As a consulting architect, I have participated in multiple data and analytics strategy initiatives. One of the tasks in such strategic initiatives is analyzing the current state of an enterprise or organization or business unit. We have to put various questions to the business and IT leaders to extract information on the current state.
As a consulting architect, below are some of the analytics questions you can ask:
1. Spreadsheets Usage – how many scenarios or existing reports do you use spreadsheets for analytics
2. Spreadsheets Purpose – do you use spreadsheets to build dashboards or do you any spreadsheet-native predictive analytics tools
3. Spreadsheets Data Freshness – how frequently do you plan to update the spreadsheets and what is the refresh mechanism? Do you generate a new spreadsheet every time you need fresh data OR you can use database connectivity plugins in the spreadsheet?
4. Dashboards – how many reports do you have currently for structured data and unstructured data?
5. Data Freshness – how many reports are built using streaming data and what is the expected latency?
6. Analytics Types – how many analytics are based on predictive, machine learning, Natural Language Processing
7. Analytics Tools – what kind of tools do you have to perform analytics on the data? Cloud Native tools vs on-premise tools vs third-party agency analytics solutions vs third-party analytics results vs BI tools native analytics capabilities
8. Proof of Concept vs Production Deployment vs Production Ready – do you have any roster on at what maturity level (PoC, Production Ready, Production Deployed) your analytics models or analytics reports are?
9. Business Functions to Analytics Mapping – for what business functions do you build advanced analytics (predictive, machine learning, NLP) based. Business Functions include Operations, Marketing, Finance, HR, Sales, R&D, IT, Customer Support, Product Development
10. Analytics to Skills Mapping – Who owns the design and delivery of analytics? Data Scientists OR Business Owners OR Data Scientists as a horizontal organization supporting the Business Vertical OR Data Scientists as part of a business vertical OR Third-Party Company OR Consulting Partners as Implementation Partners OR Business Analysts OR Data Scientists and Business Analysts
11. Analytics Yearly Projects and Budgeting – Who identifies the analytics project needs, plan the budget, manage, and controls the budget? Is it business function vertically OR analytics center of excellence or Directory of Analytics
12. Analytics for Regulatory vs Trends vs Innovation – how many analytics reports you are mandated to build and submit as part of regulatory requirements? How many analytics are in use for trend identification and new business opportunities and business innovation?
13. DevOps vs DataOps vs MLOps – what xxxOps practices you have established currently and how do you monitor those practices are followed?
14. Analytics Automation vs Analytics Designed – what analytics automation tools do you currently use (i.e. AutoML tools)? how many analytics are designed by your team but you expected them to be available in the AutoML tools (if you are using any of the analytics automation tools)
15. Analytics Data Source Types – what % of each of the following data source types do you use for your analytics?
Transactional Data; Image Data; Video Data; Logs Data; User Search Criteria data; Text Data; Structured Data; Unstructured Data (other than images); Real-Time Data; Historical Data; External partners data; third-party datasets; Customer Demographics Data; Product Ratings and Product Comments Data; Social Media data; Company News Data; Sensors Data; Geospatial Data; Audio Data;
16. Source Data Quality for Analytics Processing – how do you assess the source data quality before embarking on the analytics processing? Who confirms the source data quality? What automation tools or practices you have established to reject analytics processing when source data quality is bad? Who defines the threshold for the source data quality?
17. Source Data Usage and Availability – how many data sources do you have and how many of them are in use in your analytics processing? How frequently the source data is refreshed? Where do you store source data? Can you regenerate the same analytics in the future using the same old data? How many years old of data do you store? What are your archive solutions?
18. Security – how do you know PII data elements? how do you govern executing and viewing the analytics within the organization? who sets the security standards (access, PII transformations, masking, etc.) for your analytics?
19. Challenges – what challenges do you have in analytics adoption?
Budget Approvals, Data Quality, Talent, Data Literacy across the organization, Integrating multiple data sources across the business verticals, Technology availability, Regulatory and Compliance, Data and analytics-driven decision-making culture, operationalizing the analytics i.e. beyond PoC or production deployments of the analytics from lower pre-production environments, DevOps/DataOps/AIOps/MLOps practices
20. Value-focused questions – how to rate the success of your analytics initiatives?
Customer Churnout, Customer Satisfaction in Product/Service delivery, Loyalty Customers Satisfaction, ROI targets, top-line impact (growth), bottom line impact (less operational cost, more profits)