Manufacturing the new IT stack: unbundling the ERP

Samson and Delilah, José Echenagusía, oil on canvas, 1887

Hi. My name is Renan, and I’m currently heading the most awesome startup studio of Europe, specialized in manufacturing, based in Paris and called OSS.

In this organization, technologists, designers, coders and strategists are creating the future of manufacturing along with some of the best entrepreneurs of our time.

You may find the lines below interesting because :

1. It’s dealing with a little-explored topic, the dominance of ERP over the IT stack of manufacturing, that is effectively stifling innovation through manufacturing ;
2. It draws parallels with a trend that has been observed through past years and called “the great unbundling” ;
3. It has some pretty graphics and funny jokes.

🤔 Let’s talk consumer first

The last decade has seen some of the most incredible value creation through the relatively new business model of platformization. Platformization, basically, is the act of serving as a middle party between two ends of value creation (typically a seller and a buyer). Most common examples are Uber, Airbnb, Monster.com, but also Visa or Tinder or the yellow pages.

So, in 2020, consumers are increasingly using platforms to organize an efficient meeting of offer and demand. But before going further, let’s ask several questions. What took us so long ? And why is Uber a different company than Airbnb ?

In 1995, some visionaries understood the profound effect that the Internet would have on pretty much everything we do : dating, transportation, negotiation, getting a job, finding a permanent of temporary home … One of those visionaries was Craig Newmark, the founder of Craigslist. He put together one of the first platforms, and it was the most generalist one :

The Craigslist platform

On Craigslist, back in 1995 and now, you can :

• find childcare ;
• negotiate a new car ;
• find a bicycle ;
• Discuss ;
• Date ;
• Rent a room ;
• Buy a house ;
• Find a lawyer ;
• Perform about 90 other types of offer meeting demand.

Yet, while Craigslist is 25 years old and generates about one billion revenue, it is more than 100 times smaller than Uber, Airbnb, Tinder, Monster.com or any of the “specialized” marketplaces around.

All the vertical marketplaces that got created and their corresponding Craigslist category.

So, what happened? Why isn’t Craigslist one of the biggest companies in the world?

Well, what happened is known as the Great Unbundling of Craigslist and has been theorized by a brilliant blog post by Spark Capital.

The case for Craigslist, made as a silent agreement between Craigslist and its users, was compelling :

1. As people are used to Craigslist for one thing, they can switch to another thing for virtually no cost (it’s called horizontal synergy) and carry on the data and profile they created for themselves (it’s roughly called interoperability) ;
2. As Craigslist makes money on one thing, they can invest it to develop one other thing without requiring external capital and get the best at it (it’s called financial synergy) ;
3. As a user on Craigslist, I have to enter my information one time only and get a smooth experience over different verticals (the argument for convenience and interoperability).

The reality is, as often, more complex. The consensus among technologists is that several main factors drove the rise of verticalized marketplace :

• The promise of vertical synergies between different clients did not deliver on its premise, and the synergies between different sellers, which represent the majority of the cost of acquisition, were null ;
• The efficient organization of the meeting between offer and demand raises vertical-specific issues that have to be solved at scale by a talented team of designers, data scientists over all of the cycle of the meeting (for example for Airbnb: creating a compelling house listing, training the host, training the guest, choosing the right house for the right guest, organizing check-in, organizing communication, creating trust, create a compelling checkout experience, create operation-specific support operations, getting relevant insurances, …) and VC money fueled every single one of these verticals ;
• The surprisingly hard to attain magic number of sellers and buyers over which there are a sufficient number and concentration of buyers and sellers to prevent a “ghost town effect” in which a given user does not see enough offers to get what she wants ;
• The rise of open standards (think google or apple login, iOS apps, HTML5) permitted to get interoperability without needing to be under the same roof as a company.

Thus, over the 2000s and the 2010s, the great unbundling of Craigslist happened and tremendous value was created, teaching us some important trends :

• Horizontal synergy is trumped by the requirements and focus necessary to successfully addressing a vertical need correctly by entering into the value proposition ;
• Financial synergy hid the fact that it’s very hard to focus different teams on fundamentally different issues (housing and transportation comes to mind) and that a focused company actually can deliver more ;
• Magic numbers and vertical-specific issues made it so gaining a critical vertically-concentrated mass of user was more efficiently done in a specialized space ;
• Interoperability is not an argument when there are open standards.

In short, in the West, verticalized unbundled marketplaces won the war.

