Getting Your CDP from Launch to Live

Today we will break down the private CDP deployment process.

When we say that Velocidi is “like an in-house solution,” it’s possible that some of you are reacting with apprehension. “In-house” is a term that can carry expectations of a cumbersome ground-up build, excessive maintenance needs, and insufficient vendor support.

Let’s put those apprehensions to rest. There are a few ways Velocidi has eliminated potential issues associated with taking on an in-house platform:

1st – Lots and lots of automation

AWS already comes with its own automation. We’ve built even more on top of that.

The deployment of the platform itself is automated. Remember your CDP comes pre-built. So once you have your AWS server set up, all it takes is a few keystrokes and in as little as 30 minutes the platform actually installs itself.

Not unlike this fun visual we’re borrowing from Business Insider:


We’ve also supplemented the automation involved in server scaling. If your data volumes go up or down, more storage is added or removed automatically, within pre-set limits. So you never spend more on storage than you want or need.

automatic deployment

2nd – Automatic status reports

If you were wondering, “How will you provide support and system maintenance?” the answer is: automatic status reports.

Our team does not maintain continuous access to your private CDP. That would somewhat defeat the purpose. Instead, we support you by setting up your AWS account to send our AWS account regular status reports. That way we can keep an eye on things without actually being inside your system. If and when an issue arises, we step in to help you resolve it directly.

Velocidi is designed to be just as easy for brands to implement as a SaaS platform. But unlike a SaaS platform, it is absolutely private to your business.

3rd – Personal training sessions

The third way we eliminate potential issues that has nothing to do with automation. Alongside the technical deployment, we also set you up for success with training sessions on how use the platform, and how to evaluate its success. We do our best to make sure you draw the most possible value from your CDP.

Setting Objectives

At or near the time of installation, our customer success team works with you to outline objectives and success metrics. This helps both our team and yours to prioritize channel integrations in the system, define segmentation needs, plan experiments, and tailor predictive models to your business.

Tagging and Integrations

We went into tagging a little in “4 Advantages of Private Deployment”. Your 1st-party tracking tags use the same domain, or a sub-domain, of your website and digital properties. This allows your CDP to create 1st-party identifiers for each visitor. And it means your data is encoded to only be meaningful to your business. If third-parties can’t decode your identifiers, they can’t use your customer data for anything except your expressly defined audience activations.

Integrations are the connectors between your CDP and every other system you use in marketing, sales, and customer experience. This is how the CDP will collect data from all of your customer touchpoints and send data back out for external delivery systems.

We do a dedicated training session with you for adding and removing integrations, and tagging new properties. This also includes quality assurance checks on the data to make sure everything is valid and flowing smoothly. This step takes less than a day. In no time you’ll see real-time granular, multi-channel data streaming into your dashboard.

Segments, Activations, and Experiments

This is where you learn to create custom audience segments within the platform, and send them to external delivery systems. We will also train you in the A/B testing feature of the platform, and design a short-term experiment to verify your CDP’s effectiveness.

As soon as data starts pouring in, Velocidi’s machine-learning predictive models also begin their training. This is what allows your CDP to create audience segments based on predictive behavior attributes.

It takes 2-6 weeks to train predictive models, until you start seeing consistent and reliable predictions. Additional experiments you set up during this stage will run on a similar timeline. And we will plan time for multiple iterations so we can learn from initial results and make improvements.

At the end, you’ll have a fully operational set of smart, dynamic audience segments activated in campaigns. All of them will be optimized to help you achieve your specific goals and grow your business. Plus, you’ll always have a rich depository of quality 1st-party data to use in more ways in the future.

That’s the whole deployment process in a nutshell!

How does it look? Are you ready to take the leap with us? Send us your questions and comments at


Download a handy one-page infosheet outlining Velocidi’s onboarding process to share with your team.

Related posts