Everywhere we look, we see sparkling success stories of smart campaigns delivered by sophisticated, data-driven marketing. However, the harsh reality is that too many marketers are stuck working with manually-filled spreadsheets and reports that deliver zero value or new insights.
Moving towards more intelligent ways of getting valuable information, which can drive quality decisions and outcomes, is a must. But it’s not an overnight jump, rather a tough and long journey that will take time and effort. And it involves marketers embracing new skills, new tools, and new ways of looking at data.
This guide is here to explain in detail the five stages marketers need to go through on this enlightening journey to full analytical maturity.
This guide is designed to cover the key steps marketing leaders and their teams need to take to elevate their existing data-driven marketing initiatives. While we have taken care to keep this as simple as possible, it is a complex topic and as such we assume a certain level of knowledge on the part of the reader.
If you are new to the topic or feel some of the information here is getting too complex too quickly, don’t panic! We recommend reading our guide on data-driven marketing first to get you up to speed.
When you applied for that dream job in a marketing department, the last thing you thought you’d be doing with your days is shuffling numbers around spreadsheets. And with the rise of digital channels, things got even more complicated.
Everybody keeps talking about using advanced analytics and how you need to embrace it, but very few actually show you how it should be done. So all you end up doing is spending hours copying and pasting numbers from one tool to another, trying to make sense of it all.
Or, if you’re one of the luckier ones, the company you work for has moved away from these mundane and inefficient practices. You now have shiny dashboards and perfectly designed reports, pushing you into a presumption that your marketing department is finally ‘analytically mature’ and ready for all upcoming challenges.
But as it turns out, recent research has shown that this elusive dream is only a reality for a few. In fact, 41% of surveyed marketers still highlight manual data wrangling as a significant challenge. So instead of doing the job they were hired for - creating and executing dazzling marketing campaigns - there are still far too many people in marketing departments wasting time on manual data operations.
It should go without saying that these practices, relying on employees executing tasks that could’ve been done by a machine, are not what anyone would call ‘analytically mature’, or even remotely close.
But what then does being ‘analytically mature’ actually mean?
As with many other terms, this tends to be a point of contention (and not just within marketing) but for argument's sake, let’s define it as the following:
“Analytics maturity can be described as the evolution of an organization to integrate, manage, and leverage all relevant internal and external data sources into key decision points.”
(“Analytics Maturity Models: An Overview”, Król & Zdonek, ResearchGate)
This evolutionary process is not limited to marketing only and includes changes in many areas of a business, which is why becoming more analytically mature can be seen as a journey down a long and winding road. It is a multi-disciplinary project that takes time and resources before it produces results, not an overnight technical swap from a spreadsheet to a database.
Recognizing the length and complexity of this journey is essential in getting the right results at the end of the road. Because the end result of this transformational process should not be just another set of dashboards and reports, but a system that will allow you to make better, timely decisions that lead to improved outcomes of your marketing efforts, and ultimately the success of the company you work for.
But why is progressing down this analytical maturity road so important for marketers? You develop campaigns, think of ad creatives, organize events… and sure, you need some data to make sure you’re on the right track, but does this really require you to become experts when it comes to data?
The simple answer is: Yes, it does.
Not because you need to change careers and become data analysts instead of marketers, but because marketing as an industry is becoming more data-focused each day. Businesses that are embracing the power of data are actively gaining a competitive advantage, while less data-savvy companies are missing out on crucial insights that can improve their overall marketing performance and help grow the business.
Becoming more analytically mature means moving away from only looking backward at what happened, but starting to look forward and answer questions about future developments. Shifting from hindsight to foresight, as shown in Gartner’s analytics ascendancy model below, is the only way to increase the value of the data that you have at your fingertips.
Research has also shown that more data mature businesses are far more resilient to market shifts than those that are further down on the scale. So, the higher your organization is on this analytical maturity scale, the better chance it has to succeed in an increasingly competitive marketplace.
To visualize this journey, we’ve put together a scale that depicts the various stages a company goes through as it improves its analytical infrastructure and skills and starts utilizing data in a more advanced way.
