Talend has announced its Fall ’21 release. Key among the announcements is the focus on healthier data, something that organisations continue to struggle with. The parlous state of many organisations data means that it is often untrusted. Instead of executives making informed decisions, many are just winging it.
Krishna Tammana, CTO at Talend, said: “It’s no longer just about that first mile data movement from point A to point B. Simply moving any data into a business process, analytics environment, or machine learning model will not deliver successful business outcomes.
“Our focus is to help customers with the last mile in data management, ensuring they have healthy data when and where they need it, in the right context. When a business can build this into its systems and culture at scale, at the speed of business, is when it will consistently deliver successful business outcomes.
“Our new release builds upon the innovative quality and trust capabilities in our platform that has made Talend a leader over many years, enabling more organizations and people to benefit from healthy data.”
What is in Talend Fall ’21?
Talend has called out four highlights of the Fall ’21 release:
- Talend Data Quality Service gives business leaders a path to healthier data without technology or headcount investments. Data Quality Service leverages Talend’s market-leading data quality technology, frameworks, dashboards, and specialists to help companies monitor the quality of business-critical data continuously and achieve healthy data quality insights consistently.
- Native Trust Score for Snowflake profiles entire datasets inside Snowflake Data Cloud, ensuring data professionals can leverage healthy, analytics-ready data for their businesses.
- Self-service Data API sharing in Talend Data Fabric enables data engineers to run processes, applications, and analytics without relying on a developer because the new no-code product creates and shares APIs in minutes rather than following a process that typically takes days.
- An industry-first, Stitch Unlimited’s non-consumption-based pricing gives analysts predictable data ingestion costs and improved productivity. Now, analysts can focus less on consumption volumes that drive ingestion costs and more on getting value out of their data. Additionally, Stitch users can connect to multiple destinations from a single account.
A real lack of trust in the data
Fiorda began by addressing the issue of trust. Talking about the Talend survey from earlier this year, Fiorda said: “What we learned is that people simply don’t trust their data. Two of the bigger results are 78% of executives have challenges making data-driven decisions and 60% of executives don’t always trust the data they work with.”
Data health and cleanliness has been an issue for decades. Organisations spend vast sums cleaning and preparing data for projects but then fail to maintain it. For Fiorda, this is why Talend is focusing on the concept of data health. Fiorda continued: “It’s a concept that’s important for us to convey to our customers because it describes how well the company’s data supports business objectives.”
One challenge with data health is getting a business view of the data. Everyone is expected to be data-aware today. But that leads to conversations like “if the data can do this” or “this is the data structure we have”. What it doesn’t do is abstract the conversation to a non-technical view of the data that the business needs.
Fiorda said that Talend is focusing on connecting Chief Data Officers with their business counterparts. In doing so, they hope to get that wider set of requirements from the business. To establish trust in the data, Fiorda said there are three critical elements:
- Putting preventative measures in place: A good metaphor for that is just like with your health. You think about diet and exercise. Data needs to be maintained and managed similarly.
- Effective treatments: When something isn’t right, deal with it, especially around data transformation.
- Culture: How do people work with data? Don’t throw away data you’ve cleaned. There’s an element of data literacy that needs to be addressed.
Build an effective foundation with good governance
Fiorda believes that companies need to build a firm foundation for their data. He said: “Without data quality and data trust, data health is impossible.”
One area that is talked about a lot around data is governance. But how much of the conversation is about more than just avoiding fines or data breaches. Fiorda said: “Having a strong governance layer within this data health solution, especially as businesses become more data-driven, and data is being used in more and more places, becomes more and more critical.”
Fiorda also sees data intelligence as the key to good governance. For example, what data do you have? Where are you storing it? What data sources do you use? How are you using it? Who else has access to it? How are you sharing that data? What are the controls on data usage? What monitoring systems do you have?
To solve those issues, Fiorda believes any data platform must have intelligence in its platform. That intelligence includes the use of AI and machine learning to identify risks to data health.
Treat data as a team sport
One area where governance becomes difficult is data sharing. GPDR and other legislation require companies to do more to protect data. Yet, as Fiorda says, data sharing is critical, especially inside the organisation. But rather than rely on IT or data scientists, Fiorda sees self-service as a key part of the data value chain.
For self-service to work, however, Fiorda says you need a common platform. He said: “That means the UX and experiences that they use to interact with the data and the teams are in the context they can understand and work with. So this idea of context is very important to continue building a platform that treats data as a team sport.”
