How often?

How often do you make changes to your code, reports, or databases?

It’s likely a daily activity. A SQL update here. A visualization adjustment here. A new index on a table there.

This is the work of a data team. Daily building, updating, and refining.

———

How do you make a change?

How could you undo a change?

What’s your process for making a change?

How do your teammates make changes?

How do you know who made the change?

What if the change isn’t the right change?

What if the change breaks something else?

How do you know if your change breaks something else?

Who knows about the change you made?

Who cares about the change you made?

Do you know what changes you made last month or last year?

How were the changes you made tested?

What if you and your teammate want to make contradictory changes?

What if you don’t know that your teammate is making a contradictory change?

———

For something as central to our daily work as “changes”, we spend very little time talking about these questions.

Most data teams I talk with spend little time building frameworks and processes for doing changes well.

But if quality, consistency, speed, and scalability are a priority for your data team, it’s time to start asking these questions.

And taking steps to answer them.

I’m here,

Sawyer

from The Data Shop​

p.s. hit reply and tell me about your process for “changes”.

Humans follow their programming

The next time a business team asks for something you think is “crazy”.

Ask this: Why would they want that?

Humans don’t knowingly ask for crazy or wrong things. They follow the scripts about what they know and understand.

Another way to say that: Humans don’t make mistakes. They follow their programming.

That “crazy” request is based on years of experience and assumptions about how their job works, how data works, and how to get what they want.

It’s not that they “don’t understand data.”

It’s that they’ve been conditioned - through the company culture, team dynamics, and previous successes and failures – to ask for their request in exactly that way.

Don’t look at your business stakeholder’s crazy request as a problem to be solved through training. Look at it as a product of a system.

I’m here,

Sawyer

from The Data Shop

Flip the script

In the data world, we want more “Data-driven decisions”

It’s a mantra. Slogan. Branding.

And I agree.

Our decisions should be driven by data. When business leaders are faced with key decisions they should look to support their decision by the available data.

It’s a decision-in-search-of-data model. Which makes sense when you are a decision-maker.

But most data teams and data leaders aren’t decision-makers. They are advisors or partners in decision-making.

And as we build data products, analytics, and reporting it’s useful to flip the script on the old mantra.

Decision-driven data.

Instead of decisions looking for data (which a business perspective), we are a data looking for decisions.

We caretake, massage, manage, extort, extol, and manipulate data to support decisions.

If there isn’t a decision, then we don’t need data.

If you need to ​thin out your crowded data platform​, then remove that data that doesn’t explicitly support decisions.

Try on the decision-driven data filter.

Odds are, everything else is noise. And crowding out other more valuable data insights.

Sawyer

from ​The Data Shop

It’s crowded out here

We have a grove of pine trees nestled along the side of our house. It’s a mixture of red and white pines that are native to this part of the country. We love the pines, our boys have a tree fort in the midst of the grove and they provide a great playground.

I never want a single one to die or fall down.

But I recently walked the trees with an arborist and his main recommendation was this: Cut a bunch of them down.

Intentionally cut down the trees you love.

Why?

“They are clustered too close together and competing for resources with each other. The canopy of the smaller ones will never make it high enough to get the sunlight needed to thrive”.

So he talked me through a plan to remove a portion of the trees. Focusing mainly on the smaller thin ones. Over time, thinning the pine grove will create a much better environment for the remaining trees to grow stronger and healthier.

The same with your data.

The reports, dashboards, and bar charts your team produces are plentiful. The dozens (hundreds?) of reports your team delivers might be one of your favorite parts of your job. After all, data visualized is data realized right?

But it’s time to thin the number of reports. They are competing for resources with each other. Choking each other out. The little used ones never have a chance of reaching the surface of someone's attention.

Your stakeholder’s attention is too precious to overwhelm them. Sifting through dozens of dashboards. Choking out in all the noise.

But thinning the forest (deleting reports) can lead to a much healthier environment. A place where the strongest analytics can thrive. Where insights have the resources and attention they need to offer value. A place where business users don’t get lost in a crowded forest.

