By opting for the right CDP, your organization will become more resilient, enabling you to personalize the messaging throughout the customer journey. However, there are many factors, including AI and behavioral habits, as well as infrastructure, that can affect your decision.
In this article, we will explain what infrastructure a marketing technology team requires to sustain the challenges and be equipped to capitalize on new opportunities. This will blend four key AdTech principles that are applicable to MarTech in this new landscape.
We will first review the market of external challenges, such as cookies and OpenAI. Then, we will discuss the data platform landscape in relation to Salesforce by differentiating key features that are necessary to know to ensure you opt for the right solution. And to wrap up, we will address the key three themes that are crucial for every organization to consider in order to thrive (not just survive) within the current climate and uncertainties.
The Market
External Challenges
Not so long ago, Google (alongside other major browsers) announced the retirement of third-party cookies. Third-party cookies play a significant role in prospecting new visitors to your brand. They can impact your bottom line because of increase in return of ad spend (ROAS) and pressure on improving conversions across the funnel.
Since this announcement, additional cookie blockers have been gradually introduced by other platforms across mobile Application Tracking System that request cookie consent upon app usage. People are more aware of cookies and have a heightened concern over how their data is being used.
Alongside this, we have now officially landed in the AI-era taking digital and data concerns to another level. AI has boomed over the last couple of years, with companies like Salesforce adapting AI into their infrastructure with their Einstein products.
Up until now, AI has commonly been used for discovering insights on your existing first-party data and second-party data to improve actionable insights across a business. However, with the sudden emergence of ChatGPT in early 2023 and awareness of OpenAI has got the world talking about this (literally!).
This has also proven to be a divisive disruptor in every industry to its rapid growth, risks to employment and education but also how and where the data is collected from and given credit too. Organizations have raced to adopt ChatGPT into the platform offering, with Salesforce announcing the release of Einstein GPT only a month after ChatGPT emerged publicly.
Enterprise Impact
Organizations are now faced with the growing uphill battle of constantly reacting to the emergence of multiple data sources/channels, while also reviewing and implementing AI/Machine Learning (ML) capabilities to improve their overall activities. Although, this could mean that privacy management is somewhat of secondary thought.
However, privacy and visibility into how data is being used is extremely important to building trust with your market. People were already frightened by retargeting adverts sourced from third-party cookies – AI adoption will potentially take this concern to another level.
With this AI-fueled worry and third-party cookies disappearing, there is more pressure for organizations to refocus their efforts on capitalizing on their zero-party (declared) and first-party (implicit) data to improve overall customer experience.
Market Response
With inflation and cost of living seeming to consistently grow every year, alongside emerging constraints, organizations need to change how they target, personalize, and measure their activities. The principles from the AdTech space are necessary in the MarTech spaxe, now more than ever. AdTech is taking a big hit and it is now up to your MarTech stack to fill the gap and operate in a similar non-wasteful approach.
In the AdTech space, the following terms are used by advertising media practitioners:
Targeting / Personalisation:
- Granularity: Gaining a detailed view of your target profile.
- Aggregation: Using affinities, segments, and user cohorts to target a group (AI and ML).
The AdTech space is moving from a granularity lens into an aggregated lens in which using AI and ML will become the saving grace in the transition from third-party cookies.
Measurement / Accuracy:
- Deterministic: The accuracy of information collected by a known identified user.
- Probabilistic: Propensity to predict an action based on a user’s past engagement to determine the future actions with a probability model (AI and ML).
The AdTech space is needing to move in a probabilistic lens due to cookies being blocked across web and mobile devices (setting must be manually enabled to track cookies).
So why is this important in the MarTech space?
With such limitations being imposed and new challenges of maintaining multiple data sources/channels, organizations need to embrace all four of the approaches above into their MarTech capabilities. This will allow them to maximize overall targeting, personalization, and measurement across the organization, including your AdTech.
Debunking the Data Platform Landscape
At this point, it’s likely you would have come across the abbreviation CDP meaning Customer Data Platform. But you may have questions about what a CDP is, don’t all platforms unify data in some way? Technically, yes.
First, let us briefly tackle some potential use cases which all ingest, unify, and activate:
- DMP (Data Management Platform): Data retention is short-term, and the use case is more tailored towards the AdTech space to improve sharing of data for granular targeting. This is typically managed by the AdTech Media team.
- CDP (Customer Data Platform): Data retention is long-term and the use case is more tailored towards allowing cross-functions to be able to create segments from a unified profile record to activate across different function channels/activities to personalize a customer’s experience with your brand. This is typically managed by the MarTech team.
- Data Warehouse x BI: Is a storage location that contains a high volume of data with a fully processed structured relational database. This is typically used for Business Intelligence (BI) platforms to ingest data for reporting visual insights of an organization. This is typically managed by the Data Engineering team.
- Data Lake: Is a storage location where all types and sources of non-processed data is ingested and stored within. Data is not 100% structured within a relational database. This is typically managed by the Data Scientist team.
All these platforms have a need within an organization depending on the use case. However, from a MarTech perspective, a Customer Data Platform is needed to effectively unify messaging and customer experience across organization customer-facing functions.
Customer Data Platform Types
Now we have clarified the data platform types, we then enter the subcategory of customer data platforms. As of now, there are three types of CDP categories, Enterprise, Event-Based and Real-Time Personalization.
These CDPs can contain most of the following capabilities:
- Audience unification of a single customer view.
- Unified profile segmentation.
