CDP vs. DMP vs. CRM: Differences, Similarities, and Best Practices FAQ
How should CDP, DMP, and CRM martech figure into your martech strategy? They’re all very different solutions, and many businesses use more than one. In fact, a lot of organizations are trying to harmonize their CDP, DMP and CRM to get the most out of all their martech and adtech. Other businesses are thinking of consolidating their tech stack down to just one of these choices. What are the most effective ways to exploit all of these technologies? This CDP best practices FAQ addresses the following questions:
- What is a CDP/DMP/CRM? What are the differences among them?
- Can CDPs and DMPs work together? Are they complementary?
- Can we leverage both a CDP and DMP to have a more complete view of the customer?
- Is CDP data best integrated into a CRM, or vice versa?
- What are some best practices for dividing work between the different platforms?
What is a CDP?
A CDP ultimately ingests customer data from many different data sources to drive a better customer experience by creating a unified customer profile and a Single Customer View (SCV). CDPs are commonly used to improve ad-targeting, segmentation, loyalty programs, and personalization at scale.
What is a DMP?
A DMP enables marketers to serve targeted ads programmatically and at scale, primarily using anonymized customer data in the form of third-party browser cookies.
What is a CRM?
A CRM typically manages interactions with customers and prospects. However, they do not usually store every available piece of information on a customer, nor do they run any type of machine learning algorithms on the data to generate insights.
What’s the difference between a CDP vs CRM vs DMP?
All three tools have their own functions and use cases in the marketing ecosystem, but there can be some overlap. What they share in common is that all three have been designed to improve the overall customer experience, making targeting and segmentation easier.
A CRM is where your sales, services or marketing teams typically manage customers and update data directly, meaning the data is input manually by these teams. A common use for a CRM is to compile and manage all contacts with a customer, including sales calls, contact center records, support, and maintenance calls. A DMP tracks third-party user behavior, but it only stores data for a short period of time (mostly in the form of a third-party browser cookie ID). A DMP helps with lookalike modeling at the aggregate level and targeted paid media spend, but you cannot use it alone to build detailed profiles of individual, identifiable customers. A CDP, on the other hand, pulls in data from all of the different sources (online and offline), to create a true single view of customers, storing data as long as necessary without limitations (including first-party, second-party, third-party, and PII data). With a CDP, you wouldn’t typically update data directly in the CDP, it would pull data automatically from other systems and aggregate it.
CDP vs DMP
Can CDPs and DMPs work together? Are they complementary?
Most CDPs can work with any DMP, so when going through vendor selection for a DMP you should choose your DMP based on your requirements and the capabilities of the DMP. You then need to ensure your CDP can use the DMP Identifier for its ID resolution, and then the DMP data can be integrated into your CDP’s Single Customer View (SCV). Most CDPs will also help you enrich your DMP with data from other systems, so you will need to ensure your CDP can send data back to the DMP based on the DMP ID. The good thing with a best of breed CDP, like Treasure Data, is that customers can select a best-of-breed DMP, if they want, and they don’t have to be from the same vendor. This allows you to have a best-of-breed DMP and CDP and not be tied into a suite with inferior functionality.
What are the differences? DMPs vs. CDPs
A CDP can do some DMP functionality, but DMPs are specific to advertising and customer acquisition campaigns. DMPs work best with new or anonymous audiences, and join sessions based on third-party cookies. DMPs operate on a massive scale that is specific to audiences, and that’s why the data they process has a limited data retention period—but can help with lookalike audiences and programmatic advertising. A CDP, on the other hand, builds a persistent, true SCV—linking sessions based on first-party data and doing ID resolution across channels. That’s how a CDP manages customer experiences across all channels, helping to provide consistent messaging and exceptional customer experiences.
CDP vs. CRM
Customer Relationship Management (CRMs) platforms manage contact with prospects and customers. Sales channel interactions, as well as contact-center and customer-support calls are also typically included. But CRMs are less adept than CDPs at integrating with other sources and types of customer data. For example, Treasure Data Enterprise CDP uses a data lake approach along with pre-built connectors to many other platforms and data sources. So while Treasure Data can use CRM data, such CDPs can also integrate and use data from many more types of martech and adtech, as well as social media and even many custom sources of data.
Can CDPs and CRMs work together? Are they complementary?
Yes, CDPs and CRMs can work together, and a CRM can serve as a data source to a CDP. A CDP can use input from a CRM, but data and control hardly ever flow the other way. Many successful companies use a CRM to feed data to their CDP, and they use the CDP to control, or “orchestrate,” other marketing processes, such as campaigns, other martech.
