CDP Buyer’s Guide: Should You Build or Buy a Customer Data Platform (CDP)? 7 Key Questions
In this time of technology-driven disruption in every industry, companies are realizing that their competitive market position relies heavily on how well they can engage, service, and delight both prospects and existing customers. To do this, it’s critically important they understand who their customers are, what each customer’s journey looks like, and how to give every customer the right experiences at every point along their journey. Plus, they have to do all that at scale. Fail to do this, and competitors with a better grasp on their customers, and who can deliver a more personalized experience, will be happy to take those customers away.
Customer Data Platforms (CDPs) have become a new, hot category. The reason? Marketing and CX leaders who’ve used them say that no other technology is as elegant or easy to use. A good CDP pulls data from any existing data source, persistently stores first-party (and other) data, handles identity unification, provides plug-and-play integration with many other sources, and handles activation across all customer touchpoints, based on customer insight.
Build or Buy? Listen to Experienced Marketers Who’ve Been There
But if you’re like many retailers, consumer brand directors, hotel chains, gaming companies, software vendors, or publishers, you already know this, but you’re facing a high-stakes build-or-buy decision about your CDP. How can you get the insight and experience you need to make the right CDP decision: build or buy? What are the hidden costs and overlooked, but necessary, capabilities? What are the key questions to ask, and what areas should you consider?
We’ve listened to people who’ve been there—those who built their own CDPs, as well as those who’ve purchased a solution. Based on those discussions, here are the most critical considerations to use as a customer data platform buyer’s guide when making the build-or-buy decision.
What’s your typical time-to-market (TTM), and can you afford to wait?
Take a look at the surrounding business environment and the market your company operates in. How much time do you typically have to wait to unify your customer data and put it to use executing your company’s strategy? How much time can you afford to spend on building your own solution in-house?
Purchasing a pre-packaged software solution will give any organization a massive head start in getting a true 360-degree view of their customers. The pre-configured integrations save valuable development time and create a ”plug-and-play” environment that allows you to quickly see value from customer insights. Most companies can get up and running within months. (A top gaming company used Arm Treasure Data to unify data from 21 databases with nine different customer IDs in just over six weeks). Another benefit of working with a CDP vendor is that a proof of concept (POC) can be up-and-running and evaluated in a few short weeks, reducing the risk that the end result won’t meet your company’s business goals.
How many people will you need to do it yourself, and what skills do they need?
An often-overlooked—but critical—area is the long-term availability of skilled data science resources. A major brand’s head of data science and analytics recently told me that his team’s skill set was the biggest issue in his company’s internal CDP effort. Skilled data science resources are hard to hire, and many teams lack enough talent to handle all of the data science and AI projects the executive suite needs to deliver if there’s to be hope of making their numbers. If you plan to build a solution in-house, you’ll need to conduct a thorough assessment of your team to determine whether they have the right knowledge and expertise. While data volume and data sources are growing exponentially year-over-year, data science talent is growing scarcer and more expensive. According to Glassdoor’s data, the average salary for a data scientist is around $137,000, and trends indicate that figure will continue to rise.
Another consideration is that the “big guys” like Amazon and Facebook have a tendency to poach other companies’ employees with strong technical and data science skills—so the people who build your system aren’t necessarily the ones who will maintain it for years to come. And—by the way—you’ll need more than just software development and data science knowledge. Your organization will also have to enforce strong data governance and handle compliance with the growing array of data privacy regulations. One of the greatest advantages to third-party CDP solutions is that they come with built-in experience in these highly specialized areas.
What’s your total cost of CDP ownership (TCO)?
Your procurement team probably has experience running total cost-of-ownership models for vendor selection. But keep in mind that there are still often hidden costs to consider.
A large consideration is the cost of internal resources if you “roll-your-own” solution. Resources required for both building and buying vary with your specific needs, but if you build it yourself, you’ll need to include these costs in your estimates:
- Hiring and staffing
- Data migration
- Upgrades and maintenance
- Hot fixes
- Help desk
How much should you budget for ongoing improvements and maintenance?