This scale is largely based on Gartner’s Maturity Model for Data and Analytics and depicts five key stages for reaching analytical maturity in marketing:
At this stage data in a company is siloed, used on an ad-hoc basis, and really creates more confusion than a benefit for a company. All data operations are relying on manual effort of employees from different departments, taking too much time and causing too many errors.
As the company moves forward, there’s a concerted effort on collecting and visualizing data. But, without a parallel flow of organizational improvements throughout the company, this results in only a slightly better overview than in stage 1.
Reaching stage 3 means that the organization is much more systematic in its data collection and analysis. More time is spent on understanding and analyzing information coming from the data than on collecting and organizing data itself.
Once a company has reached stage 4, the data-driven culture is much more widespread throughout the organization. Decision-makers have started relying on data for both strategic and daily decisions, bringing a synergistic effect and impulse for growth.
Ultimately, data analytics becomes an integral part of strategic planning and decision making, because it provides insights not only on past events but also a projection of future developments that is significantly more powerful than the educated guesses found in stage 1.
However, while this analytical maturity model may show the path you and your organization need to take, it doesn’t tell you how to do it. This is why we have developed the following 5-step guide for elevating your marketing’s analytical capabilities.
As with most problems in life, the first step is to admit that you have a problem.
Acknowledging it, and uncovering the scale of the repercussions stemming from outdated and inefficient data practices, is a crucial step in trying to find your way out of the messy vortex that is marketing data.
Manual data operations waste two of your key resources: time and people. Add to this the low trust in reports based on error-strewn data, and you’ll end up in a pretty tight spot.
Here are some clear signs you can look out for that indicate your organization is still stuck in Stage 1 of the analytical maturity journey.
The first and most common cause of lost productivity and poor data quality is relying on manual data integration. We are referring to the process of exporting large datasets from various sources, joining them into spreadsheets, and then spending even more time manually harmonizing them, just to produce a very basic set of insights into how your marketing is performing.
At first, your team spends just a couple of hours exporting and consolidating reports from a couple of tools. As the company grows, so does the number of data sources, and the amount of resources spent on manually integrating all that data grows exponentially. At a certain point, this amount of data surpasses your marketing department’s capacity to process it, and you end up with erroneous reports and miserable employees.
Let’s take a look at why juggling with spreadsheets scattered around your company’s network could be potentially harmful, even dangerous:
Breaking out of this vicious circle is not easy, but employing the right tools to automate this process could help tremendously as it will reduce manual work and the probability of errors.
Research on the effects of implementing an intelligent data management solution in a global consumer electricals company has shown a potential to reduce time spent on these activities by 75% and an improvement in data operations efficiency by 80%.
The next question is how efficient are you in using the data you have at your disposal. Just think about it - if it takes you up to a week to create a report, another week to analyze it, and one to two more weeks to implement the learnings, you’re constantly a whole month behind the companies that can get it all in one go, in a matter of hours.
The real problem is deeper. Spread thin around dozens of different platforms, most of which act like a black box when it comes to how they actually spend assigned budgets, marketers are often unaware of how their campaigns perform. At least until they get a detailed report, usually once a month, from their data analytics department or agency.
If everything went well and without any glitches - great. Give yourself a pat on the back. But if something broke, not only are they looking at faulty data but perhaps also facing major losses with missed sales and tens of thousands of their advertising budget wasted on non-performing campaigns.
This is why finding things out sooner rather than later is essential, and for this, you need to have the right data at the right time.
All of these problems multiply in an agency environment as they handle several clients at the same time. The amount of work needed to acquire at least the basic information is just humongous with each new data export and created report increasing the probability of an error. And nobody wants to explain to a paying client “we got it wrong last month” and that the budget has been spent on campaigns that didn’t perform.
Sure, part of the problem is clients requiring monthly reports in static PowerPoint presentations instead of dynamic dashboards, but that doesn’t mean that the agency itself needs to remain in the “dark ages” of data management. Having a system in place that delivers the right data on time can prove crucial for client satisfaction, but also employee satisfaction.
Look at it this way: if your agency talents spend less time doing boring tasks, they will have more time to think about creating and delivering great campaigns, which in turn will make the client happier and the agency management less worried about the future of the company.
