Category: Data/Analytics

Big Data has a problem. It’s not just its bigness; it’s the rigidity of the databases that hold and that force us to make data copies. Resulting problems, from privacy to fidelity loss, are so severe, we should revisit the first principles of how databases are built. Let’s be honest, if we could build our whole data infrastructure over again, would we do it differently? 

Today’s guest says we would have built data like a network. Thankfully, next-generation technology will allow us to store data in this new way while still making use of old-style databases.   

My guest is Chris McLellan. He splits his time between the nonprofit Data Collaboration Alliance, and Cinchy, the leader in enterprise data fabric technology. Coming out of Bishops University with a degree in Political Science, his career has included stints at VarageSale, and Lyft, as well as startups like Flexday and ChangeJar. He also created the go-to-market strategy for Hailo, the taxi network of 35,000 licensed drivers. 

People/Products/Concepts Mentioned in Show

• Backgrounder on AI & Databases
• Company Chris is part of: Cinchy
• Non-profit Chris belongs to: Datacollaboration.org
• Chris’ Ask AI podcast
• Chris on LinkedIn

Stop me if you’ve heard this one before. You’ve taken on a marketing initiative that is finally going somewhere, and it’s now time for management to see what you’ve done and approve more funds, or give you props for  all your hard work. But on presenting your data, you’re only met with blank stares or nit-picking centred around how you put your data together. 

What’s going wrong when this happens? It’s probably that your audience couldn’t boil it down into something that makes sense to them. The universal structure that all of us use to do this is stories.  

My guest uses stories to present the performance of her marketing programs. It’s her contention that using a story-like framework works to your advantage when presenting data. Amy Hebdon has managed Google Ads Since 2004, working her way through at least a half dozen agencies.  In 2017, she and her husband James co-founded Paid Search Magic, which provides coaching, consulting, audits, reporting, and courses for those who want to get better at search engine marketing. She has lived in a handful of states and two central American countries, and she joined us today from the new “Home-base” which is in Tennessee. 

Fun Fact: she once worked for a man (named Mr. Schneer) who was in charge of the company website and didn’t know how to use the internet (he asked which was the “dot” key in typing “dotcom”)

People/Products/Concepts Mentioned in Show

P-A-S framework:

1. Problem
2. Agitate
3. Solution

• Rats in the Cellar, a concept popularized by CS Lewis
• Video on Story Structures, given by Kurt Vonnegut 
• Curator
• Amy’s Twitter profile
• Facebook group Google Ads for Savvy Digital Marketers

Episode Reboot. 

Frame your statistics in a way that illustrates the impact on what your audience cares about.

In its raw form, data’s not worth much. If refined and put together with other data, it can be worth a lot. Here are well-known brands that built their value by creating a useful visual experience out of user-generated data:

• Notable Examples:
  • Glassdoor
  • Nest
  • Zapier
  • Mint
  • Robinhood
  • Flipboard
  • Ancestry
  • GoodReads

This episode’s guest will help us see what is possible once you have data in your hands. Eric Boissonneault grew up loving numbers, but it wasn’t until he saw a Hollywood movie about card players at age 16, that he knew how he would apply his skill. He taught himself poker and methodically played this ‘game of chance’ so well that He became a professional player through his years at University du Quebec à Montreal and beyond. 

After cashing his poker chips in, he wanted to show the business world how they could look at the data they have on-hand as the basis for decisions. In 2020 he founded data consulting company Systematik to help businesses untangle, collect, visualize and understand their data.

Listen in this episode for Eric’s explanation of how you can put a unique transformation or twist on the data you already have, and even make an application that monetizes the data. 

Chapters/Timestamps
00:00 Intro
01:42 Meet Eric Boissonneault
02:53 Systematik Pivot Story
05:09 Clients Data Wakeup
08:22 Dashboards in Data Studio
11:38 Sheets Limits and Costs
15:04 BigQuery Data Warehouse
17:28 SQL and Transform Tables
20:41 Attribution and Platform Silos
24:45 ETL and Cloud Ecosystem
26:22 Blending Data in Studio
29:26 Telling a Data Story
34:32 Forecasting and Third Party Data
40:27 Public Data Examples
44:49 Simple Wins with Finance Data
47:15 Templatize and Productize Data
50:11 Wrap Up and Final Takeaways

People/Products/Concepts Mentioned in Show

• Tools:
  • Google Data Studio
  • ETL
  • SQL
  • Google BigQuery

How fast will you hit Google Sheets 5-million cell limit? If you have a spreadsheet with 5 tabs and each tab fills columns A to CW, and there is 10000 rows of data in each tab. It happens faster than you think.

Disclaimer: Technology vendors that come on the show are not sponsors or affiliates. They’re invited on the podcast to give a broader perspective.

There is still a good deal of information that prospects give you when visiting your site.

If you use this information in the right way, instead of annoying them, you may actually provide an experience that tailors how they’re treated in such a way that may pleasantly surprise them. 

We look in our analytics at all top-line numbers for activities , but not all that often do we drill down to the individual level. This is where we can learn about how our site is or isn’t working for a particular user. 

Wart Fransen is a longtime web programmer and entrepreneur. Freelancing in his first few years after school, he found himself  helping multinationals with their websites, including finding out what they could about who was on them. Seeing company after company wanting the same thing, he set out to build several analytics products to fill these needs. In 2014 he co-founded Leadboxer, a tool that examines a website’s visitors and aims to give insights about if they are ready-to-buy.

Beyond his fascination with technology, Wart’s a father of two and a strong proponent of causes like fighting the disinformation surrounding the Russian invasion of Ukraine. He spoke to me from his home-office near Amsterdam, the Netherlands.   

People/Products/Concepts Mentioned in Show

• Wart on twitter
• Wart’s company, LeadBoxer
• Time-To-Live (TTL) setting for cookies
• ZoomInfo
• Clearbit
• 6sense
• Demandbase
• AWStats

What does it mean to be doing analytics right? It doesn’t mean only that a company’s analysts are being accurate, it has to do with how our company makes decisions. The data guiding those decisions doesn’t just spring up out of nowhere, there’s got to be a functional team analyzing it. What’s the composition of that team? What kind of process do they follow? Do outsourced partners get involved? What technology is used? Leaders that don’t put enough thought into this can burn through both people and budgets, and their organizations never become data-driven like they’d hoped. To borrow a term from environmentalists, their analytics function was unsustainable.

Jason Thompson thought he’d grow up to be a geologist. By college though, he chose to go into Information Systems and came out of college to work at the once-mighty Novell. Over the course of several years, Novell shrank and then laid off his group. He switched to implementing Omniture web analytics for various companies; which showed him both the highs and lows of the consulting firm model. He then went to a brand with a popular web-based business, which caused him to question the value analytics really offered an organization. While there he met coworker Hila Dehan. They started imagining what analytics done right would look like and in March 2013 they started their own analytics shop 33 Sticks. 

There’s more to the story of them working together, such as how they came up with their company’s name, but I’ll let him tell you that. 

People/Products/Concepts Mentioned in Show

5 Analytics sub-disciplines

Architecture & Implementation

Tag Management & Data Layer

Reporting

Analysis & Insights

Optimization

Jason’s LinkedIn profile and Twitter profile