🏭Let’s switch to manufacturing and ERPs

In the West, in most legacy economy companies, ERPs (Enterprise Ressource Planning) systems are put in place. In this game, there are basically two big actors, Oracle and SAP.

The promise of those actors, during the last twenty years, was simple :

1. A user in your company gets a single ID, a thing in your company gets a single ID, and by putting all your business software in my walled garden, you get synergy and control (it’s called horizontal synergy, remember ?) ;
2. We are among the world’s biggest companies and can develop at scale software that will be top-notch and that a) you could not do by yourself, b) smaller companies could not deliver with a satisfying level of quality, c) our suite will be easier to maintain thus reducing the headcount of the IT team and d) we ensure maximal security ;
3. Users in your company will get one smooth experience across our suite and won’t have to bother with multiple systems.

As such, inside legacy companies, SAP and Oracle are dominant, the IT strategy is more often than not “We’re an SAP/Oracle company”.

SAP website and its 80 different pieces of software

So, have SAP and Oracle won? Is the game over and “ Platform verticalization is for consumer-facing, while bundling is for companies ?”

The counter-arguments to unbundling, in favor of the status quo and the ERP staying bundled are strong :

• In B2B software, the user is not the buyer, and the buyer can be tricked into taking a bad decision that will stick with the company for years while the decisionmaker will only stay at this place for several years. In other words, distribution is a hell of a sticky advantage of incumbents over innovators ;
• Legacy IT teams have low skills in the new technological stack and are locked in proprietary ecosystems (with their own training, skill certifications, conferences, and jargon) thus having a strong incentive to maintain status quo, making them powerful allies of incumbents, while CEO average technological literacy makes them unlikely to challenge this status quo ;
• Interoperability fueled by open standards has been consistently pushed back by incumbents through complex implementation ad-hoc coding, which effectively maintains the status quo and creates a strong barrier to entry for innovators by creating a switch cost.

But users are complaining, new innovators are starting to peel off entire bricks of SAP, no company founded in the last 10 years is powered by an ERP, even conservative Gartner describes the entire ERP landscape as “ripe for disruption”, and SAP and Oracle are amongst the largest startups buyers in Europe, consistently buying the startups that are successfully entering one of their many markets.

The arguments for successful unbundling, though, are also very compelling :

• As every company becomes a technological company, organizations are taking back control and capabilities over their software stack. At a different pace (the more disrupted the industry, the faster), technologically literate CEOs and CTOs are appointed and get decisional power over technologically illiterate CEOs and system administrators locked in proprietary walled gardens ;
• Talented innovators are finding niches in which switching costs are close to null, value delivered is high, and are de facto creating open standards. Venture capitalists money is flooding through manufacturing 4.0 revolution and other buzzwords ;
• Talented innovators are also targeting midsized companies who have no legacy ERP with higher-quality SaaS (Software As a Service) that will eventually be pushed to the BigCos after successfully cornering the midmarket which is not yet penetrated by ERPs ;
• The move to the Cloud, pushed by powerful new actors such as Microsoft and Google, counter-balances the discourse of SAP vendors and we may see open standards flourish through this movement, the same way marketing and retail created open standards. Security, scalability concerns of the buyers can be addressed by building on top of those giants eager to win new markets ;
• Users are increasingly dissatisfied with the software pushed by those companies and the gap between the level of quality of verticalized software that they are using in their daily lives, the level of quality of verticalized software that is available for small and medium enterprise, and what the reality of software is in large companies is increasingly difficult to explain.

So, which will it be ? Will ERP get successfully unbundled ? Will it stay bundled and the BigCos will successfully navigate that new status quo by buying their way into startups before they flourish ?

Our conviction at OSS is simple :

1. ERP will get successfully unbundled, and BigCos vendor will limit their offers to key processes (financial, supply chain, and HR payroll/management) ;
2. The unbundling will start with innovators proposing SaaS and platforms to midsize companies and after successfully cornering that part of the market, going upwards and start gaining significant market share of larger companies which will by then have the capabilities and mindset required for such a change. Those innovators will be verticalized giants that expand their initial category tenfold ;
3. The total value created will be tremendous and interoperability will finally get achieved thus laying the software foundation for a real 4.0 manufacturing revolution.

Agree? Disagree? Let us know in the comments !

If you want to work with us or know more, please drop by at our Parisian office @ 10 rue Chabanais, 75002 Paris. Or tweet me @siponza. Or email renan@opeo-studio.fr