Here’s what marketers really want for Christmas: 3 tips for building a better approach to data
Getting the most out of your data is on every marketer's wish list this year. Contributor Stefan Benndorf discusses ways to get your analytics infrastructure in tip-top shape for 2018.
Be honest, what is it that you really want for Christmas? It’s a more effective marketing operation, right? One in which you can make quick, informed decisions based on data. Let this be your year.
Marketers are drowning in spreadsheets, especially marketers who are overseeing performance campaigns. Even those who have sophisticated tools are still spending an exorbitant amount of time with manual reporting processes because different systems create disparate data buckets that must be manually merged.
These inefficiencies are costing you in more ways than one. You are losing time — time that could be spent analyzing data instead of trying to clean it up — and you are making important budget allocation decisions based on suboptimal insights.
The holidays are hectic, but trust me, it’s worth stepping back to evaluate your data processes, and to make changes that will set you up for success in 2018.
Why data still causes so many headaches
Approximately 78 percent of marketers believe that the CMO should be the driver of a data-driven customer strategy, according to a report from the CMO Council, “Empowering the Data-Driven Customer Strategy.” Yet only 7 percent of marketers say they are able to execute data-driven engagements, and only 5 percent say they are able to determine the bottom-line impact those engagements have on their business.
This problem — and the frustration that goes with it — starts with an insufficient data foundation. Businesses need to improve the very architecture on which their strategies are built.
Most marketers use multiple systems to manage their operations, and each system generates its own datasets. Since the tools are not integrated, marketers run separate reports and then combine the data in Excel. That leaves them spending far too much time on admin-like tasks that should be delegated to a machine, such as manually inputting and cleaning up data.
How to improve your analytics infrastructure
If you can design a more intelligent approach to accessing and analyzing data, you will save time and get access to better-quality data, which, in turn, will improve your decision-making. With better data, you will have a more accurate understanding of what is working and what’s not, so you can then allocate your budget more effectively. These tips will help get you started:
- Take stock of your current approach
Begin by accessing your current processes and tools. What systems are you using? What reports do you run and how often? What metrics are you reviewing? Think about the key metrics you are trying to measure from the beginning and set up each of your marketing campaigns with those in mind.
When evaluating your different data tools, consider user-friendliness. All dashboards should be easy to use and manage. If a technical aspect of a tool is hindering usability, think long and hard about whether or not that is the right choice for your business.
- Consistency is key
Your goal is to create a system in which data is consistent so you can benchmark across different campaigns and channels. Whenever possible, set up campaign reporting so that all reports provide the same metrics, e.g., LTV, CPA. That way you can compare apples to apples.
- Group datasets into three categories
It’s useful to group datasets by purpose. This will help you get organized and determine the right KPIs to measure. There are three main analytics categories:
- Descriptive: The purpose of descriptive data is to understand patterns and make conclusions about what is and isn’t working. This data helps you determine things like your best-performing channel, your ROI per channel, your average cost per user and so on. A subset of this category is diagnostic analytics, in which you try to understand why something is occurring, e.g., why is this campaign performing better than others we have run in the past?
- Predictive: Predictive analytics is about using data to make educated predictions about future trends and patterns, e.g., what you expect your performance KPIs to be next quarter, how many sales you will generate, etc. These are the datasets that help with forecasting.
- Prescriptive: These analytics help you make smart changes to improve or fix your marketing efforts. For example, based on data, what can you do to increase ROI? In which channels should you invest more or less?
- Embrace APIs
Make use of application programming interfaces (APIs) whenever possible to automate the task of consolidating data into a central location. APIs make it simple to bring raw data from various sources into a single database. This rectifies the “patchwork” data landscape problem and saves you time by eliminating the step of manually inputting data.
- Keep things as simple as possible
“Analysis paralysis” is a real thing! Many marketers have access to the data they need but still struggle to put that data to work for them. You want to try to make your data-related processes as simple as possible. Use as few tools and dashboards as possible. Be precise about which KPIs you measure, and create decision funnels in which you define the metrics that matter.
It will take some upfront work, but it will be worth it in the end. With a smarter approach to data, you will improve processes, save time and gain access to higher-quality reporting. Then you can make better decisions about how to allocate your budget, driving increased efficiencies and cost savings. How’s that for a marketer’s holiday wish come true?
Opinions expressed in this article are those of the guest author and not necessarily MarTech Today. Staff authors are listed here.
About The Author
What is machine learning? - MarTech Landscape