And this is where the governance piece comes back into play. It is about providing the guardrails to enable self-service users and protect the data. Talend deals with this through its data catalogue that allows organisations to set granular access to data. In a future release, Fiorda said: “We have roadmap items to automatically discover and classify data, and then do data shopping.”
The idea of internal data shopping is particularly interesting. Users will be able to choose what data they want to use in their projects. The question is, how far will Talend go? Will this include a catalogue of internal and external data? Can it restrict usage based on data sensitivity? Will location play a factor in what data can be accessed? Can users create data collections and add them to the catalogue? How will it enable users to enhance and enrich data?
Davidson says that the answer to the latter will be through APIs, including the self-service API, and users will create APIs in just a few clicks.
How does this all tie back to the Fall ’21 release?
Davidson also pointed out that Talend is running fast to build all of this out. With Fall ’21, the primary focus has been on setting out those four items from earlier. The first of those, the Talend Data Quality Service, is the foundation for better data health.
There are three parts to this:
- Scale data excellence: This is about operationalising data and data quality in new ways.
- Working smarter: Helping users prioritise data tied to their work. Enable them to prioritise their decision making and understand what they have going on.
- Accelerate productivity: Refining the ways customers get work done.
One of the challenges that all of this will address, according to Davidson, is helping companies get value out of the vast amounts of data they have. Many struggle to make that data work for them, so Talend Data Quality Service is being announced next week at Talend Connect.
Davidson also tied this back to the governance that Fiorda talked about. The first step is helping customers identify, manage and operationalise their critical data assets. After this, he says the governance and management such as “ongoing monitoring, reporting, addressing data drift, all those kinds of things” come in.
What isn’t clear is how many KPIs Talend has developed for the monitoring or whether its compliance team has contributed any. It would also be useful to know how many KPIs have been suggested by those customers on early access programmes.
Talend introducing better pricing
The pricing of cloud-based services has become a contentious issue for many customers. Computer and memory are charged at a very granular point, generally down to the hour or even below. Software and services still seem to be charged by the month, although there are some exceptions. Davidson said that with Stitch Unlimited, Talend is taking a new approach.
“Most of our customers are pricing on consumption. Talent is taking a different tack on consumption. It’s about incentivizing our customers to use their data rather than disincentivizing. Stitch Unlimited is effectively unlimited. It’s about taking that little bit of worry out of their mind and encouraging them to use the platform and their data more.”
But what does that mean, especially for a service that is part software, part hardware? Davidson insists that this is purely about being 100% predictable on the subscription. He said: “It’s subscription-based. You pay one price, and you’re not charged per row, per CPU, per connector. All of those are out the window.”
It sounds like a very positive move, but much will depend on the detail. What is the pricing per subscription? How many customers will increase the number of subscriptions they have? The latter will tell if customers like what Davidson is saying.
A big investment into Snowflake
Talend is also making a big investment into Snowflake with Fall ’21. Davidson says this is to support customers who are in the middle of modernising their environments. Davidson said: “One thing that Talend is doing is allowing Talend Trust Score inside Snowflake. It uses Snowflake processing and allows customers to use those Snowflake credits to scale out their trust score processing.”
Talend is also building out its Snowflake ELT processing. Davidson commented that this was again back to the self-service data API. He sees it as making life much easier for data analysts who: “Do all this work of getting their data together, organising and cleansing the data. What’s next for them? How are they going to enable other teams to use the data? How are they going to enable a lot of different use cases with the data?
“It might be app integration. It might just be generally ensuring that users are using the right data, the right data set. As a technical process, it can take days or even weeks to create an API. We are allowing data citizens to create no-code API’s. Click, click, click. Here’s the name of the API. Here is the security information. There’s no need for them to understand any of the complexity involved. They just click it, and it is created and deployed to a Talend processing engine.”
This is a major shift for internal data usage. It also answers many of the questions Enterprise Times raised about self-service and data shopping. But there are still issues to be worked out. For example, how do you verify the security of the API? Davidson sees this as an internal project usage. However, if it works as described, people will want to use it to share data with partners.
Enterprise Times: What does this mean?
In the call with Fiorda and Davidson, Enterprise Times could only scratch the surface of the Fall ’21 release. However, even that scratch showed that this is a significant release for Talend. Customers will welcome the promises of making data a team sport and improving data health. Just as importantly, customers will welcome the move to a better, more consistent pricing model.
There will be much more released by Talend around Fall ’21 in blogs, on the website and at Talend Connect. It will be interesting to see what else Talend has in this release.