And sometimes all you need is an outside voice to say - “you need to cut some of those down”.

I’m here,

Sawyer

from ​The Data Shop

They ignored your data

You did your job. Delivered amazing data.

Perfectly aligned with what the department leadership needed.

They took your data.

And ignored it.

Even worse, they made a decision in exactly the opposite direction of what the data indicated.

“So much for ‘data-driven decisions’” you mutter in frustration as you kick rocks down the sidewalk.

If this happens enough - and you care enough - it’s easy to get jaded, angry, or cynical about the role of data in your organization.

Hold on. Before you quit…hear this.

There’s a reason why this happens.

Incentives.

Incentives drive all of our decision-making. People make decisions based on what they want. And what they want is controlled by what they are incentivized to do.

They might have ignored your data because they knew:

• they would get a promotion if they orchestrated a different course of action that would look good to their leadership

• they would get a bonus this quarter by getting a short-term win.

• they’ve seen other people get fired for taking a risk outside the norm.

• they are stressed and overwhelmed in their personal life and don’t have the energy to take on the challenge that they data points toward.

and on and on.

Data-driven decisions are only part of the equation in making good decisions.

The human incentives are often a hidden feature/bug of the system

I’m here,

Sawyer

from ​The Data Shop

"Yeah, we do data"

Ask any CEO or executive director out there - “Do you data?”

Of course, they do. The mantra “data is the new oil” has been in the water long enough that no CEO will miss that talking point.

They know that “everybody else is doing it”. And so they insist that they are doing it too.

What if I asked you if you have a vehicle?

You might live in an urban center and embrace the car-free life, but 90%+ of American households own a car.

Everybody is doing it. Parking their car out front of their residence and using it to drive around.

But there is a dramatic difference in the type of vehicle you might own. It could be a work truck, a commuter car, a minivan for driving around kids, a sports convertible you only pull out on sunny weekends, a vehicle for your teenager to drive, or something you use to drive Uber to make extra cash.

The details of “Do you have a vehicle” vary substantially based on the purpose of the vehicle. Not just in terms of how much it’s used, but when it’s used, the type of vehicle required, how you care for it, how it’s taxed, and how much money is spent on maintenance. They are all vehicles, but people fundamentally think about these vehicles dramatically differently.

Here’s the thing.

The same is true for your data and how your organization views the purpose of data. When your CEO says “we do data” how do they think about data? How your organization thinks about the purpose of data will change how they care for it, what types of tools are required, what budget it gets, etc.

"Doing data stuff" isn't going to move the needle in your organization.

You need to understand how data fits into the shape and vision of your org.

Otherwise, you will be left hoping for a data team that leadership won't have the budget, time, energy, or attention for.

I'm here,

Sawyer

from ​The Data Shop​

The Bracket Strategy

In the US we've got this thing that happens every March. It's the end-of-the-year college basketball event where the top 68 teams play in a large bracket-style tournament that we call March Madness. The games started last Thursday

The tradition in my home (and in offices, friend circles, religious groups, knitting clubs, neighborhood association, etc) is to try to predict the results of all 63 tournament games by filling out a bracket before the games start.

My family and I filled out a bracket - (mine's in terrible shape if you wanted to know - thanks for asking).

Filling out my NCAA bracket reminded me of how many organizations approach decision-making.

More specifically, how the bias is so insidious. Without data, you tend to make decisions following one of these methods:

• Nostalgia: "I have amazing memories of this team from the past. They had a great run once in the tornament - I am sure they can do it again". E.g. Loyola Chicago, VCU, Saint Peter's Peacocks, etc.

• Personal Anecdote: "My old roommate/co-worker/mother-in-law/barber/baby-sitter went to school." This clearly makes them a great pick!

• Completely Arbitrary (Mascot style): "No way a Tiger is losing to a Jayhawk". "Does anyone actually know what a Bonnie, Zips, Hokie or Shocker is?)

• Quasi-Statistical: "I am just going to pick the higher seed all the time and then flip a coin the final four (or use one of the above methods to decide the final 3 games)"

• Trust your gut: "I just got a feeling about this team. There's a cinderella team every year - might as well be these guys right?"

