DEEPER DIVES #0004 - Relationships Are The Lifeblood Of Your Business

Good Morning,

First, as always, thank you for joining.

An unexpected twist with this week’s featured article.

Most people don’t know, but I can get hyper-obsessed with a new thing. When something peaks my curiosity or I’m told something can’t be done or is “too hard” … it’s over.

It’s been a wildly beneficial aspect of my personality for my career.

Not always the case for my sleep schedule.

Learning, innovation and offering more value are the themes this week.

Here’s what this issue brings:

• Featured Article: Business insights are built on SQL. But is it actually showing you all of the things you need to know?
  

• Broader Focus: Dramatically improve your operations by fixing one thing with AI.
  

• From the Community: Increasing the value of your logistics. How to stop selling on discounts.

Increase Performance And Profitability By Finding Patterns In The Chaos

Our brains are hardwired for patterns.

And we recognize those patterns through relationships.

Connected, stored, understood and actioned from the context we see.

I’m currently working with a large retailer doing strategic network optimization.

A huge success factor for these types of projects comes from how well you do your planning. Doing proper planning means that you have to get a REALLY good understanding of what’s happening out in the field.

To that end, I’m working with a few million lines of transaction data. Right down the individual items by day.

With that much data, there’s a lot of ways to cut it, twist it and pivot it around. The low hanging fruit is obvious and easy to see. But because there’s a need to be really tight on the implementation, deeper insights on product, inventory allocation and market position are key.

Protip: Never work with anyone that will tell you that they can help you improve if they won’t use the lowest level of detail you have.

To SQL or Not to SQL

Relational databases are the backbone of business these days.

We store huge amounts of information in tables. Lots … and lots, of tables.

With the structured nature of this data, we can exchange information across different platforms, into various systems and connect them to Business Intelligence tools. To process large amounts of transactional data, they’re great.

But how easy is it to see all of the opportunities that your data is holding?

Navigating Like Neurons

By now I’m expecting you are wondering what’s going on.

I’m an operations guy, why so much about data?

Our number one job in operations and logistics is to deliver for our customers. Understanding what’s moving, how, why, and when is paramount to executing at excellence.

As the distance to the end consumer shortens, our job is only getting harder.

To help set up this new operation, I’m interested in knowing:

• Customer (consumer) behaviour

• Product relationships and correlations

• Sales patterns

• Activity density

• Customer and order similarity

Why?

Because understanding the most representative customers (and activity) in your network is how you go from good to great with daily operations.

So what captured my attention and focus to help me dig for the deeper insights I’m designing around?

Neo4j.

Neo4j is a graph database (no, not a collection of charts like you’d find in PowerBI or Tableau) that allows you to store and query interconnected data. Graph databases take data records (things) and show you how they are connected to other records (other things). These connections are the relationships in your data, and they allow you to move through the network like a spiderweb.

Graph databases are dramatically changing how we interpret large amounts of data. They have been the fastest growing database category since 2013.

I’m setting up the data to look at a few key nodes (types of things) and some of their relationships.

Nodes: Customers, Products, Orders, Categories, YMD, City, State, Geohash

Relationships: Purchased, Belongs To, Contains, Returned

While none of this is impossible with tools like SQL, there are a number of benefits when using a graph database.

The most important to me however will be the context that will be able to be derived from this analysis and to uncover insights that might not be apparent with traditional data analysis methods.

If you’ve never heard of graph databases, or are looking for other ways to find new opportunities in your business - I highly recommend exploring them.

If any of you have worked with this type of analysis, I’d love to connect on a call to talk a more about your experience

Use Irritation to Fuel Innovation

I’m a huge believer in test and learn.

These days, too many people are afraid to try new things.

It’s easier to go with what you know, or just do what you’ve been asked to.

Innovation doesn’t come from staying in the same place and doing the same thing in the same way.

It comes from change.

From one idea that forms, and bumps into the next and into the next one after that.

But who has time for more hobbies and more self-learning these days?

Over the years, I have had a very consistent message with my teams.

Something that was the core of process improvement initiatives, the heart of problem solving, and the best catalyst for innovation.

FIX THE THINGS THAT BUG YOU!

The AI Challenge

Whether you love AI or are absolutely sick of hearing about it - it’s here to stay.

There will be two camps of people.

Those who lean on it for simple answers vs those who use it to unlock potential.

My personal opinion is that those who use it to unlock potential will be much better off down the road.

Here’s the challenge.

Find something at work (or in your personal life) that is bugging you. Something that should be automated, could be done more quickly, takes too much effort for what it is … whatever.

I am sure you have something.

We all do.

Now use AI to help you solve it - or at least, make it a lot less annoying.

The goal here is to give you a real world problem to get into the test and learn cycle. Whenever you are trying to develop a new skill, process, report, or anything else, it will ALWAYS be easier when you can connect it to something in your life you are already spending time doing.

If you can, pick something that you might be able to make solid progress on in a week or two.

Then start.

That’s the challenge. And that’s how you’ll be 1% better tomorrow.

Now, since I don’t ask anything that I wouldn’t do myself, here are a few of my recent ones.

Geohasing

Some of you might have seen a number of posts I have made recently about geohashing and how you can use it to better understand what’s happening in your delivery network.

