Learn how to set up advertising analytics and get the most value out of your marketing campaigns.
The trouble with marketing initiatives is that it is almost impossible to tell how they impacted the business’s bottom line.
As the marketing pioneer John Wanamaker said:
“Half the money I spend on advertising is wasted; the trouble is I don't know which half.”
A person scrolling through Twitter on their mobile app might have seen your ad, loved your brand, and then logged into their desktop to purchase your product. The gap between needs generated by marketing spans across marketing channels and time.
So it is hard to look back at all the marketing efforts and determine which campaigns and ads are worth repeating.
Luckily, digital marketing leaves a trail of breadcrumbs we can follow. By recording all clicks, ad impressions, and interactions, modern advertising produces a myriad of data that can be used in marketing analytics to gain actionable insights that can help your marketing team shine.
In this guide, we will look at the 8 steps to build your advertising analytics.
The role of marketing is not to post pictures on social media platforms, but to sell direct to consumers and/or generate leads that are turned into customers down the sales pipeline.
Newsletters, social media posts on Instagram, PPC display ads, SEO, and other marketing activities are just means for achieving direct sales or lead generations.
To pick the right activities, you need answers to your most pressing business questions, such as:
The specific questions will depend on the role your marketing team plays in-house and the marketing stack you work with. But starting with business questions that can change your marketing strategy - and therefore activities - is the essential first step.
Go ahead, make a list of the questions which you need to answer the most.
Business questions are seldom phrased with the same words as data engineering answers.
There needs to be a translation step between what is asked from a business perspective and what is answered with marketing data. To effectively translate business questions into data questions, you need to go through 3 steps.
STEP 1: Phrase the business question (in detail)
We talked about this above. Come up with a list of questions that need answering. The name of the game is to be as specific as possible. For example: “How many new customers we acquired via digital” is less clear than “How many new customers we acquired via all digital advertising platforms (concretely, Facebook, Google, and Linkedin)”.
STEP 2: Determine what metrics would provide suitable answers
Think of what metrics would help you answer the question. Envision how the answer would look like. In the example above, the metric would be the “number of new customers that were acquired across digital advertising platforms per week”. If you wanted to measure the cost-effectiveness of marketing, you would think about ROAS. If you measured engagement by ad placement, you would measure ad impressions or clicks.
Looking for inspiration? Check our guide to eCommerce analytics metrics.
STEP 3: Determine what data is needed to construct the metric
Marketing data by itself is seldom sufficient to answer questions. In the example above, you would need both Facebook, Google, and Linkedin ads data to determine which new customer interacted with your ads before becoming a customer; as well as information on the exact date when a customer made their first purchase to count them in the right week on the metric. That information is often stored elsewhere, such as a CRM, ERP, or your e-commerce transaction logs.
At the end of step two you should have three things:
Before you start extracting data from your advertising data sources, make sure you are collecting all the data you need in the first place.
Advertising platforms and ad networks (Facebook Ads, Google Ads or AdWords, Linkedin Ads, …) do this automatically for you. Data collection from advertising channels (usually) happens without additional hustle.
Google Analytics also collects data for you automatically. Just make sure to:
For other digital channels (newsletters, 3rd party competitions, etc.) that do not collect data automatically, use UTM tags to track your campaigns.
Extract the raw data from your data sources.
You have three options for data extraction:
Once you have extracted the raw data, pass it to the next step in building your advertising analytics.
Before you can use the raw data to get insights, you have to clean it or transform it.
For some analyses, you will aggregate data. For example, sum the total revenue per customer to build up Customer Lifetime Value (LTV) metrics.
For other metrics, you will have to join data. For instance, to get Customer Acquisition Cost (CAC), you would join all the advertising platforms’ data into a single table.
Save the cleaned data into a repository of your choice.
Relational databases or data warehouses are your best bet. Their advantage? They can be connected to other tools that help you visualize and analyze the data, can be accessed by multiple users simultaneously, and scale well as your business grows.
Connect analytic tools with your saved analytics data to put it to good use. Calculate statistics, metrics, and KPIs, visualize trends and dig deeper into interesting patterns to discover new opportunities.
In this step, you build the answers to the questions you asked at the beginning.
Technology will never substitute human creativity and ingenuity. But it can surely replace manual repetitive work.
Your personal touch will make a world of difference only in steps 1, 2, and 7. Where the heavy brain lifting happens, creative juices flow, and there needs to be a human in the loop to set up advertising analytics and interpret its results.
Every other step in between is a candidate for automation with integration tools. Integration and automation tools can help you set and forget the repetitive data extraction, cleaning, saving, and moving around.
Keboola is the end-to-end data pipeline that helps people automate their data operations.
Rely on Keboola to:
a) Automatically extract data from over a hundred data sources:
b) Clean and aggregate your data with SQL or Python code. Set up the transformation recipe once and put it on autopilot to repeat every time you extract the data.
c) Save your data to the destination of your choice. Be it a database, data warehouse, or Google Sheets.
d) Connect your data via Keboola with BI tools that help you analyze it and extract advertising analytics insights.
Do all of the above with a couple of clicks. Instead of spending time building your advertising analytics manually, or waiting for your engineers to be free, take the reins in your hands and build it yourself.
Try it out. Keboola offers a no-questions-asked, always-free tier, so you can play around and build your advertising analytics pipelines with a couple of clicks.
