If you intend to implement a Marketing Automation system, you need to know what data to collect and what to do with it. Paradoxically, having too much data is of no benefit, because it's difficult to draw conclusions from it and, in addition, you incur unnecessary costs for system licenses, the price of which depends precisely on the amount of processed data.
From this article you will learn how marketing automation systems facilitate the acquisition and processing of valuable data, and what data is worth acquiring and processing in marketing automation campaigns.
Why collect data?
Let's start with the simplest question - what do you need data for? Why do you collect them? The answer - "because they are" is not scored. Data is like fresh flowers - they cost a lot, they spoil quickly, and if kept, they cause a lot of trouble. That's why it's not worth collecting them in stock. It costs a lot to keep your data up to date, and using outdated data is worse than having no data at all, because it gives you the illusion that everything is fine. And when a campaign doesn't work, it's hard to find the source of the problem, because the timeliness of the data is the last element we verify when looking for the causes of failed marketing and sales efforts. So before you start collecting data, make sure you know what to do with it. If you are not sure whether you will need the information, you will probably never use it.
Data review
Implementing new systems such as Marketing Automation or CRM is a good opportunity to look at the data you have and decide whether it's worth moving or integrating it into the new system. Pay particular attention to the following:
- Date of last update - if, for example, the last update of contact information in a record was made more than 2 years ago, there is a good chance that the person has changed position, phone number, or employer.
- Legal status of the data – are we sure that the data was obtained in accordance with current law? Do we have the right to process the data? Are we sure that the holder of the data has not made a request for its deletion?
- Consents for communications – if we are entering data into a marketing automation system, it is likely that we will use it for communications. Make sure you have this option. Hint - the data in the CRM system is usually entered by salespeople on the basis of data they have acquired during meetings, conferences or phone calls. They can contact such people, but without additional permission you can't send such people bulk communications - e.g. from email or marketing automation systems.
- Completeness of data in the database – if, for example, in your entire database information on dietary preferences contains only a few percent of records - consider whether to enter this data into the system at all. Their use on a massive scale will be unprofitable and supplementation will cost a lot.
- Data format – here it is worth working with the IT department or a database specialist. Sometimes before importing data or merging databases, it is worth making sure that the data is recorded in a similar way. E.g., the postal code is often in the form without xx-xxx or simply xxxxx. If this is overlooked, you risk inconsistencies in the database and, for example, serious errors in segmentation or reporting.
This stage usually doesn't cause much enthusiasm among marketers, but if you want to generate quality leads, you need to ensure high quality data.
Remember, too, that this is a preliminary stage - it only ensures that the data on the day it is loaded into the system is of the right quality. And after all, marketing automation systems "feed" on data, i.e. they facilitate its acquisition, processing and use in communication - for example, for segmentation or personalization of content. This means that you also need to ensure that the quality of the data does not drop during processing.
How do you maintain data quality in marketing systems?
I've put together some good practices for using data in marketing automation systems that work well no matter what system you choose.
- Reduce the number of forms – forms on pages are the most popular method of acquiring data, so-called early leads. Data for an e-book, newsletter signup, webinar registration - all of these are handled by forms. I know from practice that it's a good idea to have as few forms as possible, because then we reduce the risk of error. It once happened that, by mistake, when creating a new form, we mapped the fields incorrectly and the town information went into the "Name" column. It's not hard to imagine the mess this caused. From then on, I was very strict about the quality and consistency of the forms.
- Limit the amount of data you collect. . Forms with five fields are already at the limit for our clients, so it doesn't make sense to risk losing a potential lead because we asked them for too much data. Another consideration is the efficiency of such a database - the less data, the faster the database will run. If you set a standard for data collection, it will be easier for you to reduce the number of forms. For the sake of argument, I will state that in marketing automation systems, one form can be used multiple times on multiple sites.
- Don't force customers to provide data. If you force, for example, a phone number on a person who simply wants to download your PDF, you will get a response like +11 111 11 11 11 11. What will you do with such information? Especially in the first stages of contact with your potential customer, keep the amount of data to an absolute minimum, such as a first name (to greet the recipient) and an email address. If you overdo it, your base will be flooded with creative, vulgar or random strings of characters.
- Don't collect data too early in the customer journey. The so-called top of the funnel are the leads least likely to buy. By converting them to qualified leads, you reduce their number - thus narrowing the funnel. So it doesn't make sense to collect data from everyone who comes to your site. Most people who visit it will not become your customers. Even if they agree to sign up for your newsletter or provide an email address in exchange for an e-book, you don't have enough knowledge of their needs at this stage to collect their data. Fortunately, marketing automation systems make it easy to determine when it makes sense to ask a visitor for data. Maybe after the third visit to our site in a week? Maybe after a redirect to our site from a precisely targeted campaign? You have a lot of options for determining the right moment, but from my practice, the first site visit is not the moment when it's worth collecting data. Again, I'll come back to the financial issue - in most marketing automation systems, the price depends on the number of records. So if you collect a large number of low-quality records (e.g., people who have downloaded your e-book, but are not in your target audience) you will incur a large licensing cost with virtually no benefit.
