Velocidi FAQs

A customer data platform, or CDP, is a type of software that collects data from all your customer interactions into one place, and creates omnichannel customer profiles and audience segments from the combined data that can be shared back to external systems. Thus, facilitating a closed loop of omnichannel customer intelligence and experience orchestration.

You might think of it as a central, unifying data management and marketing orchestration hub at the center of your tech stack.

Also Read:
Gartner: The Marketer’s Guide to Customer Data Platforms
The CDP Institute: What is a CDP

We sometimes refer to Velocidi as a “private” customer data platform because while most CDPs are offered as a SaaS (software as a service) product in a shared infrastructure, Velocidi is offered as a licensed software installed in your own cloud. It’s still managed and supported by Velocidi, but it behaves as if you had built it in-house, instead of like a third-party application. This way, you remain the sole owner controller of your data.

The advantages of removing a data management “middleman” in this way are numerous. It helps keep your data safer. It helps you protect data privacy and customer trust. It also makes you less affected by browsers’ cookie blockers and ad blockers since your 1st party tracking tags are served natively under your domain instead of a third party domain.

Also Read:
What Makes the Velocidi CDP “Private”?
Four Advantages of Private Deployment

We are responsible for ensuring your CDP is always running smoothly, as well as to keep it up to date with the latest features and security patches. Our monitoring system will ensure your CDP is functioning at all times, or warn us about any issue with it, so that our engineers can fix the problem before you even notice it. Our access to your infrastructure covers all operational support, and there is no need for your data to be exposed to our system.

Velocidi is used by DTC and e-commerce brands large and small. We are an ideal solution for DTC brands looking to accelerate growth using machine learning customer intelligence solutions.

Also Read:
How Private Platforms are Redefining CDP Expectations
Harnessing the Data Sets of DTCs

Velocidi can help optimize a brand’s retargeting by using our model for predicting propensity to purchase. Our Likelihood to Buy model quickly detects patterns in online customer/visitor behavior that indicate when a customer/visitor is likely to convert, and applies a dynamic attribute to each customer profile. Using this model, the brand is able to create audience segments that isolate the brand’s “likely buyers” so they only spend retargeting budget where it will make a difference.

Our retargeting case study with Barkyn, a DTC petcare company in Europe, walks through this in more detail. In 2019, we were also awarded Best Customer Data Platform from Digiday based on this case study.

Velocidi can help brands acquire new customers and improve prospecting by using machine learning models to identify high-value audiences from a brand’s 1st-party customer and visitor data that can be used as a base for lookalike modeling on outside advertising platforms. For example, Velocidi’s model for predicting the likelihood to purchase can be used to build an audience of visitors who are likely to make a purchase in the next 7 days. This audience can then be activated to Facebook and be used as a base to create a lookalike audience on Facebook for a prospecting/acquisition campaign.

Also Read:
How to use your first-party customer data to reach new audiences

Velocidi can help decrease churn and increase retention by predicting when your customers are in-market again for a specific product. This helps you to make communications more timely and personalized and cut down on unread emails.

Velocidi predicts the next purchase for each customer as an attribute in their individual profile. This allows dynamic audiences to be created according to optimal time frames for engagement that capture customers who are likely to be in-market for a brand’s product. Customers with multiple purchases have unique next purchase predictions based on their buying history, which can be generalized or specific to a product category. However, the model also works for customers right after their first purchase by inferring the next purchase date based on global user behavior. This helps to improve customer experience by not showing ads right after a customer has converted. Save budget for when customers are most receptive and know when your customers will be needing extra attention to retain their business.

