Deepnote is a new kind of data science notebook and the favorite tool of data scientists all over the world, enabling them to clean data, write complex queries, build predictive models, and ship beautiful apps. Jupyter-compatible and running in the cloud, Deepnote allows data science teams to collaborate effectively with engineers and business analysts, and across the wider organization.
Deepnote’s ambition is to become the industry standard for notebooks, democratizing conversations around data within every organization.
Founded in 2019, Deepnote’s founders started out with a commitment to work with best-in-class tools right across their technology stack. They wanted to implement tools that could easily scale to support the company’s growth and to ensure they didn’t have any gaps in their capabilities which could become problematic further down the line.
Deepnote turned to Twilio Segment in August 2019 to centralize and standardize its data across all of its channels. The company initially started out on Twilio Segment’s starter program for early stage, fast-growth companies, before renewing and upgrading its contract.
In doing so, Deepnote was able to:
Create a single, unified view of the customer, based on accurate, high quality and compliant data
Enable marketing teams to quickly execute customer campaigns, without the need for any engineering support
Implement a data warehouse solution and introduce advanced analytics to inform its go-to-market strategy
A focus on customer data best practice from day one
As you would expect from a company focused on the data science industry, Deepnote was committed to building a robust, future-proof infrastructure from day one of operations. It wanted to avoid the trap that many early stage companies fall into of not getting the right processes and governance in place at the outset and inadvertently creating a patchwork of siloed tools and data sets.
The founders of the company knew that they would need to integrate a wide range of different analytics tools and communications platforms as the company scaled its operations and expanded internationally.
Deepnote started to explore options and quickly recognized the benefits that Twilio Segment could deliver, both in integrating customer data from multiple sources and in enabling more personalized communications by activating this data across downstream marketing tools.
“It soon became clear that no other tool offered anything like the complete coverage that Twilio Segment could deliver. We could immediately see how we could feed data into Twilio Segment from multiple sources and very quickly push that through downstream destinations, with absolute confidence in the accuracy of the data.”
- Robert Lacok, Product Manager, Deepnote
Centralizing customer data and optimizing customer communications
Deepnote brought in Twilio Segment to integrate all data sources into a single customer data platform (CDP), to track all event and behavioral data across its channels and to create audiences for marketing communications.
In terms of sources, Deepnote has one main application, split between back-end and front-end, and one main landing page on its website. Each of these have both a development staging environment and a live production environment. This results in a matrix of sources pushing different sets of data to different tools.
The company uses Mixpanel for product analytics and deploys Customer.io as its marketing automation and email tool. The developer team uses Sentry for error reporting, in order to quickly understand what proportion of users are being affected by any performance issues.
Critically, Deepnote has recently implemented BigQuery as its data warehouse. Prior to this it was only pushing data through to downstream analytics but the company recognized a need to strengthen its data infrastructure and create its own data warehouse. With Twilio Segment, Deepnote was able to realize this ambition with just a few clicks.
“Twilio Segment has enabled us to implement a data warehousing solution almost effortlessly. It’s been a real game changer for us as we suddenly have access to a wealth of historical, trackable data. We can go back and look at trends, such as revenue and retention, over a long period and this is having a huge impact on the way we approach our customer communications.”
- Robert Lacok, Product Manager, Deepnote
Twilio Segment is used widely across the business, within every function that works with customer data. In particular, the platform is used by engineers, product managers and growth product managers, who rely heavily on the integration with BigQuery, the data warehouse.
Data analysts then use Deepnote’s own tool to analyze the data in BigQuery and to create data models, visualizations and predictions.
Elsewhere, the marketing team rely on Twilio Segment data to push out regular communications through Customer.io, whether that’s onboarding information, newsletters or marketing campaigns that are triggered by specific customer activity.
Generating a deeper understanding of customers and increasing speed and agility
By implementing Twilio Segment, Deepnote has been able to generate deeper insight into its customers.
The company’s business is built on a bottom-up approach, getting as many data scientists as possible to sign up to its free-to-use service. It then looks to take these users on a journey to a point where they are ready to progress onto Deepnote’s paid-for enterprise plan, introducing the tool into their teams and companies.
Twilio Segment has enabled Deepnote to understand where new users are coming from, building a robust attribution model across all sources of traffic to inform its go-to-market strategy. Before somebody signs up, Deepnote can look at all of the events that preceded it, right across campaigns, Google and referrals.
Twilio Segment also allows Deepnote to identify which current users are most engaged and active on its application, and therefore which are most likely to be open to upgrading to the enterprise plan. This then triggers relevant communications from marketing and sales teams.
Twilio Segment was also instrumental in Deepnote’s recent ‘2021 Wrap’ campaign, which provided users with an overview of their activity during the year, such as how many projects they’d worked on and how many people had seen or engaged with their work. The campaign received hugely positive feedback and strong engagement from customers.
“At Deepnote, we prioritize speed over everything else because in the start-up world you need to be agile to react to new opportunities. Twilio Segment supports this completely, enabling us to filter through data, integrate new tools and roll out campaigns at speed. We consider Twilio Segment such a critical part of our infrastructure now that we always check whether new tools we’re looking to bring in have an integration with Twilio Segment. It’s a key criterion across the business as it means we save so much time and hassle onboarding new tools.”
- Robert Lacok, Product Manager, Deepnote
Twilio Segment is also being deployed to bring in more advanced analytics within the business. With some downstream tools, Deepnote has reached the limit in terms of how much data they can process but Twilio Segment, along with the new data warehouse, enables teams to work around these restrictions, joining up and filtering data to build use cases and create forecasts.
What’s next? Deepnote is exploring new ways to drive further business value from Twilio Segment. In particular, the company is looking at how it can optimize tracking within its own application to improve product analytics, not just at the user level, but also at a company level.
By generating insight on how many users within an organization are actively using its tool and what integrations they are carrying out, Deepnote will be able to analyze metrics at a company level. This will enable it to take a more data-driven approach to enterprise sales, with precise KPIs around the number of net new logos and revenue per company account.
Interested in hearing more about how Segment can help you?
Data instrumentation has been significantly simplified, dramatically reducing engineering costs on building and maintaining data pipelines.