top of page
Writer's pictureJouni Leskinen

How CMO can reduce 99% and marketing data analyst 100% time spent on analytics and reporting

Updated: Sep 13, 2021


Back in the days when I was working as a consultant and we did a lot of BI dashboard projects and quite often a project seemed like it will never end. I realized that there is this infinite loop between business person and bi expert when building dashboards, that slowly consumes a huge amount of time and it’s because one person knows the business inside and out and the other knows technology inside and out. This is more true in marketing than anywhere else. Marketing data is complex and marketing is a skill that rarely exists in the same person who knows BI tools very well.


Think about the following workflow:

Task

Duration

CMO finds a change in conversion rate which she cannot explain

CMO emails to analyst

10 min.

Analyst emails back to CMO

10 min.

CMO replies

5 min.

Analyst confirms

5 min.

CMO confirms

5 min.

Analyst finds an answer form the data and emails the finding as an excel file attached

20 min.

CMO replies after studying the provided data

15 min.

CMO and analyst agree on a new column to be added to the drill down on the business dashboard

15 min.

Analyst makes the change

10 min.

Analyst tests the dashboard

10 min.

Analyst publishes the new dashboard to the cloud

5 min.

CMO checks the change and provides feedback

15 min.

Total time consumed in this loop

1 h 55 min.


The other way


CMO finds a change in conversion rate which she cannot explain. Drills down anywhere and all the way to the finest detail in the data and runs AI analysis on the data. She doesn’t need anybody else to help with access to data and finds insights quickly. Total time 2 min.


If we calculate how much less time the CMO spent in the latter way 1-(2/115)*100=99%

Because CMO was able to find the answers on her own, the time saved by the analyst is 100%

because explaining the data and creating additional reports as answers to questions is no longer needed. In reality this kind of costly loop happens multiple times a month, a week, a day, not to mention the real cost of frequent interruptions and context switching.




Our own problem, there must be a better way


About a year ago we where really frustrated with BI tools we were using to build dashboards and provide access to our data platforms data for marketers. A lot of time was used on explaining the technical details about data extracted from marketing and sales applications and how the data is arranged for analytics. Making changes to the dashboards was extremely slow and we were suffocating to the questions about data and metrics. Most BI tools are just way too technical for business users to be used efficiently and it slows down the decision making.


We started to rethink the whole process, how to enable business users to find their own insights and build reports fast without technical details and without limiting access to the data. Since the beginning our mission has been to fully automate analytics, let people focus on insight and making better decisions to improve their results, without having to work data in spreadsheets and wait for custom reports build by somebody else.


Now after years of hard work we are there, data processing is automated and Madtrix is powered by ThoughtSpot search and AI driven analytics which was the last mile of getting analytics fully automated for the business users. No more manual data preparation, data is rich and report building doesn't require help from experts. Together with Madtrix data platform business users can now search all of their marketing and sales data across the customer journey really fast assisted by AI . Analytics powered by ThoughtSpot really is like Google for data or Hotels app for hotels, you can actually find answers to your data questions quickly.


3 pieces of puzzle you have to solve to build automated analytics that can measure activities, efficiency and results.



Automate data collection

Typical marketing and sales stack consists of tens or even more than a hundred tools. The goal of analytics is to be able to get a picture about what’s happening in the data trail the customers leave into these tools and turn that data into information and accurate decisions that drive the results we want to achieve. The first step is to centralize the data and make the processing systematic and automated. When data is processed on a user level it usually doesn’t fit into excel and it is very difficult to centralize and normalize by transforming and joining data sources using spreadsheets or csv-files. This step requires some data engineering and proper tools.


In Madtrix this is done with managed data pipelines which only requires authentication from the user/admin of the marketing and sales tools and the data processing after that is automatic.


Build a data model

Data modeling is important! It’s the part of the process that brings your data to life. If you just push data into a database and call it the data warehouse, the data is still in siloes, but at least its in one place. The magic is making a record where data from social and ad platforms, website and sales, CRM or ecommerce platform’s talk to each other and the flow of customers can be analyzed and measured across their journey. Sales and marketing belong to the same funnel, salespeople don't want leads they want opportunities, and if marketing analytics and data ends to the leads its impossible to know which marketing activities help turn leads into opportunities.


Marketing data is tricky, but it should be modeled in a way that changes in dimensions or data model doesn't mean coding, data engineering and report rebuilding every time. Otherwise it won’t keep up with your business and you are pretty soon behind, chasing your data and leaving money on the table because your data is missing the needed context. Its also important to be able to collect the information which is in the people’s heads and not available in the API’s of the tools.


In Madtrix we are doing this with integrated data model and a bunch of management features where business users can feed in metadata about their marketing without anyone having to code or data engineer changes that happen in the business.


Provide access to all data and empower business people to make their own insights

People don’t need tools they need information and insights. If you invest in operational tools and data collection it makes all the sense to also be able to access and draw insights from all that data. Imagine if you could 1000 x your operational efficiency and allocate the saved budget efficiently into areas like media investments or content creation, how would that impact to the business. Modern marketing technology stack is complex and the environment where marketing is done is constantly changing, like we saw in the recent IOS 14 change by Apple. Being able to rapidly react to changes and get insights fast is vital to stay competitive in the future. Artificial intelligence can find hidden patterns from data faster than humans can. AI is here to make people more valuable not to replace them.


Madtrix has ThoughtSpot’s powerful search and AI driven analytics for insights. Its also self learning platform which uses AI and machine learning to learn about users and provides better and better answers based on the feedback.


How to move forward

To build analytics that drives results, you shouldn't start from the question what data do we have, but instead think about what is our strategy, what are the activities we do to execute the strategy and how do we measure the effectiveness and results or those activities. Then map out the customers journey and the activities together with the applications you use to deliver touchpoints to the people who you can help with their problem.


Then get data together from the applications and find any gaps that need to be filled and start improving your data quality, this includes tasks like adding context to your data, improving data modeling and configuration of marketing applications like ad platform tracking and web analytics. Most companies end the development of analytics when they get some isolated metrics on the dashboard, but this is not end this is where the journey to better data, better analytics and better decisions begins.


If you need help moving forward, subscribe to this blog or book a call with me and lets see what we can do together.


Comments


bottom of page