If you filled out a bracket what was your strategy? And more importantly, does your organization make decisions this way?

I'm here

Sawyer

from ​The Data Shop

Creativity and data team levels.

Fellow daily email list member Dan Gustafsson wrote in with a great observation about two of my recent emails: Creativity and levels of a data team.

Dan said this (shared with permission; edited lightly):

I think ’efficiency’ is key for enablers, but ’creative’ is key for partners.

Enablers = processes, and they can be ”efficient” in some sense.

Advisors= closed-ended exploration (goal-oriented analysis) can be made more efficient – consider management consultant ’frameworks’ to aid that.

Partners = open-ended explorations = never be ’efficient’ by definition.

Efficiency and creativity do move on a continuum across the functions a data team. While the detailed activities of each type of team will, I think Dan is right.

There will be functions of your team where you strive for efficiency to remove waste. And other functions of your team where you create excess capacity to explore ideas and innovation.

If you are mainly an "Enabler" then you will focus on the former. As you move into activities around Advisor and Partner you will take aim at the latter.

Thanks for the insights, Dan.

I'm here,

Sawyer

from The Data Shop

How to use data in decision making

If you want to be successful in data in your organization you have to focus on outcomes:

Here’s a mental model you can try on:

—> Use data to reduce uncertainty in decisions.

Your leaders are constantly faced with decisions.

What to buy, sell, move, invest, adjust, configure.

How much to shift, advance, retreat, scatter, consolidate.

Bombarded with decisions.

The core way they can make better decisions - and close the gap on their desired outcomes - is to reduce uncertainty in those decision.

Every decision has an element of unknown. The more significant the decisions is to the organization, the more value reducing the uncertainty is.

This is where data comes in.

To help decision makers reduce their uncertainty.

Important decision with high uncertainty? Reducing uncertainty is highly valuable.

It moves the needle on the key outcomes.

Which is the fastest way to win with data.

I’m here,

Sawyer

from ​The Data Shop​

Wanting is a feeling

They have to want the results the data delivers.

They have to want it more than you want to deliver it.

Otherwise, the data exists for you

Rather than the business.

Have you heard this conversation before in the data-verse.

Data Team - “Here’s your data”

Business Team - “Great! What does it do?”

Data Team - “Beats me. It’s a report we were told to migrate. You figure it out”

Is anyone surprised that the data collects cobwebs?

How do you get a business team excited about data?

Get them excited about what the data delivers.

Yes, excitement is a feeling.

And feelings drive adoption more than any action or belief.

I’m glad you are here.

Sawyer

Levels of a data team

Here’s a brief roadmap for where data fits into your company. Data and your data team function primarily in one of these three categories:

• Enablers

• Advisor

• Partner

None of these are good or bad.

But over time you might hope to move data to different functions in your organization.

Enablers:

• Delivers Accurate date

• Primarily focused on data storage and data infrastructure

• Basic reporting services

• Limited interaction with the business

• Support and Report Ticket takers.

Advisors:

• Offer Data Analytics, scripting,

• Business Intelligence and Data Visualization

• Centralized Metrics

• Regular collaboration and communication with business stakeholders.

• Output Focused

Partners:

• Analytics are aligned with Org and Company Objectives

• Data leadership is at the decision-making table with business leadership

• Data team bears greater accountability for business results

• Outcome Focused

Why your team functions in the category they do is a combination of the history of data at your company, your team size, organizational structure, industry, and dozens of other micro factors.

Step one for you today - analyze and own which function your team falls into.

I’m here,

Sawyer

from ​The Data Shop​

Your turn

A quick question for you this morning:

What slows down your data team?

Think about the bottlenecks, process gaps, people, technology, etc. that affect your team's ability to move effectively and efficiently in delivering data to your stakeholders.

I'm listening,

Sawyer

from ​The Data Shop

Creativity and Waste

Creativity is inherently inefficient.

To innovate and creatively explore new ideas requires waste.

This means your pursuit of efficiency can kill your data team effectiveness.

