Written: August 2025
Happy Affirmation August!
We’re claiming this month (and next) for those who are or have ever struggled with data projects. Let’s face it. We’ve all been there. But on the other end of the spectrum – the ying to our data yang – are projects we don’t talk about enough. Those inspirational events of data glory, where everything goes as planned, without a hitch, like clockwork. What made those projects so special, and most importantly, what was the secret to success? That’s what we’re going to explore in this issue of The Gestalt Gathering.
How Sweet It Is
Who better to turn to for encouragement and inspiration than our friends and team members? So, we asked them to share some of their best data stories. Here’s what they had to say. We hope you enjoy them and take comfort in knowing that data success is always within reach.
Jonathan’s Stories
#1: I worked for a company where fraud was costing us money. We needed to find a way to predict the fraud. Plus, even if we could, we didn’t know what we’d incur from deals that happened before we put the new process in place.
In order to do the analytics that would be the engine for predicting the fraud, we needed data. Not only from our internal systems, including data that wasn’t even being stored, we needed data from third parties, data that we had to buy, and data from our bank. It was a massive data project that we had to prove worthy, taking about 5 IT/data resources, 2 functional /business resources, and analysts to review deals as they came in.
Sometimes companies don’t analyze data because it’s not there. Sometimes companies don’t analyze data even if it’s sitting in front of them. The sweet spot is having the data right there at your fingertips AND analyzing it. When we finally got all of our data flowing, we analyzed it, built the models, and used it in production. Everything we did was 100% data-driven, no guessing. We went from losing money to being profitable. It was huge, and so much fun!
#2: At another company, we had many internal systems and jobs running as well as one external system. When things broke, we’d lose millions because of mismatches. One system said one thing, another said something else. There was no single source of truth, no reconciliation.
I’d ask the data team, and they said they had the data. Then I’d ask the system team, and they claimed to have the data, too. But no one was taking the data from each source directly and comparing it to find the holes. There were all kinds of holes, not to mention definitions and sources.
You may think you know the truth, but unless you take data directly from each system and compare it, all you have is a guess at the truth.
It took 4 IT/data resources and 1 functional/business resource to finally get each department to send their data directly to us. When that happened, we could find the gaps, reconcile them, and become the keeper of the truth.
#3: My final story is about a company that wanted to encourage employee participation in giving back. After tracking things like participation for 10 years, I was put in charge of it.
Tracking actuals after the fact is great, but data is more powerful if you can use it in the process. So, we created a BI dashboard for all managers and publicized it. From this we could build capacity on how many give-back activities we needed and create competition and excitement.
That year we probably had 10x the participation by using an internal BI tool that helped us focus on our goal of having fun while doing good. Data is powerful in all kinds of ways!
Stella’s Story
I got incredibly lucky to work for a very data-driven organization early in my career. The company had a culture of “Save it.” Even if we didn’t know why or if we would use it. The idea was to save it, so we could get to it without a big project.
My job as an analyst was successful because we saved everything. I can’t imagine how I would have done my job without the data. We did have small issues around two people querying the data differently, not documenting, etc. But for me, it was so powerful.
I started by doing two things:
- Reporting on the past. The typical manual process of pulling in data and building out management dashboards on how we did, if it was in line with our forecast, and if it wasn’t, why.
- Forecasting the future. This was the most fun. Leaders would ask “what if”and I could always answer.
- I had a database full of historical data. If I was asked what would happen if we lowered the cutoff score in Utah by 5 points, I could go through and determine and forecast how it would impact everyone across the board: the regional manager, the state, the company, and the sales reps.
- From there I could show my assumptions and leaders could quickly decide – run it or skip it.
- If we ran it, we could pull actuals and compare to our guess and stop the program if we got different results than planned.
Any question. Pull data. Answer. Decide. Report On. Adjust. Loved those days.
Ken’s Story
A few years back, I was running the data lake for a specialized finance lender. The Credit Risk team depended on it for insights from our Loan Origination System, but there was a problem.
Everything was semi-structured, queries crawled, and answers took so long that by the time we knew if underwriting was too loose or too tight, it was already old news.
So, I teamed up with another engineer and a credit analyst. We ditched the bloated models, built key/value pair tables that parsed XML on the way in, and trimmed the whole thing down to a lean, 10-column, 10-billion-row machine.
Now, instead of waiting months, the team could run deep credit analysis in minutes.
We went from reactive to proactive —and that made all the difference.
Wishing you a wonderful Affirmation August!
Ready to create your own happy data story? We’d love to help make it happen. Already have one? Share it with us for a chance to be featured in an upcoming Gestalt Gathering! See you next month for more data inspiration.