- Plan to update your data. People change positions, jobs, addresses. Companies go bankrupt, change names or merge with others. That's why it's a good idea to plan in advance how and how often to update your data. Sometimes it's a good idea to use data from a CRM or customer service system. If someone is updating data in these systems, it's probably worth updating it in the marketing system as well. It's also worth considering social media platforms - such as LinkedIn, where users report changes in their employment status. My practice indicates that such an operation is worth doing at least once a year.
- Centralize enrollment and unsubscription from address databases. RODO has done a lot for data protection awareness. Customers are more likely to unsubscribe from mailing lists or newsletters and become irritated if a company continues to send them messages despite this. In addition to annoyance, you may also face legal consequences, so make sure your customers are properly handled in this matter. Fortunately, good marketing automation systems allow you to centrally manage signups and unsubscribes from newsletters or mailing databases. Once the appropriate policies are set and saved in such systems, a customer who has made such a request is automatically removed from the legally required communication systems, so mistakenly sending a message to such a customer is virtually impossible.
- Automate error detection and removal. Many errors in databases come from the owners of the data themselves. Typos, incorrect abbreviations (Sp. z o.o. or Sp z oo), Abbreviation of place names make our databases constantly vulnerable to such errors. Good marketing automation systems have modules that catch and fix some of such errors. Therefore, if you intend to process tens or hundreds of thousands of records - make sure your marketing automation system has such features. Otherwise you are condemning yourself to costly and lengthy data cleaning or full database littering.
What kind of data is worth collecting?
So far we have focused on the formal aspect of data processing. Now let's consider what data is worth collecting in marketing automation systems. Here it is worth introducing the concept of Digital Body Language.
In traditional communication, we convey much more content non-verbally. By analyzing body language, we can get much more information than what we hear from the interlocutor. Also in digital communication, analyzing the "non-verbal" behavior of potential customers can greatly expand our knowledge of their needs, preferences or how they make purchasing decisions.
What is digital body language?
Digital body language is a set of behaviors of users of the Internet or other digital resources, which, thanks to analytical and marketing tools, we can observe and adjust the way we communicate with them.
It is for the needs of business marketers that the most complete concept of digital body language has been designed. It was described in the book "Digital Body Language" by Steven Woods - founder of Eloqua (now part of Oracle Marketing Cloud) - one of the first and most advanced tools of the Marketing Automation class.
According to this concept, the full profile of a potential customer consists of data that he gives us directly - "verbal" - such as filling out a form, and "non-verbal" - resulting from observations of his behavior.
Data from the form will tell us who the recipient of our content is, observation of his behavior will tell us what he is interested in and how interested he is in our offer.
What can we read from behavioral observation?
It turns out that a great deal. The easiest way is to count the number of sites that a visitor to our domain has viewed. If we add to this the frequency of visits, analysis of "events" such as clicks from Call to Action, form fills, etc., we have an interesting picture of interest in our offer. It is also worth noting whether the lead consumes differentiated content. E.g., a person who visited our website three times will be a less interesting lead than a person who downloaded an e-book and attended a webinar during that time. So we can observe not only the quantity, but also the quality of interactions. The more of these qualitative interactions (e.g., webinar, e-book, interactive course), the higher the weight of the lead.
To evaluate verbal and non-verbal data, there are lead scoring matrices available in better marketing automation systems.
In the graphic below, the horizontal axis determines the weight of the lead based on the post, while the vertical axis determines the level of engagement based on behavior and content consumption. According to this example, we will give the highest value to directors and board members who have attended a webinar or started testing software.
Summary
As you can see, it's worth considering what we will use the data for before we start mass data acquisition and processing. Data, like fresh flowers, gets old quickly and in this form it is no longer needed by anyone.
It's also worth taking the opportunity of implementing a marketing automation or CRM system to organize the data and describe the processes for acquiring and processing it.
Once the data is organized, you also need to make sure that the quality of the data does not decline in the course of its use. I hope you will benefit from some of the best practices I have shared. It's always a good idea to seek the advice of a database specialist, as you won't be without specialized knowledge for larger databases.
Finally, remember that good marketing automation systems allow you to collect more than just data provided directly - for example, by filling out a form. The best results come from supplementing them with data on the behavior of the lead - for example, the amount and quality of content he consumes on your site. It is the ability to analyze Digital Body Language that makes marketing automation systems so effective in generating quality leads.