Each step towards being more “efficient” is also driving out any margin your team has for creativity.

Creativity and efficiency push and pull against each other.

This is one reason layoffs historically fail at improving a ​company's short or long-term success.​ Layoffs are a cost-cutting and blatant move toward efficiency. During layoffs, employees feel unsafe to chase innovative ideas and create things that may not work - which are the things that historically allowed the company to be successful.

This is not to argue against efficiency. Or for creativity. It’s obvious you must have both.

But as a leader, you must acknowledge that your choices will always push you one way or the other. Failing to realize what you give up in your choice, can have negative long-time effects.

Data is a creative art. If I lose that, I might as well quit now.

I’m here,

Sawyer

from ​The Data Shop

Confidence and clarity

I paid $250 for a course last week.

About podcasting*.

It didn't give me special insider knowledge, exclusive content, or unique access to someone.

Everything it offered I could have found online somewhere else.

Was it worth it? 100%

It saved me hours of time while dramatically reducing my stress and uncertainty.

It made me far more likely to be successful because it gave me the clearest and simplest details about how to get started the right way. And how not to get lost for hours looking for "best podcast mic", or “audio editing software”.

It was content tailored for exactly my needs - get up and running with a podcast as simply as possible.

Here’s the thing.

Everything you need to know about data stuff is online. Between blogs, newsletters, technical documentation, and Stack Overflow (oh, and ChatGPT) you can find an answer to your question.

But how much is your time worth? How much is the uncertainty and stress worth? If your company is investing 5, 6, or 7 dollar figures into your data platform, how much is it worth to decrease the uncertainty, and increase the chance of success?

Probably a lot.

Most data projects fail (stats say 70%+). And it’s not because of a lack of information available online.

It’s for failing to foresee a roadblock.

It’s because of an unclear roadmap and misaligned incentives.

It’s because your team hasn’t been there before and you don’t know what you don’t know.

Here’s an option I offer (it’s just one of many available out there).

A short two-week Assessment and Roadmap project. It is explicitly designed to give you confidence and clarity while decreasing your stress and uncertainty. Depending on your situation, that could be worth a lot.

Hit reply and we can get started.

I’m here,

Sawyer

from The Data Shop​

*Yes, very soon I’m launching a podcast. Standby for lift-off.

Giving them options in Architecture

Respecting and ​partnering​ with your business teams often means ​giving them options​.

You can build this thinking into your data architecture by designing for options. It’s safe to assume that data consumers won’t want the data in the same way every time. That use cases will change, needs vary, and curiosities expand.

So give them options.

Here’s a great example from Go Daddy of a Data Architecture that intentionally provides ​data consumers with options​. It’s a fairly typical layered architecture (sometimes called Medallion).

After they described the nature of each layer in their data architecture, this part stuck out to me:

“Ultimately, clearly defining the data in each layer helps consumers determine the best place to source data for their use case. For some use cases, reducing latency is of the upmost importance so consumers may opt to consume from Clean and Enterprise layers. This means their jobs will likely have to perform many joins, process lots of rows, and implement complex logic. In other cases, consumers have a high tolerance for delays and opt to use more curated data sets to keep their processes simple, light-weight, and fast.”

Options. Does every data consumer value ease of use and minimal joins? No. Does every user want low latency and “closer to the metal” structure? No.

But presenting clarity about what’s available empowers the business teams to get the data they need in the way they need it.

I’ve talked with several Directors of Data over the last few weeks. Many of them were at the stage in their data journey of dreaming about the “art of the possible” and taking an initial step into the cloud.

There are hundreds of decisions to be made at that point. But the first and most influential decision is to embrace business-focused solutions, rather than cool tech toys.

It was good to see you today,

Sawyer

from ​The Data Shop

How much?

How much is clarity worth?

Which decision to make? What’s most important? Who’s essential?

What is clarity in those decisions worth to you?

Or another way to ask it.

How much does confusion cost?

Frustration about direction. Disagreements about responsibilities. Disappointment about progress.

What does that cost?

Depends on how you view the opportunity you have in front of you.

