Your CMO wants a “simple” report showing how many opportunities marketing sources every month and how many bookings or ARR they generate. The problem is that no one has set up the tracking and cataloging mechanisms to connect the dots between what marketing publishes into the ether and how that correlates to sales.
Marketers are in the hot seat (now more than ever!) to validate the impact of their strategies. The truth is that many marketers are not aware of the complexity involved in tracking every interaction and linking it to a person, account, or opportunity.
We’ll start by defining marketing data collection, give you a roadmap, and outline some best practices to live by when building a data collection strategy.
What is marketing data collection?
Marketing activity produces myriads of data across online and offline sources. A marketing data collection is a centralized repository for your marketing activity organized by person, account, and opportunity records. This helps marketers create a timeline of critical interactions and correlate marketing activities to revenue.
Marketing data collection requires what CaliberMind calls an “identity graph” (also known as a “golden record” system or “unified customer data”) to help tie records together across systems because systems identify people and companies differently. These graphs take the unique identifiers different systems use to label people and merge them into a single-person record or a “golden record” to tie all activities together.
For example, GA4 can look at interactions by a User ID, a Device ID (which can track people across different browsers and applications), an IP Address, or a Client ID – an ID assigned by a first-party cookie on your website. Most marketing automation systems identify people by a (hopefully) unique email address. When someone attends an event, you may get a business card or catch their name and company. These are all very different identifiers!
A data collection stores these interactions in a way that can be unified on a single timeline, showing a thorough history of what, who, why, and how someone interacted with at your company:
It would be unrealistic to state that every interaction with your brand can be tracked. Legislation and privacy-first design are making de-anonymization more difficult. Word of mouth is also challenging to track due to faulty memories and lack of tracking. However, a ton of marketing data can be captured and used to make more cost-effective decisions on how and where to spend your budget.
With just about everything we do as adults, sometimes the tricky thing is the right thing to do. Analyzing the tactics that are measurable in marketing is the most responsible way to spend any company’s money.
What is NOT marketing data collection?
While many marketing automation systems try to sell their customers on the thought that their platform is a “single source of truth” for marketers, a B2B organization’s data is too complex and vast to manage in a marketing automation system. Website interactions can quickly produce millions of records annually, making data storage cost-prohibitive in a marketing automation system.
The other significant hurdle to reporting out of a marketing automation system is that vendors (HubSpot is the exception) look at the world through the lens of individual people, not accounts or opportunities like the rest of your business. For example, Marketo (now Adobe Engage) doesn’t have an out-of-the-box account object. When the rest of your business measures everything by account, logo, and opportunity, it’s easy for marketers to accidentally over-report their impact and lose trust across the rest of the company.
A CRM is an equally unsuitable “source of truth” because of the costs of storing all marketing interactions – particularly email and website interactions. These brand interactions are essential to improving your email, content, and digital marketing strategies. Still, the rest of the business won’t view them as influential to a sale – even though we, as marketers, know otherwise.
Data warehouses, data lakes, and purpose-built analytics platforms are the only data collection options robust enough to support the volume and complexity of marketing data.
Okay, but I can’t afford or don’t have the skills necessary to manage a data collection
Data collection for marketing may seem aspirational and unrealistic, although there are many lower-cost options and “freemium” data warehouses for teams with the right skill sets. There are also tools and teams for hire that address data issues for those with the budget.
We also recommend ensuring the following questions can be answered with a solid “YES” as you develop your data collection and storage:
- Do I have key events set up in Google Tag Manager and GA4?
- Are my forms integrated with my marketing automation system?
- Do I have a process for uploading offline event data (tradeshows and other in-person events) and tracking the impact?
- Am I pushing campaign data to my CRM?
- Am I following CRM campaign data best practices? (If you have Salesforce, check out our guide to campaign data best practices by clicking here.)
- Do I have attribution (first touch, last touch, multi-touch) in my CRM?
Imperfect data should never stand between you and a data collection strategy. Data will never be perfect, and waiting to use the data you have today is never the right decision.
Considerations when building a data collection storage for marketing
Should I build or buy a marketing data collection?
With the right skill sets on your team, data collection can be extremely powerful. If you must outsource due to skill set gaps to other teams, change management quickly becomes extremely difficult. Keep in mind that marketers like to try different technologies and where your buyers expect to find information changes (the average tech stack, even after recession-driven shrinkage has more than 60 applications). Each time you change tools, you must update your data graph so you can de-anonymize the new tool’s activity and include it in your activity timeline.
How much data should we store?
It’s difficult to get data analysts to agree on a safe timeline for data storage. Some believe that everything can be kept and may have value. Others believe that five quarters’ worth of activity should be enough to show trends and answer business questions.
We recommend analyzing how long your deal cycles are from first touch (meaning first brand engagement to be captured across the account) to closed won. This will help your business understand how far back your team should look to view influential touchpoints.
At the very least, ask the business how far back they may want to report key metrics and view historical trends. For example, if Marketing Qualified Lead is a core KPI, ask if there is a reason why your business would want to report on data older than 18 months.
What kind of information should I store?
The types of information you store will depend on the kinds of questions your team would like to answer, like:
- Do you want to be able to tell your content marketing team which content prospects consume before becoming a customer?
- Do you want to compare what content customers consume to people outside of your ideal customer profile? If your answer is yes, you’ll want to share website visit information in a way that can be linked to your opportunity data?
- Do you want to be able to defend your digital advertising budget? If the answer is yes, you should have UTM parameters in place and mechanisms to capture and store that data.
Collect the questions your department wants to answer and ask your CFO and CEO what kind of questions they want answered from marketing. It may not be possible to answer all of them, but you’ll get a clear idea of what kind of educational material and data collection work you have in store for your team.
How should we present our data?
Data visualization tools are usually an add-on to a data warehouse or data lake. Marketing analytics tools that are purpose-built for marketing data collection often come with out-of-the-box reports or a built-in data visualization tool. How you visualize the data will depend on your tech stack and the questions your team is trying to answer.
Please see our Playbook section and our SFDC Admin’s Guide to Campaign Setup for recommendations on how to structure your data and get ready for data collection for marketing.
