C-suite support for data initiatives is changing. Those at the helm are seeing the advantages of shifting from a reliance on data as a tactical tool to a strategically leverageable source of rich insights. There is a new willingness to invest in the people, process, and tools related to accurately capturing marketing data and making business decisions based on it. As evidence of this, database health KPIs are now showing up on more CMO scorecards. There are also new skillsets joining the marketing ranks with titles like Head of Martech, Data Scientist, and Data Steward. And larger organizations are building entire data science teams with internal or external resources, with some companies creating data councils that include cross-functional members representing IT, sales operations, and marketing operations—to make sure everyone has a voice.
Stat from leading research and advisory company—
"Marketing and customer analytics the #1 marketing investment priority over the next 18 months."
No matter the level of martech or in-house data expertise, what’s missing from most organizations is the ability to apply best practice analyses to fragmented marketing data—there isn’t a unified data layer. This unified data layer makes it easier to get the most out of applying artificial intelligence and machine learning, it can unlock the ability to do next-gen cross-channel optimization, predict future campaign performance against business targets, guide budget allocation decisions, and drastically improve ROI for companies.
Watch the 'Data, Data Everywhere -- But Where are the Campaign Insights?' recorded webinar, to hear conDati Senior Director of Revenue Marketing, Kelly McKeown and Search Discovery’s Senior Director, John Lovett, discuss how to use AI and ML to drive pipeline and revenue growth.
A Unified Model for Better Data, Better Analysis, and Better Decisions
It starts with data quality. Better data leads to better analyses, which leads to better decisions—all helping marketers convert and win more. Let’s break it down.
Step one is to implement a data governance strategy by putting in place standards for collecting, accessing, and maintaining data and the way it flows through your systems. Next, the goal is to collect and blend data into a unified data asset, including audience, campaign, and content information. By blending it from every important cross-channel marketing data source, you’ll have access to a complete view of the buyer journey from initial interest through to revenue.
Now that you’ve got better data—you can build relevant and actionable reports and dashboards to analyze that data. Here are a few things to keep in mind as you are digging in:
- Align to a business strategy, tying campaign performance to a business outcome so it is more meaningful especially for executive meetings such as quarterly business reviews.
- Deep dive into a single channel for testing, comparing, and optimizing specific metrics and run cross channel analytics, as well to understand how channels relate to each other and determine which channel combinations result in the highest return. With this insight, you can invest more in channels that when combined effectively have a positive impact.
- Keep the end user in mind, by providing helpful filters or views for easy slicing and dicing of results by various roles and other criteria such as region, product, line of business or segment.
- Refresh data often in dashboards and reports, daily or real-time versus weekly. Timeliness of data allows marketers to pivot quickly to prevent potential revenue loss or take advantage of opportunities that present themselves.
- Make sure reports and dashboards are easy to understand, especially when sharing without narrative alongside them. Use high level visuals with granular drill down access so it is easy to drill into the “why” of trends or anomalies.
Lastly, the insights that come from data quality, and thoughtful analyses impact your decision quality. This is where the data science magic comes into play, giving you even more confidence when making important decisions. Layering AI and machine learning power into your analyses helps you quickly and accurately understand past, current, and future performance so you have information at your fingertips that you can trust when making decisions like keeping or killing programs or reallocating spend for the highest possible returns.
"With Intelligent recommendations, powered by AI and ML, you can see beyond whether performance is good or bad, you are provided a direction to take that’s backed by data science.”
Here are three data science powered use cases you may want to take a closer look at:
Campaign optimization leveraging machine learning scans hundreds of variables and combinations such as day of week, time of day, region, campaign creative, and demographic information and provides intelligent recommendations that can quickly and easily be activated - shortening optimization cycles from weeks to days. Campaign optimization now goes beyond looking at high-level reports and deciding whether to turn on or off a campaign, it can look for performance lift at granular incremental levels and know at what point you will hit diminishing returns and are wasting spend.
Performance forecasting provides dynamic forward-looking insights to help you understand whether you are on track to hit targets before it’s too late to do something about it - alerting you to anomalies, good or bad, that fall outside of your range of normal. Leveraging machine learning forecasting becomes more accurate and gives you the ability to build in the seasonality of your business that is often done poorly when data science is not applied.
Customer journey analytics using statistical models more accurately calculate contribution and influence helping marketers understand the most valuable touch points, determine the next best action for engagement, and prioritize spend allocations. This approach statistically calculates each channel’s influence on the desired goal - most often revenue - rather than assigning specific credit to touch points as seen in first, last or evenly weighted attribution models which provide a biased view.
If you’re at the stage where you want to explore this for your company, but the idea of building out a data science team internally seems overwhelming, take advantage of conDati's marketing data science as a service platform. Try us on for size. Get insights and optimization recommendations from your combined Google Analytics, Google Ads, and Facebook data within 7 days of providing conDati access, when you sign up for our free assessment.
In previous blogs, I detailed the new pressures and accountabilities CMOs and digital marketing leaders face today, which is daunting. We also looked at how to identify where you are on the digital marketing maturity path to get where you need to go, and reviewed how an AI model can give you the full hindsight, insight and foresight view into your marketing analytics data to confidently deliver on key business goals.