- Predictive modeling (analytics and/or content/product recommendations).
- ID management (multiple ID stitching).
- Collection of zero-party data, first-party data, and personal identifiable information (PII).
- Activates data with other platforms.
- Handles data in real-time or near real-time.
Below outlines the slight differences. “Real CDP” only refers to the right CDP your organization needs.
Type | Function / Need | Salesforce Solution |
---|---|---|
Enterprise CDPs | Large volume and velocity of data needed to be accessed and activated across the enterprise (brands, functions). Near real-time in data accessibility and allows you to create segments from a unified profile easily with clicks not code. Improves overall operations when engaging with a customer and the customer experience. | Data Cloud (Salesforce CDP), powered by Genie. |
Event-Based CDPs | More of a technical need to improve tag-based solutions and ingest large volumes of real-time event data to activate across different channels and platforms. | None owned by Salesforce. However, many Event-Based CDPs can integrate directly with Salesforce, such as Tealium AudienceStream / EventStream. |
Real-Time Personalization CDPs | A CDP that is built for real-time interaction. This management system targets, learns, and segments real-time personalisation across web, marketing, sales, and service (customer-facing) functions. Allows the ability to view all interactions over time, including when visitors are anonymous, enabling organizations to have deep insights to a person’s engagement. Improves overall customer experience with personalization and relevancy from recent interactions. | Salesforce Marketing Cloud Personalization (Interaction Studio). |
Salesforce’s CDP Platform Key Features
Think back to the four principles highlighted within the AdTech space earlier in the article. We will use these to explain key features of Salesforce CDP solutions:
- Targeting / Personalisation: Granularity, Aggregation
- Measurement / Accuracy: Deterministic, Probabilistic
Data Cloud (Salesforce CDP)
Area | Principle | Features (not limited to*) |
---|---|---|
Targeting / Personalization: Granularity | Gaining a detailed view of your target profile. | Data ingestion from multiple sources, including AWS Data Lake data | Batch and streaming real-time data | Data normalization techniques (Streaming Transforms, Batch Transforms, Formulas) | Segmentation filters. |
Targeting / Personalization: Aggregation | Using affinities, segments, and user cohorts to target a group (AI and ML). | Harmonize data-to-data model | Calculated insights (eg. propensity to spend, LTV, etc) | Real-time Streaming Insights | Segmentation | Activation of segments | Data Action targets can drive event-based orchestration | (pilot) Lookalike Segments. |
Measurement / Accuracy: Deterministic | Accuracy of information collected by a known identified user. | Profile Unification (a unified profile created only from anonymous source profiles is considered an anonymous unified profile) | Identity Resolution Rulesets link multiple sources of data into a unified profile. |
Measurement / Accuracy: Probabilistic | Propensity to predict an action based on a user’s past engagement to determine the future actions with a probability model (AI and ML). | Leveraging Einstein Discovery AI / ML features on CDP data to create probability and propensity models to support data-driven marketing and feed back into central CDP data. There are a few features in pilot: Bring Your Own AI/ML Model – connect CDP data in Snowflake to AWS Sagemaker/Databricks | Customizable | Propensity Scores (best for Website/Email Engagement or Loyalty data) | Segment Lookalikes API (AI finds users like a provided segment) | Rapid Segments (hourly activations based on a 7-day window) | Audience Insights in Meta/Google Activation Target in Advertising tier of Data Cloud. |
Marketing Cloud Personalization (Interaction Studio)
Area | Principle | Features (not limited to*) |
---|---|---|
Targeting / Personalization: Granularity | Gaining a detailed view of your target profile. | Unified Customer Profile | Unified Account Profile (B2B only) | Places a first-party cookie on the user’s browser which then tracks their interactions, regardless if not known, and collects overtime to create profiles. |
Targeting / Personalization: Aggregation | Using affinities, segments, and user cohorts to target a group (AI and ML). | Einstein Recipes leverage all the data captured and stored in the Unified Customer Profile with defined targeting. Recipes consist of a set of key attributes to listen for, which is then targeted with the appropriate messaging. |
Measurement / Accuracy: Deterministic | Accuracy of information collected by a known identified user. | Reports, including Einstein Reports, store all the rich customer insights which can be used to shape future campaigns. |
Measurement / Accuracy: Probabilistic | Propensity to predict an action based on a user’s past engagement to determine the future actions with a probability model (AI and ML). | Einstein Decisions use existing data to make real-time personalisation decisions on what content/messaging is going to be most relevant to users. |
Final Thoughts
In summary, collection and orchestration of data to overall improve your business activities and customer experience will place your organization in a good stead to tackle ever-changing challenges. This will leave you with a solid infrastructure to manage and govern data for advanced data-driven ways of operating across functions.
By opting for the right CDP, you can improve your organization’s business resilience and be able to deliver greater customer experiences. This is both online and offline by using the customer journey to provide experiential personalization messaging.
An organization with a solid infrastructure to collect, manage, and active data will naturally lead to stronger enterprise productivity due to increase of data collaboration, hyper-automation enabled, and overall digital scalability. With the right CDP, your organization will be ready to centralize and always adhere to your customers’ privacy concerns and access to their data easily if requested.
And finally, the icing on the cake, using the right CDP and applying Business Intelligence with AI/ML will allow your organization to lead business decisions with insight and intelligence. Enabling self-serve analytics/visuals across functions will enable you to engage with data better, whilst leveraging AI/ML data discovery for new opportunities and risk averting activities where needed.