Also Read:
Reduce Churn Using Next Purchase Predictions

Yes, first-party tagging and identity resolution gives us the foundation for more accurate multi-touch attribution. A brand’s customer data platform is already connecting customer touchpoints across devices and channels. Machine learning helps decode which touchpoints are separate attributable events, and which can be identified as one multi-screen session. Velocidi’s attribution reports use multi-touch attribution modeling to decide which of those touchpoints have the biggest impact on revenue, now and in the future. This allows brands to know which sources, mediums, and campaigns are responsible for specific conversions and how much revenue they drove. Velocidi also offers a long-term perspective on revenue tracking by predicting Customer Lifetime Value by channel, enabling clients to invest in the channels that are driving the most value.

To understand channel performance, Velocidi offers a long-term perspective on revenue tracking by predicting Customer Lifetime Value by channel, enabling clients to invest in the channels that are driving the most long-term value.

Velocidi’s segmentation user interface can leverage any combination of user attributes and custom rules to generate segments. The user attributes available for segmentation can be onboarded from offline sources or generated from live data collected in the CDP.

Included in the attributes available to you are the predictive attributes automatically generated within the CDP based on real-time behavioral data. You can easily create segments based on predictive CLV, likelihood to purchase, and more.

Only your existing marketing smarts! No technical or programming skills are necessary. The segmentation tool has an intuitive UI using rules and logic that enables anyone to build detailed and multidimensional segments.

In case you have any doubts, your onboarding process does include training sessions to ensure your success.

Also Read:
Getting Your CDP from Launch to Live

Velocidi can help decrease churn and increase retention by predicting when your customers are in-market again for a specific product. This helps you to make communications more timely and personalized and cut down on unread emails.

Yes. Our user interface utilizes Boolean logic to create rules. Rule-based segments are groups of customers/visitors that match a given segment rule. Segment rules act on customer and visitor attributes as keys combined with operators and values, enabling the implementation of full Boolean logic expressions. Attributes can be created from collected events, onboarded from CRM systems, or provided by third parties.

Yes, using a subset of SQL.

AOQL is the language developed and used by Velocidi to query event databases. Its main objective is to provide a set of operations common to all data storage backends in use, such as relational databases and Elasticsearch. The language is a subset of SQL and its semantics try to mimic the latter’s behavior on a standard relational database. It is intended to be suitable both for “table scanning” queries (e.g. selections, filters) and aggregation queries (with GROUP BY). It most notably differs from SQL due to the absence of subqueries – AOQL intends to be targeted at data retrieval with a low server-side computational cost.

Some of Velocidi’s API’s are also available through a gRPC interface. In particular, both the Event Storage and User Profile Storage modules provide gRPC interfaces, giving access to the systems’ captured user events and user attributes.

Yes. User segments can be activated to external platforms on multiple channels, including social media, email, and display. Activating segments to marketing platforms enables the creation of campaigns that target these segments.

Your private CDP license is priced based on the volume of customers and events processed by the system. The price is not influenced by the number of campaigns executed in external marketing platforms.

Velocidi provides multiple prediction models, tailored for the e-commerce market. The models are trained automatically right from the moment the CDP starts receiving its first events and, as time passes and more data is captured, the predictions will gradually become more accurate. No configuration is required on the client’s side.

Velocidi currently provides predictions for customer lifetime value, likelihood to buy and product recommendations.

Predicted CLV tells us the expected dollar value from a customer in the upcoming months. Likelihood to buy tells us the probability of a customer making a purchase in the near future. Product recommendation models give us a list of products most likely to be purchased by a user.

All models provide results on an individual customer basis, so each customer will have attributes in their profile that come from the predictive models.

The machine learning process generates new user attributes to be used in the segment creation process, or directly activated to campaign platforms. For example, you can create an audience segment of customers who are predicted to have a high likelihood of buying within the next seven days. Then you would activate it to your campaign platform to start delivering ads to that segment.

Velocidi uses a managed license model, which means that Velocidi manages the software, but the CDP is deployed on the client’s own AWS account.

The only thing you’ll need before installing the CDP is your own AWS account. Velocidi is responsible for managing the whole cloud infrastructure required to run the CDP.

