Content Published Last Week

AWS, GCP, Azure, Oracle & Cloud Hyperscaler Competition

There are many interesting parts of this interview with a Former Director of GCP who worked with Thomas Kurian at Google and Oracle and who was also at Microsoft during the launch of Azure.

Microsoft pretty much gave away Azure in the early days:

Everybody has at least Windows and/or Microsoft Office and Teams now, and Satya leveraged the heck out of that. They said, if you're a Microsoft customer, it's easy to add Azure. During the early days we would add Azure credits and people would ask what the hell is this $20,000 worth of Azure? We would say, don't worry about it; we just took $20,000 off your SQL Server licensing and gave you $20,000 monthly credit. That's how we planted that seed, and now they're using it and consuming it. - Former VP at Google Cloud

Microsoft had the enterprise customer base, FCF generation, and a great leader in Satya to plant the seeds of Azure.

Compare this to Oracle…

Oracle who at that time was one of the largest database companies in the world, shifted onto the cloud. When I was at Oracle, I realized how late Larry Ellison's strategy was, because he is very arrogant. I think his hubris let him down thinking they will never get off Oracle. He even challenged AWS and said Amazon would never get off Oracle, but they did it within 18 months. They shifted their ecommerce database consisting of millions of customers with billions of SKUs, over to Redshift. Then they realized they needed to do something about it, but it was too late. - Former VP at Google Cloud

Historically, GCP’s go-to-market strategy was built around BigQuery:

Google's business model is about capturing data and turning it into a profitable ads business. 80% of Google revenue comes from the ad side like YouTube data. They already knew how to capture that data and use AI and ML. They took those learnings from the Google side and engineered them to the cloud. When I was at Google, our pitch was that we had best in class data analytics. BigQuery was our flagship we were trying to pitch, and we had the most advanced AI to wrap around those products. That was our differentiator…The differentiator was you have this huge data you need to consolidate into a data warehouse and we had the best-in-class tools and the most intuitive interfaces, plus we can leverage data on your ads business, because many customers were spending millions with us on ads. Why don't we leverage that data to give you more of a 360 view of your customer and partners? - Former VP at Google Cloud

GCP seems to have an advantage acquiring customers that have large GOOG ad budgets (mainly retailers that don't want to be on AWS) or companies with high volume and high velocity workloads which work best on BigQuery.

But could this also mean GCP is more at risk from SNOW and Databricks that offer just as good analytics services but across all clouds?

Maybe this is more of a two-horse race between AWS and Azure than it seems.

AMZN was first to market with the greatest market share and every large enterprise customer either uses Windows, Office, or now Teams. If GCP’s GTM leads with BigQuery focusing on those with high volume, real-time workloads and a large ads business, GCP’s terminal market share could be lower than expected.

One final interesting comment from the interview was the relative attractiveness of usage vs seat-based SaaS models:

If you're going to invest in a company, I would invest in a consumption business versus a user-based business. The Rule of 78 is very important for cloud consumption. This is the reason why companies like AWS, Azure and GCP can grow at a much faster scale than someone like Salesforce or Workday. When you have a user model, you're limited by the number of users. If you have a user model as big as Exxon Mobil was, they only had 20,000 users that actually used a SaaS application, you can't grow that. Walmart have 100,000 employees, but they are not going to double that in a year. You can with a consumption model because the number of applications being built can double or triple. - Former VP at Google Cloud

It’s an interesting principle: typically, it’s easier to grow with usage-based models rather than seat-based.

This is mainly because customers can grow the number of applications faster than FTE’s.

Assa Abloy & Electromechanical Locks

ASSA is a serial-acquirer in the safety locks and access solutions industry. Over 27 years ago, it was created by a merger between Assa and Abloy, both companies over 100 years old. The merged entity owns brands such as King, Yale, Sarget, HID, and Lockwood. ASSA has ~10% of the $100bn access solutions market.

Source: ASSA CMD Presentation
Source: ASSA CMD Presentation

This interview was with a Former President and CEO of HES, a company that was acquired by ASSA and integrated into its electromechanical division which accounts for~30% of group revenue and is one of the fastest growing segments of the industry.

This comment reminds us of TDG; great serial acquirers are rigorously disciplined to the core strategy.

