You need to learn to see user requirements as an end-to-end solution and work to understand exactly what ROI can be driven. Lots of companies these day build huge big data lakes, but they don’t think to understand what value the data has. It’s better to have a little data of excellent quality, rather than billions of records of 20% quality data.

Aim for what will make a difference, and the easist dfference to make is saving money or making money.

The post Go for the ‘Dough’ and not just ‘what’s nice to know’ appeared first on KieranPoynton.com.

If you ever hear “oh, we didn’t touch it because it’s always worked”, it can certainly be improved.

The real skill is knowing when to stop developing.. and just release. Aim to satisfy your customers and your timelines. Perfection comes with constant feedback and small alterations to an existing product.

Aim to deliver 80% of the work in 20% of the time. Release and improve. #agile #bi #analytics #timelines

The post Be Curious and improve daily appeared first on KieranPoynton.com.

 What’s your “if only?”

• If only I knew what the average stock cover days I have on all products, then I could manage stock more effectively.
• If only I knew how often products were often reduced out of policy, I could manage markdowns and risk.
• If only I knew how often machines failed, I could send engineers to check machines earlier.

“So what?”

A few years later it became apparent that some requests did not provide the business impact I was expecting. I started to realize that some users would just ask for data or reports without thinking about possible business value based on the solution delivered. So, to ensure requirements and demand were gathered correctly, I started to ask my users a different question: “So what?”

Example:
Person A “I want a report showing me our top 10 products.”
Person B “So what?”
Person A “So I can understand where we make all of our money and where we need to prioritize.”
Person B “So what?”
Person A “So I can ensure that these products are always in-stock in every major store”
Person B “OK … what if I supplied the following?”:

• Information report showing top 10 products sold
• Exception report showing stores that run out of this stock along with an estimated loss calculation
• Advanced analytics to work out average stock cover and automatically order new products from suppliers based on the top 10 current products sold.

“Have you thought about x?”, “Explain how that would support your process”

Due to the complexity of data in businesses today, it’s now even more important to ensure you really understand your users (and that your users understand the data). The “If only?”or “So what?” questions don’t narrow down the use case enough or provide enough detail to form an agreement between business and analytics teams.

I now try to use coaching questions (often the same questions that I pose to my team), these start with: “Have you thought about x?” , “Do you know how this would impact the business?” , “What can you learn from this?” , “Is there any other data that would enable you to make a faster and more confident decision?”, “Are there any political blockers that you might face”?. These questions help the users to really think about their request and already do a lot of the ground-work in their head. You can validate the request together and ensure that the final decision does create value and business impact.

I’m interested to know if anyone else has any other methods of gathering demand/requests from your users and how you ensure success?

The post Gathering User Requirements: “If only”​ and “So what?”​ appeared first on KieranPoynton.com.