What are the differences between CDP vs. CRM?
CRM is mostly sales-oriented and support-oriented, designed as a tool to help sales, marketing and customer support carry out their day-to-day business. It relies mostly on these departments to enter customer data directly into the CRM—or in some cases, to take data from a phone system or call center to work effectively. A CRM has virtually become a requirement to do business today, for any company from a small business to a multinational enterprise, and is no longer a differentiable capability.
A good CDP incorporates data from all customer-facing systems including website analytics, in-store POS systems, billing systems and more. And thanks to powerful machine learning, CDP also offers predictive analytics capabilities, creating deep customer insights and actionable next steps. An enterprise CDP such as Treasure Data also handles orchestration, so it can use the unified profiles and SCVs to trigger such marketing communications as texts/push notifications, emails, calls, advertising, in-store personalization and personalized web/mobile browsing experiences.
Can we leverage both CDP and DMP to have a more complete view of the customer? If yes, how do we integrate CDP data with DMP data to create a more robust customer view?
In short, yes! Again, you need to ensure your CDP can use the DMP identifier as part of its identity resolution, and only then can your DMP’s data be integrated into your CDP’s SCV. Treasure Data, for example, can ingest your DMP data to improve segmentation, and optimize the experience for visitors across all channels. A good CDP should allow you to store your DMP data for as long as you need (not be limited to the last few months), and help you store the data in any format. This makes it much easier to get data from all of your systems. This allows you to create true data science scores (like CLTV) and not have a limited subset of data in your models. You should also be able to push your CDP data back into your DMP to improve the accuracy of your lookalike audiences. Once you have attracted that customer to your site, the CDP can take over—providing real-time personalization.
Is CDP data integrated into a CRM, or vice versa?
Yes, both in fact! You typically update customer data directly in your CRM, and in order to leverage that data you would need to pull it into your CDP as part of your single customer view, as it is a rich source of PII and behavioral data.
The CDP is also getting data from all of your other systems, and is running data science and machine learning on that data—so you are best to push some of this data back to the CRM to enrich your CRM for your sales or service channels to utilise. This could be the next best action, customer lifetime value, or propensity scores—using data from all of your channels and not a single channel.
You do need to ensure you have a master for each data attribute though, so you are not reading and writing the same attributes from CRM, otherwise you will end up with potential data integrity issues. Think of the CDP as the hub in your hub/spoke model—ensuring data is read and sent to all of the systems in your ecosystem—including your CRM.
Activating to a DSP
How would you match identities when building an audience to push to your DSP for advertising?
There is no intermediary between Treasure Data and DSPs. If the DSP ID for a customer profile has been retrieved when browsing a specific website, the profile would be able to be targeted through the DSP. Additionally, first-party data onboarding partners like LiveRamp can be leveraged to increase the match rate. Active audiences within the DSP itself would then be dependent upon the onboarder’s match rates.
How both DMP and CDP Align to Customer Journeys
As an example customer journey, your DMP will target lookalike audiences and help with retargeting, then your CDP would take over at the “consideration” stage right through the “retention stage.” Don’t forget when evaluating both DMPs and CDPs that your CDP can enrich your DMP data to improve your lookalike modeling, and retargeting campaigns. This would be an area of focus.
Are CDPs a solution for the cookieless world?
Yes! Third-party cookies are going away, and this is going to prove a challenge for most DMPs, as you won’t be able to stitch sessions or see a user’s full journey. A CDP like Treasure Data Enterprise CDP can match data on any customer attributes, actions and consent, so you can build a true single customer view easily, without relying on third-party cookies.
Are audience activations done via CDP, or do businesses need to work with a partner?
An enterprise CDP, such as Treasure Data, will activate audiences via your existing tools—being able to select best-of-breed solutions for your overall martech stack is a CDP benefit. Your marketing team is already used to working with activation tools, and these tools are industry-leading with key functionality—so it is more effective if marketing can continue to use current tools and not have to learn entirely new ones.
This is why the right CDP is a good choice to orchestrate marketing across all of your activation channels. A CDP with orchestration capabilities can determine which channel to use, which message is best suited for the target audience, and when to send specific messages—for every single customer. Your marketing team can build templates and manage content in the activation channels as they do currently.
And that’s part of what’s driving today’s move toward CDPs: Marketers don’t have to switch, and they don’t have to learn a new (potentially inferior) tool. Rather, they can focus on building successful campaigns, exploiting new insights from their CDPs, and creating profitable, personalized customer journeys that acquire new customers and build loyalty in current ones.