The thing making Assa Abloy unique is they stay true to the core of their business. When Assa Abloy first acquired effeff, 50% of effeff's global business was an access control company. The first thing Assa Abloy did, when they bought effeff, was split the company in half and divested the access control division. They've done that time and time again, and the message from the beginning of the original management team was to dig where you stand; don’t get too broad away from your core business. The strength of Assa Abloy is they stay true to that. They stay true to staying core to the business. That’s branched out to doors, anything from a small door to an airplane hanger. Again Assa Abloy wants to own the opening and everything that goes on inside that opening. They've stayed very true to that. - Former VP at ASSA

Adyen, Stripe, Checkout, and Payment Acquirers

A Former Stripe VP believes acquiring is becoming commoditized:

Acquiring is becoming more and more commoditized and it's becoming more and more under pressure. It's more about the value-added services for protection solutions or how they help customers to increase their conversion authorization rates. That's probably what is more important for merchants because finally negotiating on a few basis points is less impactful than maybe 10% higher conversion rates. That's how Adyen can probably maintain their profitability and also sell a premium model which is quite similar to what Stripe are doing as well. Stripe has a similar approach where they really sell at premium value but within an overall proposition, there's so many components. Sometimes they drop their pricing on a particular product in return for maybe more profit on a different aspect of the proposition. - Former Managing Director at Stripe

Wayfair, & European Online Furniture

We asked a Former International Director at What drives conversion in online furniture?

His answer: lead-time.

Lead-time is critical, I would say less so now than at the time, but we saw that having a showroom or physical awareness also drove conversion in the cities where we also had conversion. Obviously, you would feel less comfortable purchasing a sofa without having tested it, seen it, or touched it. Those were key growth drivers for Made, and now it's more common; the market is more mature. Also, online penetration was a key growth driver. In the end, as with any online brand, it's about trustworthiness, credibility, authority, presentation, and branding that must be spot on. - Former MD at

This is an interesting answer given’s model was built on shipping well-designed furniture before it’s even produced in Asia. This leverages net working capital but extends lead times for customers.

Since listing in June 2021, is down 95%. It is not alone: Wayfair and Westwing are both down over 80% since June 2021.

Although the IPO may prove to be bad timing, MADE’s business model seems fundamentally challenged at scale. To grow beyond early adopters, MADE needed to reduce lead times. There are two options: own more inventory or offer 3P brands drop-shipped to customers.

MADE is trying both. Increasing inventory reduces the negative working capital that was funding the business and injects more complexity into operations. MADE also acquired Trouva to offer 3P brands and is pushing towards a marketplace model, similar to Wayfair.

Although the macro backdrop is significantly challenging for large-ticket, discretionary items like furniture over the next 12 months, it seems that Wayfair has designed the superior business model to win in online furniture; a wide selection, shortest lead times, effective pricing, and no inventory risk. MADE is yet to prove a scalable model beyond design-focused millennials who are willing to wait 3 months for products.

US LTL Trucking: Old Dominion Freight Lines vs YRC Freight

A Former YRC Freight executive on why ODFL is so dominant: pricing.

One of the pieces was our pricing structure. I never had a lot of confidence in our pricing structure. The pricing piece is so important for success at an LTL company. How do you maximize the dollars per cubic foot on your trailer and maximize the dollars, not only from a sheer number of dollars standpoint, but what are the operating characteristics of those dollars per cubic foot? Our pricing structure was certainly complex but it was so many averages of averages. It was an average line haul cost, average P&D cost, average sales cost and that band of whatever percent the band was of the shipments in the middle, of what the average actually were the amount of shipments that were priced properly. But the vast majority of the shipments, because they were out of that average band of shipments, were priced improperly. - Former VP at YRC Freight

Spirax Sarco: US Steam Positioning and Distribution

Steam is relatively more expensive and harder to manage than hot water, which could replace steam systems for certain market segments. The executive believes that ~40% of SPX steam business could be a risk.

The biggest risk for Spirax is the erosion in the trade who understand how to manipulate, operate and maintain steam systems. More systems are converting to hot water because they are less complex to manage and less expensive than a steam system. Healthcare and food and beverage processes require steam so they have operators who will continue to maintain and operate those systems, but if there's a choice, people will move away from it. If something doesn't go your way on the regulatory side, it creates an even bigger challenge, so you operate in fewer areas and are still very susceptible to the ebbs and flows of the oil and gas and food and beverage industry. Healthcare has a lot less volatility and is a good example of where they're moving towards hot water and away from steam. - Former VP at Spirax Sarco