Introducing any change that will improve the agency’s data maturity level is a clear win-win for all parties involved.
For all agency marketers, we have prepared a two-part strategic playbook for transforming agencies into data-driven creative factories.
You can check out part 1 here and dazzle your managers with fresh ideas on how to use data (and your time) better.
In a time when decentralization is a popular buzzword throughout the business world, rooting for a data centralization initiative might seem out of place. But if you plan to break down silos within your organization, this is imperative.
Moving from stage 1 to stage 2 is perhaps the most important step a company can take in its analytical maturity journey so we’ve broken down some key steps you need to take.
Once you have established as part of your overall marketing strategy the metrics and KPIs that you need to track, the next step is to identify the sources that provide the necessary data for them. Although this may look like a straightforward job, it isn’t, because the number of necessary data sources grows constantly.
Let’s start with some of the basic ones:
Even though you see only three bullet points here, in a mid-sized company this equates to over 20 different tools, platforms, and consequently data sources, while larger companies are already in the hundreds when it comes to the number of platforms they regularly use.
Once you have made a list of all the platforms and individual data points within them, make sure to take a minute and think about how they correlate and fit into a larger picture of a company-wide (or at least department-wide) analytics strategy.
This will help immensely with the steps that come later down the analytical maturity road.
With a list of data sources and necessary data points finalized you can move to the next step, which is fetching the data. At the risk of repeating ourselves, it should be clear that doing this manually is not a good use of your time and resources, so automating this process is a must.
And yes, Adverity is a platform that does this (and lots of other things), but that’s not the point here.
The point is that you need to develop and execute a data strategy that will allow you to move upwards on the data maturity curve. This is not just a matter of pulling ones and zeros and pushing them into a centralized database.
It’s also about the way you organize your data because that will determine how you can use it in the future.
Once you have the data fetched, transformed, and harmonized, you can move it to the location of your choice. In most cases, this is some kind of a database, and depending on the rest of your company’s tech stack and IT infrastructure, it can sit on a local server or somewhere in the cloud.
Deciding where and how to store data is not normally the job of marketers but your colleagues from IT can help you find the best solution based on your needs. Regardless of the technology that you’ll use, the point is the same - you now have a centralized dataset, from which you can extract valuable information.
Before we move further, it’s important to emphasize the importance of data governance. Data governance is a set of internal rules and processes that determine the way data is being handled within an organization.
Although this seems more like a technical or administrative issue handled by IT or Legal departments, a marketer needs to understand the basics of how data is collected, organized, and used.
This will not only help ensure legal compliance, primarily in the area of privacy and personal data protection, but it will also allow the marketer to see the complete scope of available data and, potentially, come up with new ways of using it.
If you want to learn more, check out our blog post on the 7 building blocks of good data governance.
Once you have gathered all the data you need in one place, you can start to think about how to visualize it in a way that will allow you (and your bosses) to get a better overview of your results, and make conclusions and decisions faster, with more confidence. Based on your data strategy and the metrics and KPIs you’ve decided to closely monitor, you can create a wide range of different types of charts and dashboards.
The story of data visualization is a wide and complex one that is beyond the scope of this guide. And, we’re sure that by now you’ve mastered the art of creating some basic linear charts and rich colored pie charts.
Nevertheless, a well-thought-out and designed dashboard will allow you to look at data from a different perspective, and think of new ways to combine it and use it. Apart from displaying things that are more or less obvious, data visualizations can be a very useful tool to uncover new, unexpected insights.
Let’s take a look at a couple of examples.
Regardless of the type of information you are trying to relay, using visuals instead of text is always better. A study by MIT has shown that humans can understand the meaning of an image in only 13 milliseconds, while processing words takes twice as long.
Of course, transforming data into visual form doesn’t automatically mean that this information will be clearer. Without a definite idea of what the key data point is, and a vision of how it should be represented, the message can be lost. This means that you need to be extra-cautious when designing your dashboards, as they should make the story they are telling clearer, not more opaque.
Looking at the numbers in a table sometimes might be enough, but to really understand what these numbers are saying, it’s much better to display them in a chart. A simple linear chart does the trick in many cases and is not to be underestimated as it can speed up and simplify the process of understanding the data and making proper decisions.
