A large number of products come out in the data & analytics space every so often. We read about the Analyst reports, new features, and VC fundings. Long & colorful white papers, detailed comparisons with other products ensue. Companies conduct webinars to spread the word.
But, wouldn't it be beneficial rather understand the challenge the prospect customers face? What are the business ramifications of those challenges?
How product or solution set overcomes it (if at all)?
Finally, what is the business benefit?
The purpose of a platform-led solution is to shorten the business value creation time.
Here are few commandments to succeed with a platform-led solution :
- Think business backward, not data forward.
- Prove hypothesis, Fast.
- Do not boil the ocean. Look for quick wins. Deliver business value.
- Get more cheerleaders among the leadership.
- Do not focus on engineering efforts alone. Focus on end-to-end use cases.
- Decentralize decision-making.
- Make security and compliance part of the culture.
With those commandments at the helm, let us deep-dive into Snowflake's data cloud.
I will walk you through few value propositions.
Each of them covers an enterprise challenge, feature & the business value play.
1. Thinking beyond Data Silos
- A large amount of data in enterprises stays in Silos.
- For CDPs, it is the main plot for unifying data.
- Bring everything together.
- This logic led to Data warehousing solutions.
- Data lakes took that further and added semi & unstructured data types in the mix.
- Here is a typical enterprise data silo example:
- Inventory data lives in WMS(warehouse mgmt).
- Finance, Planning data lives within ERP.
- Customer data lives within the CRM systems.
And so on.
Then, Some systems have data overlap.
They become partial sources of truth with redundant data.
These data silos & systems turn into behemoths, getting beefier capturing data for years. Then, you can not leave.
It becomes impossible to apply Data Standardization, Quality, Cleansing, Governance to these silos.
Data Cloud focuses on breaking the data silos. With data being at the center, it enables seamless sharing across the organization.
Data Encryption, Role-based Access controls, and Multi-clustered shared data architecture help rid silos. Business still Security, seamless Access, and Governance around data.
Separation of Storage and Compute enables infinite scale, without one affecting the other.
For business users, it enables data discovery across the organization. They can connect dots with data dimensions that were not visible before. Using the external data sources is seamless as well.
Data cloud doesn't need expertise in hyper scaler’s proprietary.
It is a business user, business analyst-centric offering.
The SQL skilled analyst can make full use of the platform and discover data and hypotheses.
2. The rise of Multi-Cloud juggernauts
With the growth & omnipresence of hyper scalers, It has become too easy to spin up a database or a compute cluster. Setting up a data mart is a walk in a park. It means it is too easy to create a new data silo in the cloud.
Think about it!
IDC's Research says - IT managers would like to use at least 2 Cloud providers as part of their strategy.
They can hedge their bets. They can ensure optimized costs, resilience, and meet any regulatory and compliance requirements. This aggravates the data silos problem. They also have to deal with the learning curve across hyper scalers.
These hyper scalers are the backbone of the data cloud. They are completely abstracted yet. The end customer defines the underlying cloud provider and regional requirements. Data cloud takes care of the Rest - Provisioning, Control panel, Billing, and management.
Business users can build resilience using hyper scalers and their presence across regions and zones. Sharing a dataset within their organization across countries or cloud providers is seamless. Data redundancy is not required. It also takes care of Data residency requirements.
3. Bump up the business value
We are creating data at an unprecedented rate. This data spans geographical boundaries, on-premise, cloud, types of workloads, and data types. 75-80% of the data generated is of semi-structured and unstructured type. It requires Complex ingestion, Data cleansing, Governance efforts to crunch such data. Specialized tools required are expensive, time-consuming, and need customizations.
Engineering efforts often become more important than the value it deems to deliver. I have seen this happen firsthand in many customer situations.
Engineering complexities take over the initiative.
With data cloud, the focus stays on the end goal.
It abstracts engineering efforts. It has built-in connectors for AWS, Azure, GCP. It has built-in tools for data governance, Security, and Compliance. This ensures the enterprise focuses on the hypothesis, outcome, and not on engineering decisions.
Data Cloud takes away the heavy lifting of engineering efforts from the enterprises.
It saves enterprises precious time and effort.
It is simple to use compared to the last-gen platforms & databases.
4. Problem with wasted resources
Building data initiatives require ingesting data, processing, different domain views for various purposes. ( Reporting, various use cases including data science, data warehouse).
Central to it is the wastage of resources.
Every zone - RAW, Standardized, Processed is a copy of data. A little different from others.
When needing to share data, copying of the data is the most prevalent solution. It immediately makes the data stale ( Data is not fresh if not real-time).
Think about it. Copying and sharing of data mean you have to:
Consider data security - Who is going to access it? Are they authorized to do so? Its Storage?
You would need to repeat the same activities If the other party will enrich the data & share it back.
With hyper scalers sharing buckets is easy but it creates copies of data, each a little different. Moreover, no organization gives access to the actual bucket. They will make a copy and share a bucket ensuring security aspects. It is a management nightmare.
Data cloud simplifies real-time data sharing.
There is no data movement.
No need to make copies.
It has built-in access Controls which are simple & granular.
5. Compliance, Governance, and Security
Once past ROI, the next challenge lies around Security & Compliance. An organization must stay compliant with its industry requirements and keep things secure. It is a non-negotiable need, irrespective of the size of the company or the type of data platform - SaaS, PaaS, etc.
Things were simpler on-premise.
Defended a perimeter and laying Security was simpler. Cloud has changed it for good.
Security becomes crucial with the multi-cloud, hybrid cloud, and distributed nature of our work.
The data platform architecture needs to follow complex compliance & security needs. They must cater to the growing data, its format, sensitivity, data residency requirements, and much more.
When architecting from the hyper scaler toolsets, meeting all requirements can be challenging. As Engineering teams evolve the product, One can hope it is scalable and meets security and compliance requirements.
Easier said than done.
Data Cloud ensures sensitive data security while providing seamless & real-time access governance. It applies to industries requiring tight governance requirements like Healthcare or Defense.
Data Cloud is HITRUST certified, HIPAA Compliant.
The Data cloud offers a modern solution built for the new-age business environment.
It offers agility.
It is architected for complex hybrid cloud requirements, compliance, and security needs.
It's an excellent choice if your organization focuses on business value, vs building a hand-crafted engineering solution.Saurabh Mittal July 14, 2021