Shan Newton is a former Senior Developer and Information Architect at American Airlines and is currently the Head of Business Intelligence at Pixelz. Over the last two years, Newton has been responsible for transforming and standardizing how data is requests, processed, analyzed, and communicated at Pixelz.

If your business is like most, you are tracking as much data about your operations as possible.

But just collecting data isn’t enough. How effectively an organization turns data into action is what seperates businesses that are as lean and efficient as possible from those that are plagued with unnecessary waste.

Over the last two years, Pixelz Business Intelligence Leader Shan Newton has been tasked with overhauling and standardizing how data is requests, processed, analyzed, and communicated at Pixelz. His work has resulted in a Pixelz that works smarter, more efficienctly, and is more aligned across the entire organization.

We spoke with Newton about the improvements he has made at Pixelz and how other businesses can use effectively use data experts to help drive positive change in their organization.

Table of Contents

Issues We Faced With Our Old Model Of Handling Data At Pixelz

Over the last year since you joined Pixelz, how have you changed the way teams at Pixelz access the data they need to make decisions, track performance, assess risk, etc.?

Before I joined Pixelz, our data team played a passive role in our daily operations. What I mean by that is, the data team would wait for a manager to submit a data-request ticket and then the data team would pull the raw data from our database. We would then send that raw data back to the line manager with the expectation that the line manager analyze the data themselves.

There were a few drawbacks to functioning this way.

One concern I had was that the data team was often a bottleneck in our operations: while a manager may be “working on three projects,” that often meant they had tickets submitted to the data-science team and didn’t feel like a decision could be made or plan executed before they received their requested data.

We also faced alignment issues as a result of allowing managers to request and analyze raw data themselves. We saw a lot of low-value data requests that we knew probably wouldn’t unearth anything interesting, or requests that unintentionally ask for different methods of calculating a KPI -- as an example, one team using averages and another using weighted averages.

Obviously, I wanted to eliminate those drawbacks in a way that allowed both for my team to use their skills as data experts and our organization to work for efficienctly.

Improving How Pixelz Accessed and Leveraged Our Data Team

What was your vision for improving how we accessed and utilized data at Pixelz more effectively?

There were three pillars to my approach.

The first thing we did was automate as many reports as we could. We worked with our team leaders to decide what data was crucial for them to have access to on a daily/weekly/monthly basis and, using both RJ Metrics and Google Data Studio, we automated as many of those reports as possible.

Today, not only do managers have easy access to their most important data, those reports were all built by my data team, so we have also eliminated the alignment issue from before, helping ensure that we are all looking at the data the same way across the organization.

Transparency provides a touch of magic by helping other managers understand what other teams are doing, what KPIs are most important to different teams, and how each team fits into the context of the larger machine that is Pixelz.

Automating reports decreased the number of data-request tickets that my team needed to process. For the tickets we did receive, I encouraged my team to be more critical towards the data requests made by managers. Rather than processing every data request as soon as possible, we began to ask managers about why they needed the data they were requesting, what they were trying to measure, asking managers to provide business context and priority with their requests, and we provided them more of our thoughts and expertise to help make sure they were getting useful, high-value data. Today, because we take a more active role in the process, we are able to provide data-insights that the requesting managers may not have seen on their own.

The third pillar came with our C-Suite empowering my team to be more exploratory in nature. They asked us to propose new cost-saving initiatives or revenue enhancing opportunities and to introduce new projects for other teams in the company based on what we were seeing in the data.

Of course, some of our explorations end up in dead ends, but we find those dead-ends much faster and spend less resources than if that exploration comes from outside our team. And explorations that have been successful have led to measurable changes/efficiencies in our operations that are now creating hundreds of thousands of dollars in annual cost savings.

I am thrilled that we now function much more like internal consultants and project stakeholders: we control the entire flow of data throughout the organization and managers involve us in their projects from start to finish.

Turning Data into Action

Your team has been successful at turning data into action in a way that not all data teams are. Is the shift from being simply a data-producing team to being used as expert consultants within the organization part of that success?

Yes, and I think leveraging your data experts as consultants the way we are now makes for healthier operations overall.

Because our data-science team isn’t just sending back raw data for managers themselves to process, we are able to provide insights that help managers be more efficient. Not only do they no longer have to spend hours doing their own pivot tables and analysis, we can also help steer them away from, say, exploring an area that has low business impact and towards areas of higher value.

