That's probably the first piece. The second piece isn't having the tools to follow up on this. So we see when companies adopt Supermetrics. Usually, they're pretty excited about, "Hey, let's get all our campaigns in one place, and let's build these marketing data governance reporting dashboards, and we're going to make sure everything's super clean, and everything's super measurable."

And that happens, and that's super great for maybe a month or two. And then after that, it's like, "Ah, shoot, we need to launch this campaign tomorrow." "Well, what about these things?" "Oh, forget about it. We need to get it launched." And then suddenly, "Hey, we launched this campaign, and it doesn't really follow any of the rules we had outlined, but it's fine.

The media's in the market." "Okay, that's fine." And then one thing leads to another, and then eight months down the line, someone says, "Hey, we're going to go kick off the MMM process. Let's go ahead and apply the taxonomy to our media." And then you have five or six campaigns with a lot of spending behind them, and you realize, "Hey, these don't fit the walls that we defined, and it's going to be hard to categorize these." "Yeah, that's going to be challenging." And so the follow-up of that, I think, is important.

So we see our most successful customers implement systems to keep themselves in check. If a campaign doesn't meet certain standards associated with naming conventions or audiences or something like that, there's a dashboard that shows it. We're seeing as well centralized administration and control of who has access to the data and ensuring that all data access is run through something like Supermetrics, where you can essentially use Supermetrics as the interface to all of your different marketing reporting platforms.

That helps ensure that you limit the number of people with administrative access to your marketing platforms, and you get people access to the data they need to do the work they need to do. The same thing with data security is close collaboration with the administrative teams who are in charge of IT and security and things like that. To ensure that at the same time, we're following rules and we're following up, we've all agreed and decided on this thing, but keeping those follow-ups in place is important. So I'd say those are probably the two main things: executive buy-in and the follow-up piece, making sure it doesn't become a big exercise that nobody ends up following in the future.

Edward:
Yeah. For sure. And I think, as ever, not getting that buy-in is going to make it almost impossible, I think, to get anywhere with implementing some form of unified marketing data governance structure. So yeah, super good point. It's great to hear a bit about Supermetrics. What are some other ways you're seeing people use Supermetrics, for example, to help with some of the challenges of marketing data governance and getting started with marketing data governance?

Evan Kaeding:
Yeah. So we've got a couple of features I can discuss here. With data access, like I mentioned earlier, we have Shared Connections, which is powerful. That allows you to limit the number of administrative users of your marketing platforms and share that access pretty easily to the underlying data with the teams who need access to it to do whatever analysis they have. That's one really good way to, number one, ensure that people have access to the data they need. Number two, reduce the security footprint associated with what you need. So Shared Connections can be great for that. The link-based authentication sharing is pretty sweet as well for agencies who want to onboard brands quickly.

So you can pop into Supermetrics and generate a few authentication links, kick those links over to the client, and a couple of hours later, they're authenticated. They have taken their client onboarding from anywhere from three or four weeks down to a couple of hours just by getting access to those systems, federating that access out to the clients who own the system.

So that's just one powerful way we see Shared Connections being used. Another feature that we see being used is a feature we call Custom Fields. So Custom Fields is a tool in Supermetrics that allows you to essentially transform your data before it goes to your destination. Why is this important? It's important because we actually see customers using it to run their data through this set of Custom Fields, and these Custom Fields can be used to check and see whether or not your data is actually conforming to the naming conventions that you have set out to try to find. So let's use an example. Maybe you're separating your campaign names by underscore. You've got the market that you're targeting, you've got the product that you're targeting, and maybe you've got the funnel position as well.

So you can set up Custom Fields in a way that will automatically check it, is this campaign using a market naming convention that's on my approved list? Is this using a product that's on my approved product list? Is this using a funnel position that's on my approved funnel position? And if it is, great, you get a star. And all of a sudden, that kind of data flows through. It's showing up in the dashboards, and by the time you're ready to do, whether it's a last-click attribution analysis, whether it's a statistical analysis or assessment that campaign's performance in some way, you know that data's good and it's guaranteed because it's run through Custom Fields. If it's not, then you can set it up to have a variety of different tools or a variety of different outcomes. What I usually recommend is saying, "Hey, if it doesn't meet the selected criteria that you want it to, just flag it as non-compliant."
And then you can pipe all that data into a dashboard that shows, "Okay, here are the campaigns that are running. These are the compliant ones; these are the ones that aren't." And hopefully pretty quickly, if you've baked this into your campaign trafficking process, you'll be able to take a look at that dashboard maybe once or twice a week, and you'll be able to see, "Hey, that campaign we launched on Wednesday, it doesn't look like that's fully compliant. Let's go in and fix that." You fix it right then and there while it's still fresh in your head rather than 6, 12, 18 months from now when you're trying to remember, "Gosh, what did we run this campaign? Oh man, I need to dig into every single asset to figure out what this was." Fix it right as it goes live; it's going to save you and your team a lot of time going up ahead.

That's one of the ways that we see customers using Custom Fields for marketing data governance. You can use Custom Fields for a variety of different things like currency conversions; you can use it for categorizing your campaigns, you can use it for adding additional information with lookups or classifications, but using it to put data governance best practices in places is one of the strong use cases we've seen for it since we've come to market with it. I'd say those are probably the two main pieces. Obviously, data access is important, data security as well for customers that have a requirement to store data in a particular cloud vendor or in a particular region. Of course, Supermetrics doesn't store your data, so if you do have a requirement to store data, we can send that to whatever cloud environment in whatever region you need to as well. So we can help out with those data sovereignty pieces as well. So a couple of different pieces where Supermetrics can help out, thankfully with Shared Connections, with Custom Fields, and with data exports to your data cloud of choice, wherever that happens to be.

Edward:
Yes. Gold stars for everyone. And any final passing words of advice on marketing data governance?

Evan Kaeding:
Yeah, I would say the best time to start is now because what we find is that a lot of people don't realize that the data that exists in these marketing data platforms, for example, a Facebook or a Google or a Google Campaign Manager or whatever that happens to be, that data is not always going to be there for you. Each one of these platforms, the same thing with Google Analytics 4 and Facebook, only retains data for three years. Google, at this point, retains data for a lifetime, but who knows when that could change? Google Campaign Manager only retains data for two years. Google Analytics only retains data for 14 months.

And so, if you want to actually build a strategy around statistical measurement, you need a lot of historical data. And if you want to build a strategy around historical measurement, you're going to need good campaign naming conventions. And so the best thing to do is get started, store that data, clean up the data that you have now, and then put processes in place to make sure that the data that you generate in the future is also clean. So that when you come to try to solve this problem, two to three years down the line, you've made sufficient investments in marketing data governance, and you're not historically looking back for data that either you don't have access to anymore or just hasn't been cleaned or fit to your current process. That's pretty much my biggest recommendation.

Edward:
There you have it. Awesome, Evan. Well, thank you so much for joining us at the Marketing Intelligence Show.

Evan Kaeding:
Of course. Thanks, Edward. Much appreciated.