This summer I immersed myself in customer data platforms having decided earlier in the year the technology is transformative. I believe CDPs are the biggest thing in Marketing SaaS since the advent of site analytics. CDPs are a generational, transformational, business technology.
The power of CDPs includes setting the stage for true customer identity resolution in a post GDPR world where first-party relationships and primary evidence are essential to survive. I believe over time CDPs will become the database of record for everything relating to customers. Everything.
Enough hype. The fact is we’re in the early innings as far as adoption and knowledge. So let’s talk about the problems and challenges. Here is some of what I learned over the summer relating to CDPs. Also, thoughts and ideas I hope can to be helpful solving some of those problems and challenges.
The first problem with transformational technology is getting people onboard with transforming. How often do major martech integrations happen? Not too often, because change is scary and change management is difficult.
When the fear of not transforming is even scarier than the integration fears, the CDP market will be ready to blossom. That will be driven by the C-Suite since CDPs are major investments in time, money, talent and org. It is starting to happen. With competition for customers a click away the fear of not knowing who your customers are when Google, Facebook and Amazon already do is real. Customer intelligence forms the basis of any brands future and any management’s responsibility from here on in.
I spent a large part of my summer speaking with brand executives and getting primary evidence myself on their data analytics and marketing problems. I learned the many flavors of what they need a CDP to do.
I also spoke to over twenty CDP vendors trying to wrap my head around their issues and challenges for adoption. I met with engineering teams and product vendors that are building CDPs directly for brands. In addition, I read everything I could, listened to too many webinars saying the same thing and attended some conferences.
The CDP space is evolving and with that are issues you see in developing software markets. Product fit, market fit, talent and customer success to name a few. That said, CDPs Datorama and Treasure Data were purchased this summer for a combined $1.2B. I’m reminded of martech circa 2005 when site testing and optimization technology were brand new and the site analytics space was starting to bustle. Omniture, WebTrends, Visual Sciences, Coremetics were all coming of age and Urchin was bought by Google becoming GA. I think a lot of what happened during that period repeats itself – the market leader goes public, the other top players get bought and a major ad platform buys the one that can get the broadest adoption. That last one might actually happen twice.
The best thing I read this summer on CDP was a piece “Why You Should Build Your CDP.” I agree with most of it. Except the advice.
Deciding to build a CDP is a decision with many variables that need to be evaluated before you can be certain it is the best path. Build is not right for everyone. In some instances organizations have the cycles and expertise to roll their own. The more advanced you are with modern database technology, APIs and real-time data streaming the better off you will be. But even with that it still might make sense to use a vendor.
Here are some reasons to use a vendor:
Cost of hosting: Economies of scale here and this could become a large bill depending on the size of your data and the size of your business. Plus you need DevOps and teams of people to support this. Maybe you have these teams already. Then you’re hiring more of the toughest people to find.
Speed to impact: Building will take you a year. With a vendor you probably cut that time in half. That’s two quarters of acceleration. Not insignificant to most. That might mean tens-to-hundreds of millions of added value. It might save someone their job. Time is money in standing up a CDP.
Org readiness: There are numerous challenges that confront an org anytime a new product at the size and importance of a CDP is being built. Your engineering team is probably not sitting around without a roadmap. You have already approved and budgeted important projects over the next year. And this is far more difficult build than some microservices migration since the consumers of a CDP are outside engineering.
Learn before building: Vendors having gone through many installs and know things about the process that your teams do not. They probably know stuff about APIs and other technical things that your teams are not going to know about CDPs. Going through the process with a vendor will make you smarter when and if you decide to build it yourself – saving you time and money with a better process.
Be active as the market evolves: Why not be an early adopter? It’s hard to argue with the power of unified identity. And even if it’s not perfect (it isn’t) why not do the best you can with imperfect data. Whatever you do it is going to be better than the customer data you have now.
I want to be clear I’m not against the idea of building and owning your CDP. I’m actually someone with a build background and there are clear paths and benefits to building part or all of a CDP. I am seeing brands building their own systems for multi-touch attribution and identity. It’s understandable why. These are core parts of marketing and certain businesses and their business models have what might be considered edge-cases at a vendor but are core-cases that demand internal specificity (of course everyone thinks they are edge-cases and most are not).
In addition to the value of truly owning your customer database the process of building while frustrating provides unequalled learning. There’s also a middle road. It’s possible to build pieces or light and quick proof-of-concepts serve as a test and help you make the big decision and can integrate with whatever future decision you make.
All through the summer the number one question from brands about CDP was “where do we start?” My answer is the same for everyone. Business and use cases.
Business and use case development is an important starting point for all CDP decisions. They are the first steps in building a prioritized roadmap that understands trade-offs, tech and talent gaps, and creates a time-to-benefit analysis where quick-wins (something everyone wants) can be obtained while a longer range plan is being executed on.
Business and use cases are different for every brand and business and this is important because CDP is not a one-sized fits all. These cases also bring to light issues and deficiencies with the existing marketing stack and vendor selection that may need to be addressed before a CDP is stood up (garbage in, garbage out). When experienced and knowledgeable operators lead case development they will cast a light on organization and talent gaps – the most critical elements for a modern marketing practice and for executing post-CDP success.
There are two large problems with CDP business and use case development I noticed over the summer working with brands.
The first issue is that many brands do not have mature product development or product leadership that can work with marketing. Brands historically have not built many products for internal stakeholders. Business and use case development and the product work that rolls off it are not something engineers or marketers are expert in. Senior marketers at brands often confuse business cases with use cases. Those days and businesses are slowly coming to an end. Every business at its core will have to be a platform in order to compete on data as I wrote about over the summer.
The second large issue I saw this summer is that marketers are creating their business and use cases off their current marketing data, analytics, segmentations and vendor capabilities. This doesn’t make sense. Use case development for CDP is about meeting future needs of marketer with what is possible to create value. A big part of the foundational work should be to validate important assumptions about customers based on your existing data (quality usually fair-to-poor). Business and use case possibilities for CDP are more than mind expansion – they require mind reprogramming.
One major breakthrough of CDPs is their ability to unlock segmentation from the channels and verticals and truly understand the customer from a top-down view across the company. This also creates additional consumers of the CDP across the organization. This is where and why business and use case development should also exist outside of marketing. Value across many work streams of a business is a big reason CDPs are so valuable. It’s important to capture all the value.
Summer’s Almost Gone
Ultimately what I can report of the Summer of CDPs is that everything you would expect to find in any burgeoning software market exists. Hype, excitement, confusion, fear, restlessness, dissatisfaction, competition, talent shortages and service gaps. Underlying all of this is some extraordinary technology and great entrepreneurship.
From the brand side, there is a growing understanding as they get educated that CDPs are only the beginning. The machine learning, predictive analytics and automated decisions that will emerge from the CDP will power a wide range of optimizations across multiple business lines, from merchandising to supply-chain to media spend. It will power the algorithms that understand the needs, intentions and behaviors of the customer in order to control the experience and improve relevance. It will become the customer database and power a platform.
Now that summertime is almost gone some brands will turn their focus to their existing capabilities and teams in preparation for upcoming code freezes and holiday shopping. Others will hope to finish the CDP installs and product dev taking place now and light-up Q4. Many will move one-step forward the rest of the year to better understand their needs and challenges as it relates to customer data in preparation for big changes in 2019 and beyond.
For me, it was an amazing summer of learning about vendors, clients and services related to CDP. I will be paying it forward to brands and helping ensure the success of the market and most of all, relevance to consumers. Cowabunga! And feel free to reach out if you want to discuss CDPs.