Delivering personalized and relevant customer experiences is now the expectation, not the exception.
And year over year, these expectations are getting higher. According to our 2022 State of Personalization Report, 62% of consumers say a brand will lose their loyalty if they deliver a generic, unpersonalized experience, up from 45% in 2021.
In practice, however, delivering these personalized experiences is an empty rallying cry rather than a reality.
The problem is that gathering all the data about your customers and merging it into one record is a largely unsolved challenge for the vast majority of businesses. Stale, inaccurate, and fragmented customer profiles are still the norm, leading to impersonal (not to mention inefficient) customer engagement.
For over a decade, Twilio Segment has been on a mission to change this. With our industry leading product – Segment Profiles – businesses can merge the complete activity history of each customer across web, mobile, servers and systems into a single profile, giving companies complete, accurate and trusted customer profiles automatically – no data modeling or extra transformation required.
Now, we’re taking things one step further. Two upcoming features – Profiles Sync and Reverse ETL– make our industry-leading customer profiles portable, allowing businesses to create hyper-tailored audiences in their domain-of-choice, the data warehouse.
These features, alongside Twilio Segment’s other products, such as Engage, are the next step in our vision to empower businesses with real-time, accurate, and unified data together in one platform.
Since Segment’s founding in 2011, open data and interoperability has been at the heart of our approach to customer data management. Unlike legacy marketing suites, our open platform of 400+ pre-built integrations ensures customer data is not locked in any one system, and can be activated in real-time in your marketing and analytics tools of choice.
So, when we heard from some of our most data-savvy customers that they were performing advanced analytics, Machine Learning, and trait computation in their data warehouse, it got us thinking about what customer profiles in the warehouse looked like.
Using new modern data platforms, we saw 1-3 person data teams running powerful cloud data warehouse systems alongside their CDP to enable mission-critical analysis and impressive personalization use cases on par with big tech companies staffed with armies of data engineers.
But when it came to building customer profiles, we observed a recurring stumbling block.
These small but mighty data engineering teams were creating “shadow profiles”, custom-built identity resolution pipelines within the data warehouse itself.
Custom-built, identity-resolved customer profiles are very difficult to create and maintain in the data warehouse directly, sometimes taking years to model and refine to a point that they are trusted.
In some cases, data teams were able to stitch identities in their warehouse by performing transformations or working with additional vendors, but for many businesses these “shadow profiles” became incomplete, out of date, and less accurate than using Segment’s identity resolution system built over several years.
Additionally, more and more customers showed us how they were driving business impact with Machine Learning. To enable this, they wanted to send sources and data processed through Segment into the environments their data scientists already worked – the data warehouse.