One advantage to purchasing a solution is that vendors will continue to improve the platform by using the latest technologies, adding new integrations, and providing new, valuable modules. Because they can leverage these investments across their whole customer base, they can continue to innovate. For example, Arm Treasure Data’s latest product release included the following:
- New policy-based permissions for better compliance to increasingly stringent regulations
- New integrations with tools such as Box, Snowflake, Sansan, Mailpublisher and Karte
- In-app purchase events for Android and Unity
The convenience of continuous product updates with new functionality and integrations—as well as new and updated connectors for your constantly evolving marketing stack—make third-party vendors an easy choice for many companies.
In addition to taking advantage of new developments, the vendor also handles all of the ongoing maintenance and other IT tasks that can drain resources from your company for years to come. These include things like backups and recovery, security, support, updates and bug fixes. If you develop a solution in-house, don’t assume that work on the platform will end when it launches.
Perhaps the most underrated factor to consider is whether your organization has the expertise required to navigate this new world. Having witnessed many other implementations up close, a CDP vendor can supply the skills and knowledge required for success in this area. You can often lean on vendors to supply the best practices in customer data management, data enrichment, and unified customer profiles. After these golden profiles have been created and are being constantly updated, they can also show you how to activate campaigns at scale. And of course, the best CDPs provide and maintain the AI for real-time segmentation (in areas such as high-value customers, churn risk, cross-sell/upsell opportunities, etc.).
Security and privacy concerns
In the modern era of consumer privacy regulations such as GDPR and the California Consumer Privacy Act, connecting all your customer data can pose security and privacy risks. Where did your identifiable customer data come from? Did those customers opt-in to share it? Where is it stored? What does it tell you? How is it being used or shared?
Given that the typical organization uses an average of 91 marketing cloud services (according to Mary Meeker), it’s easy to lose track and fall into noncompliance. Going with a reputable third-party CDP vendor gives you peace of mind that you’re putting your customers’ data in the most capable hands.
What are the ongoing opportunity costs of a slow or inflexible system?
Many who try and build their own solution overlook costs of a system that isn’t optimized for speedy results and usability. In a recent Forbes survey titled “Data vs. Goliath,” 47 percent of participants said that it takes more than a week to get results of campaigns or customer experience deployments, and another 47 percent said the wait was three to five days. In some markets, that’s not good enough, and the extra engineering required to achieve near real-time reporting and retargeting would be prohibited.
Unless you have people dedicated to ensuring you can personalize each of your customer journeys in near real-time, and get the results of each campaign in time to capitalize on opportunities as they arise, you might be incurring significant costs due to missed opportunities. This is particularly true in retail and consumer packaged goods (CPG) businesses, as well as in gaming and entertainment, where consumer emotions, whims and up-to-the-minute shifts in tastes and fashion are paramount.
A CDP can offer marketing departments the ability to get results and react quickly to changing needs, as part of their standard offerings. These capabilities are attractive to CMOs whose volatile competitive environments require quick responses to win or even stay in the game.
CDPs: A Unified View of Each Customer, for Personalized Experiences
Businesses today need a unified view of each customer to succeed. The right technology can empower marketing, CX, and product teams to personalize customer journeys with one unique profile per person, driving revenue growth and increased customer loyalty. A CDP offers a 360-degree view of the customer, a detailed view of the customer journey and an ability to execute on a personalized omnichannel strategy.
But what’s really priceless is a deeper understanding of individual customers, even if you have millions of them. While “building” may seem to be an attractive option, the time-to-market considerations, technical requirements, long-term support, costs, and potential of project failure or delays all make “buying” a less risky option.
Request a demo today to see how Arm Treasure Data enterprise Customer Data Platform can help your business avoid the potential headaches of building a CDP.