Observing website visitor data in a time series line chart will immediately show you relevant information, such as seasonality, weekly oscillations, growth or decline over time… and more. And if the linear chart is not clear enough (i.e. because of too many oscillations) just add a trendline to it and immediately you have a clearer picture of long-time performance.
Besides the most commonly used chart types, such as linear, bar, or pie charts, there are many others out there. One of the particularly useful ones is a map chart, which can have many unexpected uses.
For example, showing sales results from various regions in a table can clearly show you which ones sell more than others. And for the sales department, this information may be enough to move on. But for operations and logistics, understanding which neighboring areas sell more by seeing them grouped on a map could be a crucial piece of information needed to improve delivery times and reduce operational costs.
On the other hand, spotting adjacent areas with poor sales might help marketing to create special campaigns targeted at these territories only.
For inspiration on how to introduce new views to your reports, read our blog post on 20 chart types you can use to visualize your data
When you look at a simple linear chart of a very basic metric, like the number of users on your website, you might notice some discrepancies that can raise alarms if looked at in isolation. Understanding a wider context can be crucial before jumping to any conclusions and a well-designed dashboard with carefully selected chart types can be of tremendous help in these situations.
For example, a spike or a dip in website traffic could point to a technical problem with the website, but also can be a consequence of greater or fewer visitors coming through paid channels. By easily distinguishing organic from paid traffic in a single chart, website managers can quickly see in which area the problem lies and continue to investigate in the right direction instead of raising false alarms to multiple stakeholders.
Besides helping marketers to navigate the mountains of available data and dozens of reports, data visualizations are an extremely useful tool for company-wide data democratization. First of all, you need to make sure all your colleagues within the marketing department have access to all data relevant to their daily duties. Once you have that covered, easily sharing insights with colleagues from other departments, sometimes sitting on the other side of the globe, is vital in speeding up the information flow within the organization.
By seeing the progressions of leads through stages in a practical chart and being able to drill down into details, sales can understand the incoming pipeline much better, as well as the quality and complexity of leads coming from marketing. In an alternative example, giving HR the opportunity to dig deeper into the website data of the hiring pages will help them to understand the profiles of job applicants, and try to optimize open position descriptions to fit better to the wants and needs of the talents they want to recruit.
Once you’ve created a set of dashboards that show you exactly what you wanted, why stop there? The world of marketing data is as exciting as the world of marketing itself, so digging deeper into the nitty-gritty details and finding nuggets of information is a logical next step in your journey to full analytical maturity.
Now, when you have all the data in one place, you can start developing new, original insights based on it, and the options are practically endless. Due to the vast amounts of this data, it would be very complicated to find these new insights on your own, but luckily, you can rely on the help of technology. Various techniques used by modern martech tools can be used to browse through your data and identify opportunities or threats to your marketing performance.
Scenario 1
Remember what we said before about the importance of data granularity? This is where it comes into play.
While understanding the overall performance of advertising campaigns launched on all channels is a key piece of information for the business decision-makers, marketers want to know more about the individual performance of each element of the campaign.
If you are a small company, handling only one ad channel with a fairly limited budget, you could get away with manual analysis of ad creative performance. It will take some time, but it will produce valid results and conclusions. But as the company grows, so does the complexity - the number of ads spread around multiple platforms can easily reach thousands, and there’s no way you can manually process or analyze data on this scale.
The digital team at Colgate-Palmolive in the EU faced the same problem. With members spread across over 20 countries and without any automated data collection, understanding the performance of individual ad creatives was more wishful thinking than a realistic expectation. Once they managed to centralize the data, the team was able to easily identify the better-performing ads.
This allowed them to structure future campaigns differently and optimize on-the-fly (without the 2-week delay they had previously) resulting in improved budget pacing and lower CPC values.
Scenario 2
Behind the term “proactive” lies the secret of many successful companies - they differentiate themselves by trying to identify and address detected issues before they become a problem, such as declining sales, increased competition, rising costs of delivery, or any other internal or external factor that can seriously damage their business.
