line graph, bar graph, two women looking at data on a computer, large group of people walking and wearing masks in a city, and two males a manager and frontline employee checking on products in a warehouse
January 19, 2021
Data Intelligence

Data Delivery with Blaine, Former CEO @ Panera

Last week, I virtually sat down with Blaine Hurst, former CEO and now Vice Chairman at Panera (among a long list of previous leadership roles). 


We talked about data for almost an hour. Selfishly, I loved the chance to nerd out with an industry veteran, and we uncovered some really good “nuggets” from the discussion that I wanted to share.


But before we dive in, you may be wondering what successful data delivery looks like? It’s getting the right insights to the right people at the right time in the right format. 


That’s clearly easier said than done. So here are a few highlights to help you get there. 


It all starts with alignment across tech and the business.


If there’s one pattern we see across just about every client we work with at RevUnit, it’s this. We often come in to tackle a new data initiative or product build and find that there is a lack of understanding between tech and the business. Not gonna lie, this is often the very reason we are brought in to help: the business unit is struggling to find success because their needs are buried deep in the backlog of IT, or they’ve simply lost trust in tech delivering a successful solution. So they look outside of their organization.


Blaine has felt this first-hand, too. “So often we approach this as ‘I'm the technology god and I'm going to tell people how to do it’ or ‘I'm the business person and I know everything, and the tech guys are just idiots because they haven't delivered for me.’ I've heard both. I've probably said both.” 


It’s easy to let silos keep you from creating buy-in on a strategy, but without understanding (and input) from both groups, you are setting yourself up for a rough journey ahead. So the first step is forming a data strategy that both business and tech are aligned too. That means bringing key stakeholders and end-users of the data together to make sure your strategy is practical at every level, taking into consideration the data structure and how it will be used in reality.


“Both of those are critical: how we organize the data itself and then how do we plan to use that data,” Blaine said. “Without that, I don't think you have a great data strategy.”


Observe your data in its real-world context before determining the structure.


Blaine summed this up nicely when he said, “Understanding data as it exists is step number one. And I think we've missed that with some of the newer techniques around data. I cannot tell you how many schema-level discussions I've had, even in the last six months, and I'm looking at their data going, ‘What were you thinking?’ because they didn't reflect how the data actually existed. In fact, there wasn't a comprehension of it. It's a mistake if you don't start with an understanding of how data exists before you go build a version of it.


That matches what we're seeing with our enterprise clients. As SaaS, ERP systems, and custom enterprise solutions have increased, organizations now have a range of different data sources to grapple with. Not only is it important to observe the business, but you then have to figure out how to marry up the data from all of these different sources to reflect reality. 


Everyone likes to geek out on all the technology stacks and all the new things you can do (myself included), but in many cases you need to first work at the data modeling level to get it right there before you can do anything else.



Democratizing your data is bull@*%.


A pretty hot take from Blaine, huh? 


But it’s true, this is a buzzword in the industry right now, and just like most of these hyped concepts, it’s not very practical. Many hope that democratizing their data will solve all of their data problems. But that is simply not true.


We already spend way too much time analyzing data that doesn’t matter — so allowing anyone across an organization access to data is only deepening that problem. 


Democratizing data for analysts across the organization is a good idea because they need it to do their jobs. But it’s their responsibility to sort through that data and pull out the key indicators that everyone else needs to take action on. 


As Blaine put it, “I don’t need all this sh*t. I just need to know what I can take action on now.”


Which leads me to our next highlight:


Getting to the data that matters most starts with key metrics.


With so much data available (and that amount only growing), pretty much every enterprise is struggling with sorting through the sheer amount of data to get to what actually matters. To try and tackle the tidal waves of data is overwhelming, so instead, work backwards from the use of your data in the wild.


Start with the key metrics that are most important to the decision maker, then find core drivers of that metric through correlation. 


An example Blaine shared as we prepared for our discussion was customer warmth at Panera. A key metric that was paramount to the success of any restaurant was the in-store customer experience, which they call customer warmth. After analyzing data from restaurants, they found through correlation that two metrics were the most responsible for success: tenure of staff and the amount of staffing. These two inputs would reliably lead to improved customer warmth AND were very controllable inputs by the restaurant managers. 


Managers on the front line didn’t need extensive reports with tons of data in order to improve their store’s success, they just needed to keep tabs on these two metrics. 


Showcase trends, exceptions, and patterns to decision-makers when they need to take action (even on the front line). 


Even if you are able to determine and pull out actionable insights, they won’t affect the success of your business if you don’t deliver them at the right time.  


Blaine talked about three key metrics that are key to showcase:

 “It's ‘how am I trending?’ It's exceptions, ‘Am I above or below threshold?’ And it's pattern matching...But if that trend I spot is a week later or a period later, what am I going to do about it?”


In general, we need to remove the need for periodic reporting and replace with actionable insights given at the time of need. Blaine gave a great example.


“What I need to know is today, right now. I'm doing lunch and that salad station is slowing down and not keeping up with targeted throughput. Why? Because that's what's driving my line times and making them worse. I need to know that right this second because that's when I can take action on it. When I need it isn't generally after the fact, it's while I am actually in the middle of the process, and I can do something about it.”

___

Getting the right data to the right people at the right time is arguably the most important factor in adapting your business to any environment. And if we learned anything business-wise from 2020, it’s that adaptability is critical


Good luck in your data journey. If you want to chat more about any of these ideas, just send me a message here on LinkedIn or email me at: michael@revunit.com.


Watch the full recording of Blaine and Michael’s discussion here.


Data Studio
Google Cloud Build
App Engine
Cloud SQL
Cloud Storage
Cloud Functions
BigQuery ML

Jumpstart your data innovation with repeatable frameworks and attainable ROI (for real)

Diverse team in front of a computer working together, innovating

Using startup frameworks, we can help you leverage your data to drive new revenue opportunities, innovations in technology, or even new business models. Learn more about our Innovative Data Services.

Tell me more →
RevUnit icon
WRITTEN BY
Michael Paladino
Co-Founder and CEO at RevUnit, Michael leads the company's efforts to stay on the cutting edge of emerging technologies. His current areas of focus include AI, machine learning and conversational interfaces which can all be used to help RevUnit clients Work Better.

Related Posts