What does a modern marketing reporting stack look like?

Lessons I learnt from working for startups with strong analytics needs and various data sources.

The main issue around marketing reportings

This is 2019, and sometimes reporting looks like it has been stuck in the 2000s.

Today more than ever, marketers need to have an instant access to numbers and being able to build their very own reports, to monitor those numbers on a daily basis.

They have never been given so much data, but they are struggling to be able to gather it and actually get actionable insights.

Indeed, it has now become standard to have your data scattered all over your marketing stack: leads in Hubspot, clients in Salesforce, acquisition spendings both in Facebook Ads and Google Ads, payments in Stripe…

And the marketing team does not have anymore the luxury of time to wait for the IT to build dashboards they will actually never use. Either because they do not have the right information in it, or the data lacks of actionnable insights.

Being able to think and make the reports autonomously is key for the marketing teams, and the time you got weekly Excel reports for each data source should be over.

Excel no!

Copying and pasting data from several Excel files to get the big picture is a tremendous loss of time, not to mention the master file size issues and dealing with the history.

It cant take up to one day a week to a marketing manager to do this data gathering job before even trying to get insights from it.

And we have not even talked yet about the integrity of the data and its format. One of the major issues in setting up a data pipeline is to ensure the quality of the data itself. If the data gets corrupted on the transport chain, the analysis will be biased and sources of errors.

Who has never had issues with weird encoding or missing columns when you paste data from a file to another?

But I have some good news: this is 2019, and there is actually a way of solving this issue.

Building a modern marketing reporting stack

A plea for a marketing database

I strongly advocated in favor of setting up a Customer Data Platform (CDP) in my previous article. A CDP is a game-changing tool in the hands of marketing teams, because it enables them to have a clear vision on each customer journey in real time.

But this kind of databases is a really advanced use case for mature companies. Actually, there is a first step on the road to becoming a modern marketing team: the setup of a marketing database.

By marketing database, I hereby mean a unified source of truth for your marketing team, gathering data from your acquisition campaigns, leads, sales and clients.

Gathering this data in one single place thus enables your marketing team to plug a BI tool on it and get a clear overview on its actions and their consequences in terms of business.

NB : Business Intelligence (BI) and/or Data Visualisation (Dataviz) tools such as Google Data Studio, Tableau or QlikView enable you to build reports and get value from your data.

How to setup your marketing database

Whereas this step looks like being the most difficult, there are actually a lot of very good tools to achieve this in a few minutes and this is pretty straight forward.

There are two distinct tools you need to use to get your marketing data into a database:

  • The database itself.
  • The tool which will managed the data pipelines to your database for you: an ELT software (Extract, Load and Transform data).

The main advantage of this architecture (ELT + PostgreSQL DB) lies in its ability to answer one of today’s most struggling issues: scalability. Indeed, this approach lets you add data sources very easily without having to think your entire data architecture again (and thus your reporting stack, too).

Setup a database in two minutes

Since you want to setup a database for your marketing team, I strongly advise you to use a managed database solution, so that they do not have to care about maintenance and backups. AWS RDS and DigitalOcean both have really good offers for this, starting ~ $15 a month.

I strongly advise you to choose a PostgreSQL database since this is a very popular open source database which is widely supported by ELT softwares.

NB: here is my DigitalOcean referral link if you want to get a free $100 credit to try it out.

Move your data to your database in five

Once your database is setup, you now have to choose the pipelines for your data. Segment or Stitchdata (Talend) are very popular solutions for this.

I strongly advise you to start with Stitch, which has a free plan until 5M records a month (whereas Segment is a more expensive solutions for mature companies).

Stich data sources
Stitch data sources

Stitch offers a large variety of connectors to plug your favorite data sources to the destination of your choice (the PostgreSQL database you just set up in your case). The user interface is pretty intuitive and straight forward.

All what you have to do is select your source, choose which data you want to replicate to your DB, and the frequency of replication.

And, that’s it! A few minutes later, all the data from your different data sources are now synced to your database.

Setting up a modern acquisition reporting with Data Studio

The tool itself

Data Studio yes!

Data Studio is a reporting tool from the Google Marketing Suite.

It has several huge advantages:

  • It is free.
  • It is very intuitive for non-technical people.
  • It offers several native connectors to plug your data sources and start building reports, including Google Analytics, Google Ads, Google Sheets, BigQuery… and PostgreSQL.
  • You can easily join data from different sources to build actionable dashboards.
  • Its dashboards are refreshed live.

If you want to see what Data Studio reportings look like, here is a summary of 30 templates for your reports (in French, by Cartelis).

NB : Data Studio has some limitations regarding the joins and the data export, but this is something we can talk about in another article.

The final step: plug Data Studio to your database

Pluging Google Data Studio to your marketing database is really easy since Data Studio offers a native connector for this kind of databases.

Juste click on “Resources > Add a new data source > PostgreSQL”… and that’s it! You are now able to query all your marketing data gathered in your very own database, and build real-time updated reports.

The final stack
The marketing reporting stack

Coming soon: OpenClassrooms case study

A case study case will follow on how we helped OpenClassrooms to setup their marketing reporting stack after they raised 60 millions euro. Keep updated!

OpenClassrooms is a French online education platform for vocational training, providing courses in IT, technology, entrepreneurship, and digital skills. Courses are conducted fully online, through a mix of video resources, online reading, real-life projects and individual mentoring sessions.