Right now for many organizations, self-service analytics is a wasted opportunity. At one time, the idea of self-service analytics was attractive and people believed in its potential. But, somewhere over time, companies went off track. The promise and value exchange between data creators in the center and data self-servers out in the business was never actually fully negotiated (or fulfilled, for that matter). For true success, teams need to align and work together from the beginning in order to turn missed opportunities into untapped opportunities with data.
However, because of varying organizational maturity, there is no true one-size-fits-all for self- service analytics. So, to avoid making the concept something that is (even more) daunting to or wholly avoided by business teams, this blog outlines some of those nuances (many of which stem from semantics) and highlights the future trajectory of self-service analytics for data executives.
Let’s say someone on your central data team is struggling to service a lot of requests and prioritize the highest value work. The idea that they can get other people involved is very attractive. So, for IT-driven buying and thinking, it’s a huge win. The business users, though, need to find the time to do the work themselves without the expertise of the data experts.
How Can Your Teams Move From Traditional Self-Service to Self-Service in the Realm of AI?
It’s less about understanding the past or predicting the future, and more about empowering people — because we’ve reached a point where people (such as an analyst or someone on the business side who is doing much more with data than ever before) want to drive additional insights. Dataiku makes all of that possible and connects doers with data by:
- Bringing people of diverse skill sets together to work with data in a common ground
- Providing a remotely accessible platform that provides a centralized place to find data and previously built datasets and projects by other colleagues
- Enabling the users of data to make more of the consequential decisions about their data by encouraging them to help build the data project from the get-go
- Giving business users autonomy when it comes to creating said data projects (via best-in-class data exploration, data access, and data transformation abilities — all without having to learn to code or put additional strain on IT)
- Granting users automation capabilities (in the cases where an insight needs to be created repeatedly) to update the insight every hour/day/week without the need to stop their current day-to-day tasks
- Making it second nature to productize new insights and data preparation methods (in a way that is easily reusable) so that all business users will benefit from the expertise of the many as timely new data products are created and maintained across the whole business
GE Aviation provides a compelling example of AI-driven self-service through their optimization of the self-service system, utilizing real-time data at scale. This strategic approach facilitates quicker and more informed decision-making throughout the organization.
Don’t Forget About Governance
It’s important that these guardrails for data quality and governance aren’t overlooked. Taking another example, the FP&A team at Standard Chartered Bank has developed their own brand of self-service analytics with Dataiku. The idea isn’t that individuals can do and build whatever they want with data (which would lead to data chaos), but rather that a center of excellence owns the core structured intelligence of the bank. This means having enterprise-level data connected to a homogenous pool, with designated product owners for each dataset and clearly defined governance that connects to the entire organization.
From there, the team builds specific experiences on top of that data that can deliver answers through core apps, and the ultimate self-serve flexibility comes from how people around the organization use those apps to solve business problems day to day. On top of it all, Dataiku enables all of this work through visually legible data pipelines and not on desktops, which helps IT teams sleep better at night.
Achieving Time Savings and Scalability Through Unified Projects
Further, when it comes to time savings and efficiency, analysts and line of business managers can capitalize on their existing BI data pipelines and turn them into AI data pipelines by creating AI data products. To transition from building BI data products to AI data products, staff can reuse the data assets and infrastructure they’ve already built and understand well, shorten the learning curve by leveraging an identical user experience in the data pipeline, and scale faster and more economically by doing both BI and AI projects in one environment.
Now, while BI and AI analytics can certainly build off each other, it’s only part of the story. It goes beyond using data to just prove what you already know. For Dataiku customer Rabobank, one of the biggest advantages of Dataiku is that data projects progress. Often, the business starts out with a simple insights question, but those insights then lead to new initiatives. For example, if you know certain customers have certain risks, that starts out as a dashboard but quickly gets into predictive analytics and machine learning.
If you start with a BI tool, then you have to do all kinds of work to set up a new environment once the project progresses. Dataiku allows us to start out with relatively simple insights questions and grow toward a more specific predictive question, developing a model all in the same tool.
-Roel Dirks, Product Manager Big Data Lab, Rabobank
Why We Call All of This Everyday AI
There are people in the business who have the ambition to go on their own data journey and will do it if the points of friction are reduced and they are enabled to do so. And for those on the business side who aren’t ready to jump into AI, they can start somewhere and work their way up to it — but they might as well do it in one place that’s collaborative and future proof, like Dataiku.
When people of different skill sets and backgrounds start using Dataiku, they get inspired. For example:
- Unilever designed a responsible, self-service tool for natural language processing.
Unilever’s People Data Centre (PDC) teams across the globe deal with vast amounts of unstructured text data on a daily basis to gain insight into their customers. The answers that the company’s marketers, product research and development, and supply chain specialists need also require analytics approaches tailored to the business. Data scientists and software engineers in PDC built a range of NLP methodologies, including a plugin. With Dataiku, they can easily assess how and where people across the company use it across different research projects. - Westpac has created a collaborative, self-service operating model to upskill and drive a closer alignment between the business and tech teams.
Westpac developed a new operating model and new processes to ensure a strong alignment between the Discovery Lab (consisting of technical players) and the business teams, while enabling them to broaden their understanding and gain new data skills throughout the project. When a new team member joins, they attend an orientation session so they understand how they can leverage Dataiku and go through an assessment to further define their use case, the objectives and expected impact, and how they visualize the outputs and outcomes of the initiative. The tech team is enabled to serve the business, help them upskill, and work together to drive innovation at scale. - Standard Chartered Bank transformed its FP&A division through Dataiku, achieving an extraordinary 30-fold increase in analyst productivity.
Previously constrained by traditional spreadsheets, Standard Chartered Bank struggled to manage large volumes of data effectively. However, with Dataiku's self-service capabilities, just two individuals now accomplish what once required 70, streamlining operations and boosting productivity significantly. Additionally, Dataiku democratized data access across 12 bank communities, enabling collaboration and unlocking substantial business value.
At the end of the day, Everyday AI brings down the temperature of self-service analytics and makes it more consumable. It involves enabling cross-functional teams and people with diverse skill sets and levels to work together, understand what each other is working on, understand what has been published by whom, and how they can rebuild data products on top of what already exists.