Marketing Data – Aggregation & Segmentation
Define, Operationalise and Automate.
Understanding marketing data is a tricky task. I could write a 5-book series about it and it would still not be enough. Plus, no marketer would read it. Data – boring stuff, right?
Nowadays marketers have too much information and consuming it is difficult. Just think about easy email campaigns that use Marketing Automation tools. We know who we are targeting, we know who opened or clicked a URL in the email, we know who visited a web page and if they bought a product or service we were selling. We can track every step that customers make. But, so what? What do we do with that information?
How do you use marketing data?
The obvious thing would be to use the customer’s activity as a trigger for follow up communications, or next steps, as a part of a journey. Mr X received an email and opened it, so we can send him a second email that continues on from the initial communication.
So we send him another campaign, and another one, and we gather even more data. But do we take any lessons from the stored rows of information we now have in the database?
Do you know which customers are the most loyal ones?
Do you know who is reading your emails most often? Who is really interested in what you write?
Do you know that Mr X visited your website for the 5th time this week? Maybe he’s looking for something? Do you know what he was searching for?
We cannot expect marketers to be data scientists, fluent in data mining. However, the role of a Marketing Data Analyst or Marketing Manager includes designing a data flow so that marketers have all the vital information available, in a very clear way. And I will tell you how to do just that.
1. Define what you need
At the global level – what would you like to know about your customers?
Let’s take as an example an engagement measure. ‘Who is reading our emails?’.
Start here and ask yourself what you would do with this information? How would you interpret it? You can for example use this information to create customer segments and use them as additional targeting and reporting dimensions. Wouldn’t it be interesting to know that only 20% of your recipients are opening and reading emails? And who they are? Or to know the 35% that never opens anything? Customer segments can provide this insight.
As a marketer you defining what you need is important. Let’s say your need is to divide your customers into five groups:
- Not interested
- Slightly interested
- Sometimes interested
- Rather interested
- Waiting for your every email
2. Operationalisation
This is the tricky part of the process and it’s important to make sure it’s a team activity. You need a marketer and data analyst in the same room. Ideally you’ll have a Martecheter in the mix too. It can be the same person, but then there’s no fun team bonding, probably no meetings and no cookies and coffee. Well, we are all working from home anyway, so forget about it. The important thing is that this part of the process involves the marketing and the analyst minds coming together.
Back to your five segments – would you rather have them equally distributed in quintiles of activity? (Division Type A) Or have strict rules and see how they split between them? (Division Type B) And more interestingly – how do they flow between the groups?
And now we are talking about data.
In Option A you could define the last six months as a time frame. Search for every open event activity and every sent email for each customer. Calculate how many emails they received and how many of them they opened. Sort all customers based on that percentage and divide them in the five different groups based on that value.
In Option B you are deciding and defining activity tiers – like in my simplified example. Then for each customer consider their last four mailings and verify if they were opened or not; then calculate the value for each customer. Easy!
3. Verify your data
Don’t simply trust your data, question everything and make sure you understand what’s there. It’s important to fully understand the process and data before you draw conclusions based on one excel chart.
It’s full of traps but you can do it.
Think back to your initial needs. Did you mean only marketing emails or operational emails too? What if a customer opens the same email twice? Take the Data Analyst on this journey with you. Be as precise as possible. Question the final value and repeat until you are sure you have what you want.
4. Automation
That was easy! Now is even easier.
Speak with your DWH (data warehouse) team about maintaining the same process on a daily basis, preferably at the data servers. That’s the optimal solution.
They should keep all the calculations and run them daily/hourly to supply your marketing team with the freshest possible data. On a daily basis your marketer shouldn’t have to worry about calculating all of this from scratch. The final and up to date information should be provided for them in the marketing data mart and become one of many dimensions describing customers. Pretty useful!
5. Don’t limit yourself
All the above was just an easy example. I’m sure you can make a long list of dimensions you would be interested in. But the process is similar – define, operationalise and automate. Raw data is beautiful for data freaks like me, but painful for daily usage. That’s why reasonable aggregation and segmentation helps to get the best out of your marketing database.
To find out more about aggregation & segmentation, read the section devoted to this topic in Purple Square’s eBook, An Introduction to Successful Marketing Automation.
If you would like to discuss further how to manage your data, please do get in touch to speak to one of our experts.
Like what you see?
Subscribe to our newsletter for customer experience thought leadership and marketing tips and tricks.