End game series vol. 4: Building an eCommerce platform is like building a skyscraper

One of the things that i’ve found interesting over the years is patterns… in particular patterns that end up being mistakes. As they become predictable. And so when i see the pattern repeated… I like to see if the outcome holds true to the previous examples. Now when you logically analyze the mistake in hindsight, it is fairly clear why it was a mistake. And yet you continue to see others make the same mistake. Now begins my story…

The context: I have had the opportunity to ‘look under the hood’ in 9 different large eCom companies

In addition to working at a few places… I’ve done a lot of contracting the past few years and so it has the benefit of allowing me the rare opportunity to have seen inside a lot of ecommerce companies.

This list includes (roughly estimated $GMV):

• Asia: Lazada ($10bn+), Alibaba ($1 trillion+), Pomelo ($50m+), Landers ($30m+)
• Europe/Russia: Ozon ($3bn+), Yandex Market ($2bn+)
• US: Wayfair ($15bn+), Branded ($200m+)
• Africa: Sokowatch ($200m+)

For each of these companies, I could pretty much tell you things like:

• How they organised their tech team
• What were the core pieces of the software architecture and how they connected
• What some of their main problems & mistakes were?
• What were some of the things they did that worked well?
• etc.

With each peek under the hood I learned more about what i consider ‘optimal’

Early on in my product career (back in 2014–7) I got to see the inside of quite robust platforms like the one that Lazada used (mainly built by Rocket Internet) as well as Alibaba. This sets a very high benchmark… but you then realize that other smaller companies do not need this level of complexity and customization.

And so you kind of learn which blocks make the most sense for which size & type of ecom company.

And there are a number of elements at play like..

• what do they already have and how tangled is it?
• how good is their team?
• how much funding do they have and what is their longer-term ambition?

And what I generally concluded was that building an ecom platform should probably be viewed the same way as building a skyscraper.

There are building blocks like the front end, OMS (order mgmt), LMS (delivery), WMS (warehouse), the middleware (eg. Celigo, Mulesoft), etc. And you put these building blocks together in an optimal & sequential way.

It probably holds a ton of parallels to how building skyscrapers has evolved over the last 100 years or so.

• There are architectural rules you need to adhere to.
• These have typically been learned over time the hard way by other companies.
• There is a lot of know-how involved in building a foundation that scales well for both volume and complexity and withstands the test of time.

As you read these statements above.. they generally sound like i could be talking either about building a skyscraper or about building an ecom platform. I mean think about it… is the core ecommerce experience of Amazon that different from a Lazada or a Shopee? And are the differences in all the core elements of what makes a great customer experience (like fast delivery, great warehouse operations) because of ‘ideas’ that someone came up with? No… they aren’t.

And so i’ve generally concluded that there is a formula and pretty hardset rules / guidelines about how you build an ecom platform.

So hiring a CPO for an eCommerce company that is not from eCom…

…is the equivalent to hiring your landscape architect to build you a skyscraper, in my view.

Would you hire a landscape architect who has focused on building parks his whole life to build you a skyscraper? No, it is pretty logical that this would be a mistake. And yet tons of companies repeat this same mistake.. and each time i see it, i just shake my head.

I’ll give on example of a company i was contracting for some years ago, which hired a CPO from the gaming industry. I was there to support some of the fulfillment & logistics tech plans, and he had just been hired. He was chewing off my ear about ideas he had and some of the things he had done in gaming. I quietly listened for like 5–6 minutes… And after he finally paused to take a breath, I said something like…

“Ok.. but just remember…. eCommerce is a business model that has been around 20+ years. They have been operating warehouses and logistics for like 100+ years. You don’t need that many new ideas here… “.

He looked at me with that look that said… “ok… we are not gonna hit it off.” LOL

Sorry… but it’s more or less a formula. You just gotta know it and follow it… kind of like a standardized chess opening like the old “Queen’s gambit”.

Fast forward a couple years…

• The company was completely whooped by at least one of its competitors, which i know had a tech mgmt team that had been doing ecom for many years.
• That CPO was gone pretty quick.
• The CEO that hired him and pretty much that company’s entire mgmt team are all gone ive heard

Note that on top of this example.. i have at least a few others i can think of where more or less the same held true. And i still struggle to think of an exception.

So in hindsight…. does it make logical sense that if you’re following a formula, it would probably be a good idea to hire someone that knows that formula? And will not waste resources on a bunch of ideas?

Now does this hold true JUST for eCommerce? NO, the opposite holds true in my view.

A travel company once approached me about a CPO role and i immediately said no for the same reasons. How the hell could I be a CPO in travel? I don’t know shit about it…

And travel is an industry like ecom… it has been around awhile and the folks that have been years at Booking.com probably have a ton of know-how they’ve built up. So why would I assume i can just go and take my ideas from product ecommerce and beat them? It just doesn’t make sense.

Now if its some new fintech biz model where nobody really has any experience.. then yes.. I think pretty much any smart, experienced product person has a pretty good shot because you’re basically navigating unchartered territory. And there is a true need for lots of ‘experimentation’.

Are there some exceptions to my rule, probably yes.

If we look up every mature ecom that took on a CPO directly from a non-ecom I am sure you’ll dig up some exceptions…. but when you bet on a number at a craps table it also occasionally wins.

Is it the same for CTO’s? In my view no.

Why? Because in my view it is product that needs to understand the logic of the systems. We are the ‘architects’ who decide what to build and in what sequence. Whereas engineers need to be great electricians and plumbers. Can the plumber of my house install the plumbing of a bathroom on the 35th floor of the Empire State Building? Yeah probably.. and so a great engineer probably can transfer between Internet business models quite easily and lose relatively little along the way.

So to conclude…. knowing what the eCommerce foundational architecture should look like and then building it… is like knowing your end game in advance and then playing to it.