This approach requires marketers to actively seek issues, trends, and anomalies in campaigns and other marketing activities. And again, if you run a small operation, this is manageable, but as the business grows, you have less and less time to look closely at the constantly growing amount of data.
This is where the technology kicks in. A machine can sift through data and notify you of any type of anomalies that have occurred previously, or that might be happening right now. Through utilizing AI technologies any type of negative trend or anomaly will be detected on time and you will be notified about it before it creates any major damage. And if the detected trend is positive, you’ll be in a position to turn it into a new business opportunity.
For many agencies, optimizing client campaigns on-the-fly presents a major challenge. Not only does it mandate a very advanced data approach to get the insights, but also requires a lot of resources to implement the findings. Unless they find a way to focus on the things that make the most effect. And this is exactly what the team at iProspect Canada did for their client, a global automotive group.
Through smart analyzing the performance of video ads of their client, the team at iProspect was able to quickly identify and stop the ads that had a significantly lower CTR, leading to savings for the client through lowered CPC values. The data has also shown which ads resonated better with the target audience, providing valuable feedback for the client and their future campaigns.
Even though the word “predictive” sounds more like witchcraft than business practice, with the introduction of AI-supported solutions, predicting the future is something that you can actually do with a decent level of accuracy.
When you talk to your colleagues from Sales, they sometimes sound like fortune-tellers with their sales forecasts and projections for the upcoming year. But if you start relying on the same techniques applied to marketing data, you as marketers can also become forward-looking visionaries, with projections firmly grounded in historical data and advanced predictive models.
A great example of this forward-thinking approach is the way the marketing analytics team of the popular US basketball team, the Philadelphia 76ers, uses weather data. Using historic venue attendance data combined with historic weather reports, they have created a predictive model that is able to foresee future game attendance based on weather forecasts for the dates of the games.
Not only is this important for capacity planning on the day of the game, but it also means they can intervene if the forecasted attendance is low by investing more in advertising before the game. Besides that, they can also provide additional intelligence to the food and beverage suppliers, making sure that there are no hot dog or beer shortages during the game.
Remember, analytical maturity is the ability of an organization to “integrate, manage, and leverage all relevant internal and external data sources into key decision points.” When you look at it, this is exactly what the team at the 76ers did - they used data from internal and external data sources to create a true decision support system for them and their key partners.
Armed with a basic knowledge of data management, marketers can employ some truly out-of-the-box thinking by combining totally disparate data types, leading to unexpected new insights.
Telecommunication companies are widely known for their advanced use of various datasets to improve customer experience. Through automating the acquisition of data into a single database, the digital marketing team at Vodafone Italy was able to link online prospect activities with lead, sales, and activation data handled offline by call centers using different CRMs.
Thanks to the cross-analysis of all of this centralized data they have managed to identify previously undiscovered opportunities for upselling existing contracts and reducing customer churn. On top of that, the intelligence from the performance of existing campaigns was integrated into future customer acquisition and retention campaigns, making the centralized data the gift that keeps on giving.
In the largely accelerated world of today, marketers often neglect the importance of banging their own drum when they’re successful. You might get discouraged by others taking the credit for your work or by a lack of understanding in the C-suite who don’t see the connection between your hard work and the company’s bank account statements.
To avoid this, you need to demonstrate your value. Advanced data science techniques applied to your data and displayed in custom charts and dashboards will help immensely in demonstrating the direct correlation between the thousands of dollars invested in marketing campaigns and actual revenue figures.
Agency marketers face the same challenges in this area as in-house marketers, as we can see in the example of Havas Middle East. Faced with requests from clients for improved reporting, this agency created an internal analytics framework. The “one tiny problem” was data integration, but once they had resolved it, the agency was able to present fully-automated, timely reports to clients.
Apart from the less tangible, but very important benefits of increasing trust and transparency in client relations, this approach brought much shorter turnaround times for campaign optimizations.
The improved reporting helped decrease the customer acquisition costs for one of Havas’ clients working in the very competitive mobile app market, where even the smallest optimizations can mean a lot.
As your role as a marketer progresses, so do your responsibilities. Once you reach the top and become a CMO, the burden of management becomes an overwhelming force, overshadowing everything else, including worrying about data. At the same time, your C-Level peers keep asking tougher and tougher questions, to which you will never have an answer if your marketing department’s analytical maturity is at a low level.