That has meant less meetings discussing KPIs that are low-impact or that are already where the business needs them to be and in turn has allowed managers to move on to other areas of the business that can be improved.

Greater transparency has also played an important role in our ability to help drive change. Because my data team controls the data-flow in Pixelz, we have the power to give all of our managers access to dashboards that cover the operations of nearly all of our teams. That transparency provides a touch of magic by helping other managers understand what other teams are doing, what KPIs are most important to different teams, and how each team fits into the context of the larger machine that is Pixelz.

What we see now is managers asking other teams about their KPIs and forming a dialogue about what KPIs are most important for them. The result has been that, with a better understanding of other teams, managers can optimize their own teams in ways that also help drive the KPIs of related teams.

That cohesion, especially in a big company like Pixelz, has been critical in translating data into action.

Three Pixelz staff standing near computers talking about data

Integrating our data science team in our creative operations has been critical to our success.

The Role of Pixelz Data Team in Decision Making at Pixelz Today

What role does the Pixelz data-science team play when we are discussing potential solutions to challenges we are facing?

One of the great things about Pixelz is that data is really valued by everyone.

Our C-suite at Pixelz has decades of experience with our operations, and our team leaders have unique perspectives related to their team function, so both are in a position to ask some very thoughtful questions about the data my team presents. It isn’t unusual for us to have passionate discussions about the data itself, debating the date ranges from which we are pulling data, potential influencing factors from when the data was pulled, how much data is required to make an informed decision, etc.

Our job on the data team is to either anticipate the concerns of our audience and have data ready to help answer any questions they might have or be able to highlight what data we don’t have and explain the potential risks that those unknowns might pose. Then, it is up to individual managers to our leadership team to execute on a decision.

Communicating Complex Data to the Organization Effectively

Like many organizations, our data can be really complicated. What advice do you have for organizations for taking complex data and making it digestible for everyone?

Like any team, each member of a data science team has different strengths. Some may be more communicative and comfortable sharing their work with non-technical staff. Others may want to take on the hardest technical problems and be less involved in communication with other teams.

My job coming in to Pixelz was to figure out who on my team had which strengths and then leverage those strengths so that everyone felt they were in a position to succeed.

In terms of how we share data, we always make sure to include three components any time we are sharing analysis with a wider audience:

1.) The problem statement to provide helpful context about the problem we are trying to solve.

2.) How we gathered and processed the data (date range for the data, what variables we considered, etc.) in as simple language as possible.

3.) How the data or analysis impacts the stakeholders.

The third component of tying the data or analysis back to people’s teams in a concrete way is really important. If you are not always tying your analysis back to how your findings impact your audience, you are going to lose people’s attention. For the C-Suite, that might be tying things back to cost-savings or revenue-enhancements. For our Production team, it might be monthly hours saved as a result of efficiency gains.

Whatever team it is, we always want the audience to understand how our analysis or findings impact their work.

How to Integrate a Data Team to Your Creative Operations Effectively

If you were consulting a company that wanted to integrate a data-science team, what advice would give them to make sure their data team is able to deliver as much value to the organization as possible?

To start, it is critical that your data team is aligned with the leadership team or C-Suite, or at least has a really deep understanding of what the primary objectives are for the company in the upcoming quarters.

Aligning your data team with the goals of the company allows your data-team to help drive the company in the right direction. My team can explain to managers exactly why a data-request may not be high-value or prioritized and we can then help push them in a direction that is of higher-value because we understand the bigger picture of where the company is headed.

You also need to make sure that the tools and workflows for requesting and accessing data are efficient. If you want people leveraging data, they need to know how to request it, where they can see it, and it those processes need to be as easy as possible. The ultimate goal is for your data team to build data reports which are downloadable by others, so workgroups can "self-service" their own data requests.

Third, centralizing your data with your data-team is crucial because it ensures that all your data and the language you use to discuss data is aligned across the organization. By centralizing our data with our data experts, we were able to help define metrics and ensure that everyone across the organization was using the same ones. Establishing that common language has been really important for how communicate with each other.

Interested in making your business more scalable? Sign up with us to recieve 10 free images so you can try Pixelz products and service for free.