If you don’t see meaningful ways for data to influence the growth of your business, then clarity isn’t worth very much.

If you think the current status quo for your data team and technology is good enough, then clarity isn’t worth very much.

If you don’t see any value in improving the current morale and energy of your team through improved processes, then confusion doesn’t cost very much.

If you think being ignored by executive leadership just comes with the territory and doesn’t bother you, then confusion doesn’t cost very much.

But if that’s not the case. And perhaps there is a big opportunity. And you think these challenges are holding you back.

Well.

Then clarity might be worth a lot.

And confusion could be costing you a ton.

I'm glad you are here,

Sawyer

from ​The Data Shop

Time to Transformation

Pre-S: TSDL Registration closes in 8 hours (4pm ET). I have one spot left in the cohort. It reply to snag it. Cohort launches next Tuesday!

-------------

Things you can offer your company/stakeholders. In order of increasing value.

Time: I will work this shift. Show up at this meeting. Punch the clock.

Inputs: I will answer these phone calls. Respond to these support tickets. Check these boxes.

Outputs: I will give you lines of code. Deliver these reports. Look productive.

Outcomes: I will move the metric that matters most to your business goal. Tweak the outputs until they influence the outcomes. I win when you win.

Transformation: I will own the company's mission and vision as my own. Closing the gaps in our culture. Exceeding the market’s perception of us.

Make this personal. Or shift the “I” to “We” and make this about your team.

Sawyer

from The Data Shop

Why data exists

Pre-s: Registration closes tomorrow at 12 pm ET for The Technical and Strategic Data Leader. This is your last chance to join the cohort! We meet for the first time on Tuesday, March 12th for content and discussion about building great data teams.

--------------------

A simple three-step framework for data at your organization.

Businesses want outcomes.

They get outcomes by making good decisions.

And good decisions come (partially) from good data.

-------------------

or to phrase it another way.

-------------------

Data is needed to answer questions.

Questions show up because we need to make decisions.

Decisions exist because we want to achieve certain outcomes.

I'm glad you are here,

Sawyer

from The Data Shop​

Things that compound

Pay attention to things that compound. They grow slowly at first but pick up steam after years.

• Interest in a bank account (of course)

• Knowledge in an industry or domain.

• Skill in a professional or personal craft.

• Trust between peers and business partners.

In these instances, stay on the bus. The longer the better.

But other things also compound. They pick up gravity and keep rolling.

• Fear of failure.

• Dishonesty between peers or business partners.

• Apathy towards a craft or skill.

• An obsession with looking for shortcuts.

In these instances, get off the bus. The quicker the better.

I’m here,

Sawyer

from The Data Shop

The job interview

I had reached the final interview for the first real data job of my career. I really wanted this one. I turned down a job offer at another company the day before and was all in on this job.

The final interview was a presentation where I was asked to deliver a data story using a data visualization tool. Any topic or dataset I wanted.

It was grueling to pick a topic to present on. I spent a lot of time working through ideas. Then my wife said, “just pick something they will be most interested in”.

Then it clicked. I had been looking at all sorts of random datasets from the public domain. Everything from housing prices to interest rates and population data. It was all pretty boring.

But the moment I turned my focus on the two people I was presenting for - the CEO and COO - it became clear. Both of them traveled a lot for work. They were based in a smaller city but were within driving distance to three larger city airports.

I expected they spent a lot of time evaluating different airports, airlines, and flight options when they needed to fly somewhere for work.

This is where it got fun. I immediately had a dozen questions that I expected my audience to wonder about. Once I found the right datasets, the analysis and visualization came together quickly.

When I sat down to present for them at the interview, they both said “Ah, something we care about”. Based on the response I got, I expect most candidate presentations were about generic public data - without any personal connection to either the presenter or the audience.

Data fundamentally needs a human connection.

We need a reason to care about a piece of data.

Interesting facts, figures, or trends get rapidly boring.

We need data that gives us something useful. That we can change our lives around.

Something to help us make better decisions.

That’s human connection.

I’m here,

Sawyer

from The Data Shop​

p.s. I got the job.