The first two steps, which are setting up an AWS account and enabling data collection, can be done in two days, depending on the availability of the client. Once the setup is done, data ingestion starts, and after two weeks the first machine learning models come to life. After 4 weeks you can start using the predictions generated by your data.

Also Read:
Getting Your CDP from Launch to Live

The software is updated every two weeks. Each new release includes bug fixes and new features. There is no additional cost for keeping the system up to date. Urgent updates or fixes are pushed live as soon as they are ready.

Users can access the CDP using their web browser. Each user will have their own username and password to access the system.

When you license the Velocidi CDP, the installation and onboarding process will include hands-on training sessions with our team to make sure you are getting the most out of your CDP as possible. You will also be guaranteed a set number of hours each month for support, depending on your license tier.

Also Read:
Onboarding Infosheet: Setting You Up For Success
Getting your CDP for Launch to Live

The CDP takes care of that automatically after the initial setup. Primarily the platform uses javascript tags and direct API connections, which capture data from virtually any source. A javascript tag will be used to record customer behavior as customers browse your online store. An API connection might be used to import customer data from another platform in your tech stack.

Setting up Velocidi is very similar to setting up Google Analytics and detailed instructions can be found here. Placing the 1st-party tag on your website will enable you to collect customer data from your website and online store.

Velocidi also offers the ability for users to set up imports and simple transformations of customer and product data through a simple form-fill style UI.

Typically, real-time events are collected from the website using a 1st-party tag. Website data can also be collected via batch, streaming, and API.

Yes. Your Velocidi CDP directly collects first-party data and recognizes unique users across devices. That includes both web browsers and mobile apps, as well as tablets, and even in-store purchases. If your customers are using multiple devices and channels to make purchases, they will be recognized as the same individual by the CDP. This ensures you maintain a clean, unified view of omnichannel customer activity.

Yes, Velocidi has mobile SDKs for iOS and Android.

Velocidi supports uploading offline PII data assets via the GRPC API or file imports.

No, Velocidi does not require a 3rd-party onboarder.

Your Velocidi CDP captures granular customer behavior from any channel, such as browser, mobile, email, in-store point of sales, or any other network-enabled device. There is a multitude of default behavior attributes built in to include purchase and browsing behavior, interactions with emails and ad campaigns, etc. If there is a specific behavior attribute you need that is not already built-in, it can be easily configured.

Velocidi provides individual records of each customer’s cross-channel interactions, resolved to account for multiple IDs and devices. This information is used by our machine learning algorithms to create intelligent audience segments for personalization. Velocidi also provides an aggregate level view of all customer activity across channels.

Also Read:
Beyond the Single Customer View: Capturing Omnichannel Customer Journeys

It is a 1st-party cookie to the client site, set via HTTP headers. It is not a JavaScript set cookie. This means it is not affected by Safari’s ITP, among other cookie-blockers.

The user id cookie only stores a UUID-type user identifier.

Yes. Users can be identified with multiple user id types, specified via API / HTTP parameters, e.g. CRM IDs, email hashes, etc.

Your Velocidi CDP will generate its own unique ID for each user based on a first-party browser cookie. This first identifier will be used as a starting point, and additional identifiers will be added to each customer’s “ID graph” through specific interactions. For example, when a customer logs into a loyalty account or provides contact info during a purchase, that information is anonymized within the CDP and used from then on to link customer activity across devices and channels. The same thing can be done if you upload a static file of customer records from an offline system.

All known identifiers for each customer are collected into one customer profile, so their activity across channels and devices can be gathered in one place.

Yes, Velocidi is designed to ensure that the client maintains all ownership and control of their customer data and it can be configured to meet the most stringent privacy regulations (including GDPR, CCPA, etc.).

The system is deployed on a private AWS cloud that is owned and controlled by the client and all of the data collected is completely owned and controlled by the client.

By design, the client is the Data Controller and Processor for all of the processing activities, except for CDP maintenance and update operations, providing services per the client’s instruction/approval, and modeling activities per the client’s instruction/approval.