Even though it’s been around for a while, it’s not a main stream data element that most people use (even though it’s the concept behind Uber’s mapping technology).

What was bugging me?

That unlike geocoding, it’s much harder to find services / options to batch geohash a list of address. Even harder if you are looking to do any real project level amount.

Since I couldn’t find what I wanted, I built two tools

A Python script (with a GUI) that allows you to select the input file, confirm the column names for your latitude and longitude, and tell it whatever field is your UID (unique record ID). This thing batch geocodes like a boss. Python installation is required, as well as the supporting libraries - but don’t let that scare you, it’s easy to do.

If Python isn’t your thing, or you don’t have that many to do, I also did a version in Google Sheets. The only thing you need to do to get this working (once you save your own copy), is to load up App Script (sign into your Google account) from the Extensions and run the code in there once, before jumping back into your sheet.

(Can you tell the level of frustration geohashing was for me a couple of weeks ago!)

FuzzyWuzzy

Another project I was working on was KILLING ME with funky address data. While easy enough to understand for a person, the typos, spaces and abbreviations wreck havoc on lookup functions.

Fuzzy Matching allows you to compare two strings and see how similar they are. The most similar they are, the higher the likelihood they are the same.

Totally disappointed again with the options and quality of results, it was back to Python to leverage the fuzzywuzzy library - which is one of the best tools to use for this type of work.

This script is also GUI based. You pick your masterfile (file A) and your reference file (file B), give it the columns to match, and where you want to save the output file (it will be a all of the data from file A, with the matched UID field from file B - with the accuracy scores).

Now, before you get totally frustrated with me and pick me for the thing you are going to “fix” this week (as I’m suggesting to work with AI while dropping Python code on you) here’s the thing - I haven’t written code since year 1 semester 1 at university.

Everything I put together here was using new AI tools.

I explained what I wanted, I tested and debugged (found mistakes in the output files) and kept going back and forth until everything was good.

I didn’t write 1 line of that code - but I did fix some indentation.

Your turn.

Think Like a Salesman, Execute Like an Operator

The most common question that’s coming up from conversations in the community: How do I stop selling on discount?

Here’s the most fundamental thing to embrace.

If a customer will only buy your service when you slash your rate, they don’t see enough value in anything else you are offering.

Lowering prices is the oldest, and honestly, most effective way to dramatically alter the value proposition.

When someone feels they are getting a lot more out of a deal than they have to give up, they say yes.

There are two major issues with this strategy, which is why I work with my clients to avoid this at all costs.

1. Once you cut your rate, you have established a baseline for your service with the customer. From that point forward, any time you try to increase rates, you trigger a sense of loss in them. Never good.
   

2. The more capacity you allocate to low margin (or even negative margin) customers, the more pressure you put on your team to maximize the revenue from other customers - they likely have to pay more than their fair share so that your business can be at the right level of profitability. Also, never good.

I’m going to use warehousing to illustrate a the concept of selling on value.

Why warehousing? Because warehouses are sexy (You can thank Kevin for that one)

Warehouse Services

Here’s the hard part.

Receiving, storing, picking, packing, shipping. Nothing here is selling on value.

Basic fulfillment is the expectation. It’s what everyone is offering.

There is not one competitor you have that doesn’t do those things. There isn’t one customer who is looking for a warehouse, that isn’t expecting you to be able to do those things.

Better inventory management to not lose anything?

Not good enough. The customer doesn’t expect you to damage or lose their stuff.

Great management team?

Again, this is what a customer expects.

Remember, there’s no salesperson anywhere cold calling customers and winning deals by telling people that his operation is average or does alright most of the time.

So what’s selling on value?

It’s providing services that the customer sees as making things better or easier for them.

• Customized services:
  This is an easy one. Work with your customers to develop solutions that respond specifically to the needs of their business. Selling a fixed service set with no flexibility is guaranteed to go right to a rate discussion.
  

• Real time inventory tracking:
  This one is also becoming much more common. If you aren’t doing this already, it’s a must for where things are going. What’s needed here is a system in place that knows (and can communicate) the exact state of the customer’s inventory. What’s there, what’s allocated, what’s available, what’s defective. Bonus if you can accept and incorporate business rules or custom instructions to support their customer management or promotional activity.
  

• Warehouse inventory management services:
  More on the B2B side, or if you work with manufacturers in the CPG space, having a team that can take ownership of ordering and managing what needs to be brought into the warehouse takes the work off your customer’s plate.
  If the client uses multiple providers in different regions or countries, you could take on the management of the flow of goods, acting as the primary hub for distribution.
  

• Dynamic Slotting & Bin Configurations:
  Build the capability to offer dynamic slotting and bin configurations to match seasonal trends, promotional periods, new product launches or line extensions
  

• Custom packaging & inserts:
  Retailers are interested in personalization more than ever before. The ability to create small batch custom boxes, to support a brand’s raving fans or trigger white glove service from the customer and order level all leave you offering your customers more.

That’s not an exhaustive list.

There will always be ways to connect with your customers and help them solve more pain points.

While being better than their last provider is great - that may be a case of how bad they were versus how much more you are able to do for them.

That’s it for this week. Thanks for being here.