The ultimate question is very often “what is our ROI?” and not many CMOs know the answer to this question. Not because they’re not capable, far from it, but because it’s such a complex question. Establishing the effectiveness of marketing and its direct link to the success of the overall business can be extremely difficult, especially in the digital age, where you have numerous touchpoints that influence the buyer’s journey.
Getting to a dashboard that clearly shows the complete customer journey and marketing’s contribution to each of these stages is a process that might take years. Reaching this should remain a long-term goal of any organization, but that doesn’t mean that you can’t answer other tough questions in the meantime.
Let’s take the example of Return on Ad Spend or ROAS. It’s an important efficiency indicator of your campaigns and is calculated by dividing the total revenue generated by ads by the total costs of these advertisements. If you use just one platform for advertising and sell only on your website, it’s really simple to calculate it.
But if you use dozens of advertising platforms at the same time, like many companies do nowadays, the situation is far from simple and easy to manage. With the constant increase in the amount of competition online and growing ad costs, managing budgets effectively and finding out the right allocation among channels should be one of the key priorities for you as a marketer.
Once you have the aggregated ad spend from all platforms, and the total revenue from all sales, it’s easy enough to calculate it. The problem is that if you haven’t gone through all the previously described steps, then you won’t be in a position to do this at all.
Popular US-based automotive marketplace, Cars.com, had to go through a set of business transformations, process improvements, and technical innovations to reach this stage. But once they did, they were able to calculate the ROAS of all of their advertising campaigns. The company works in a very competitive market, as an intermediary between the car buyers and car sellers, so knowing how much they can invest in advertising and how much return they (or their partners) get from it are crucial pieces of information for the management of Cars.com.
And this is a great example of how marketers can clearly demonstrate the value of not only the investment in advertising but also in the underlying analytics infrastructure that allows them to be efficient in their job.
However, to enable the CMO and the rest of the marketing team to answer other tough questions, the organization’s analytical capabilities need to successfully shift to the final stage on the analytical maturity scale - prescriptive analytics.
The ultimate goal of predictive analytics is to eventually become so smart it becomes prescriptive, helping you to be more successful by advising on the next steps and actions that have the greatest chance of reaching your targets.
Based on the inputs that you give, the prescriptive analytics system will return a set of suggestions on what to do next to maximize the effects of your actions and investments. This concept can be nicely explained with a popular methodology called Marketing Mix Modeling, used to optimize budget allocation across multiple advertising channels.
In the previous steps of the analytical maturity journey, you’ve reached a point where a chart shows the current allocation of your advertising budget over multiple channels. It was a bit of a pain to get there since the chart combines data from multiple sources, but now it’s all set up you know exactly how much money you spend on each channel.
By applying some statistics to this data, enriched with conversion data, you can see how reallocating these budgets between channels would impact the overall performance. These statistical models can evaluate various scenarios and can tell you if a budget shift of X% from Facebook to Google Ads will bring more or less revenue.
The smart, prescriptive part comes when this system is supported by artificial intelligence, so it can analyze the correlations between all the provided data, ultimately delivering suggestions on which channel to invest in and by how much to get the maximum returns with the available budget.
Of course, the application of prescriptive analytics is not limited to advertising. Advanced AI techniques like machine learning can be used to help with pricing strategies, optimizing store inventory, and many other areas of marketing or business.
If you feel like all of this is overwhelming, and that the higher levels of analytical maturity are something you might never reach, know that you’re not alone. This is a massive challenge even for the most sophisticated companies out there.
A survey from 2018, six years after Gartner published the Analytics Ascendancy Model we mentioned above, has shown that only 13% of companies have reached the highest level of analytical maturity.
This means that the remaining 87% are still somewhere on the bumpy and winding road to analytical maturity, and if your company is among them, it’s nothing terrible.
Our final advice?
Start this journey sooner rather than later, and you will be able to enjoy all the benefits it will bring, one step at a time.
Want to improve your analytical maturity?
Book a demo with one of our advisors and see learn how Adverity can help your company move further up the analytical maturity scale.