Everything from Magento to Facebook. Your private CDP can import offline data from POS systems as well as receive real-time event data from your online store and digital campaigns. And it will also integrate with advertising delivery systems so the data generated within the CDP can be used in campaigns and personalization.

There are no limits to the number of applications you can integrate or the volume of data passed through each integration. For examples, see our Integrations page.

Velocidi supports a customizable Opt-in / Opt-out mechanism, that enables users to provide an opt-in or opt-out selection. When users opt-out, their data is no longer processed. The default behavior of the opt-in/opt-out mechanism can be customized and integrated with existing Consent Management Platforms that adhere to the IAB’s framework.

The length of time data stays linked to an ID can be any amount of time and is customizable by the client.

Data storage time frames can be customized by each client.

Typically, user profiles are stored in the live environment so that they are readily available for programmatic use forever and events are stored in the live environment for 3 months to be readily available for programmatic use and then moved to an archive. In the archive, the events are not readily available for programmatic use but can be made available, as needed. However, in either case, data is typically not discarded, unless instructed to do so by the client.

Clients might choose to discard data after a certain time period to control the costs of data storage from their AWS account, for example.

Yes. Velocidi has a standard taxonomy (see here for more detail: https://docs.velocidi.com/collection/collection-overview/), however, the taxonomy can be customized for each client and new attributes can be added if needed.

Taxonomy rules can also be added after data ingestion.

Yes, Velocidi accepts both structured and unstructured data.

Yes, using deterministic ID matching. User profiles are merged into one profile upon a deterministic match of two previously separate user IDs, e.g. two cookies with the same e-mail address/hash.

For customer data, the identify graph is deterministic, using the client’s login data. For activation across devices, Velocidi can also support probabilistic matching, using ID graphs powered by third-parties, contracted by the client.

Velocidi has an interface to match multiple IDs, e.g. e-mail hash plus a cookie (which comes automatically with the website tag), an e-mail hash plus a mobile ad ID, etc. The use of this interface is open for the client to apply in any existing channel, where it is possible to observe or capture multiple user identifiers. For example, website login can be used to link e-mail hash with CRM IDs with the first-party cookie of the browser.

The platform also supports creating separate user identity match configurations, to enable turning matching on and off from individual sources as well as reporting on them specifically.

It is updated in real-time as new matches occur.

This is customizable by the client, but an unlimited number of devices can be linked to a user ID. Some clients may choose to limit the number of IDs per user for data cleanliness’ sake.

All of your data is automatically stored in your AWS server. All permanent data storages are encrypted using AWS state of the art encryption, supported by the AWS Key Management Service for secure key management of all encryption keys.

The client is the owner of the infrastructure and all data in it. Velocidi’s access to the system is limited to operational purposes like updates or client support. No data will be moved out of the client’s infrastructure.

All Web traffic is protected by SSL encrypted connections, preventing any potential for data leakages along the way. And all data stores are encrypted with state of the art techniques, ensuring that your data will never be publicly exposed.

Velocidi records customer activity and identifies unique customers while minimizing the use of personally identifiable information(PII). Anonymized versions of user IDs are used within the CDP environment, and when packaging customer segments for external systems. The definition of PII varies by country and jurisdiction. Specifically under GDPR, any identifier, even randomly generated ones, such as a those stored in 1st party cookies, are considered PII. The difference is Velocidi keeps these identifiers under the brand’s complete control and doesn’t permit, by design, cross-relating these identifiers with each other (e.g. the same visitor on two different sites powered by velocidi will have two distinct cookie ids)

Admin-level users of the platform can manage user access to the platform. The CDP is installed with a dedicated Identity Provider (IDP). You can use the IDP interface to add and remove users from the system.

When a contract is terminated, the client will have the opportunity to export all their data beforehand. Upon termination the whole infrastructure and its data are destroyed.