There are so many acronyms in the software space and this can make it difficult to know what DMPS, CRMS, and CDPs really are, why an organization may have one or all of them, and the impact these systems can have in relation to infusing customer insights into your organization. In this post, we hope to level set and to provide context as to what these are and how they can impact your role in your organization.
CRM = Customer Relationship Management
A CRM vendor will tell you they help an organization focus on relationships. Under the hood, most CRMs are the place you store all your customer data. Sometimes they get a bit unwieldy as they tend to be a dumping ground for any data you happen to have, but it is not always organized or managed well. There are many players, but SalesForce is the obvious standout.
Common CRMs: SalesForce, HubSpot, SugarCRM
DMP = Data Management Platform
This term gets used a lot and tends to cause the most confusion. In the early days of the internet, websites wanted to know what the traffic to their site looked like. Were people clicking on the site? If so, what places on the site had the most traction? This was the genesis of analytics tools like Google Analytics or Adobe Analytics. You have a lot of data, but it is pretty anonymous, so you really don’t know much about the individual users. Analytics got a lot smarter by tracking users across thousands of sites rather than just one. They also allow tags and other metadata to teach you more about the interactions on your site so you can learn something about the user by the things they do while they are visiting.
To solve this problem, DMPs were born, and today there are many players in this space. The common theme among tools is that they mostly collect website traffic and clicks collected anonymously. Then, they try to layer in other data such as point of sale, social, and census data, but it usually starts with web traffic. The data tends to avoid anything personally identifiable, but these toll providers were also among the first to get into trouble with data privacy and are one of the main motivators for new internet privacy laws. Ever wonder why so many sites have the annoying "accept cookies" popups? Cookies are the dominant mechanism DMPs use to track users across websites.
Common DMPs: Google 360, Adobe Experience Cloud, Lotame, Oracle Bluekai
CDP = Customer Data Platforms
CDPs are the newest entrant to the customer data space. They focus on 1st party customer data rather than the anonymous data a DMP holds. They also attempt to hold customer data from all sources. This means they can contain full or portions of CRM data, but they can also have website data, social media data, and product usage analytics. They attempt to pull in data from all sources and help you understand it and provide control and governance. Some CDPs brand themselves as MDMs (master data management systems). Other than more letters, this usually means a CDP with more governance controls so the data can be used by more than just the marketing functions.
Common CDPs: Segment, Signal, DOMO
Why Do Organizations Care?
The main reason for the collection and analysis of customer data is not new. Organizations want customers to buy more of their products and services.
What has changed is how customers actually make these purchases—largely because of the internet. All of these systems—DMPs, CRMs, CDPs—provide data that underpin marketing campaigns, website ads, website content, and sales funnel activity. Every one of these activities relies on one key ingredient to be successful: knowing as much about customers as possible so the activity is as catered as possible.
Organizations generally achieve this level of customer data analysis through fancy things like data science and machine learning. Under the hood of these processes is a certain amount of inference and assumption. They all try to imagine what customers like and what they think by watching activity across the internet, across the company website, through things they bought, or through things they have "liked" on social media. Armed with a customers' imagined preferences, individuals are segmented into like-minded audience groups for the following types of activities.
Email campaigns
We all get marketing emails from organizations trying to get us to buy stuff. Some may use the strategy of more email equals more sales. As users, I think that as we get more emails about stuff we don't care about, that results in more unsubscribes... not more sales. Worst case scenario, it means you get reported as spam and you can't email anyone anymore. Targeted emails mean fewer emails and higher sales. The more accurate the targeting, the more effective the campaign.
Ad targeting
In the past, internet ads earned a bad reputation and spurred ad blocking add-ons across the net. Today, they are maturing into personalized, relevant, and less intrusive content, which is a good thing since they fuel more of the internet. Like marketing campaigns, less is more. This is achieved through a better understanding of the customer and what they want. Advertising sellers go to great lengths to understand the viewer so they can better understand their preferences. Arguably, this is the main driver behind Facebook, Pinterest, YouTube, and Twitter as some of the biggest online advertisers in the world. They are all DMPs of a fashion since they hold customer data that can be used for better targeting.
Sales Funnel Targeting
Sales funnel management is a critical process for many organizations. The top of the funnel can have thousands of prospects—more than your sales team can handle. Being able to segment the population into bite-sized groups that allow you to focus on the people most likely to buy is everything. Similar to all the rest, inference and data science are used primarily.
Website Content
When we want to know more about a company and their products, one of the first places we go is the company's website. Often we go there several times. This makes the website a key tool in lead generation. Armed with my segmentation information, websites can dynamically alter the content of the site to cater images, blog posts, service offerings, and other items to better suit my motivators.
How Should Insights Professionals Be Thinking about These Platforms?
There is certainly an argument to be had over the efficacy of behavioral data to accurately predict someone's tastes, preferences, or options versus the declarative data traditional market research gathers. To some, it is a hot topic, but the smart play is to avoid the fight. The marketing engines in most corporations are funneling large amounts of investment into predictive platforms tied into their websites and campaign systems.
Insights professionals should continue to do what they do best: provide insights by uncovering the why behind the what. This is exactly what the teams at a large Canadian retailer did. The insights professionals and the data scientists collaborated to use their insight community to test the underlying data science models used to drive the segments and campaigns. Armed with a new way to validate their work, the data scientists learned that their modeling could be enhanced. Through the simple conversations available through a customer insights platform, they made improvements to models that predict things like the life event of moving. Better models mean better results and that is exactly what they got. Read the full story here.
