Joe Mosby's blog

Salesforce buying Informatica is a signal that enterprise SaaS is now a mature industry

He starts with a point of view that the semiconductor business has matured. The businesses must be run differently than when they were growing up.

Silver Lake partner Kenneth Hao on Broadcom CEO Hock Tan, just before the attempted Qualcomm acquisition

In early April 2024, the news broke that Salesforce was in advanced talks to buy cloud-based data management service Informatica. (UPDATE: the deal fell through, but the fact that they even tried makes the rest of the post relevant) Salesforce currently operates its own Data Cloud, a broad enough software offering that it can probably do everything. It owns automation software MuleSoft and is still in the process of incorporating its mega-acquisition of Slack. So, what does it need Informatica for?

Broadcom's CEO believes that semiconductors are now a mature industry, and as a result, businesses must be run differently. What does that "mature" designation mean, and how would it change strategy?

Mature industries share a few relevant characteristics:

  1. Technological advancements or shifts in consumer preferences have reached a plateau, and fewer new customers are entering the market as a result.
  2. There are a few well-established competitors with stable market shares and high barriers to entry.
  3. Products are highly standardized.
  4. Companies have a focus on cost efficiency and operational optimization.
  5. Regulators are involved and scrutinizing the industry.

Broadcom's top customers include original equipment manufacturers (OEMs) like Cisco, Huawei, and Dell; cloud service providers like AWS, Azure, and GCP; telecommunications companies like AT&T and Verizon; and custom buyers like financial institutions or governments. All of those customers were launched in the 1990s or earlier. Even if different companies have grown or lost market share, there hasn't been a new class of customer in twenty years. And while semiconductor chip manufacturers continue to look for ways to make faster, smaller, more efficient chips, the general design of a chip is also much the same as it was in the 1990s. We're etching circuits onto tiny pieces of silicon. And as far as regulators go, the Qualcomm acquisition referenced at the top of this post blew up because of regulators.

So Broadcom's in a mature industry. And what are they looking to do to grow? They're certainly still looking to improve their own products, just like any other competitive business, but that's not where they're seeing the massive growth potential. That comes from buying large, established chip businesses with healthy profit margins, incorporating them into the larger Broadcom infrastructure and selling off or spinning down the parts that aren't successful. They did it with LSI in 2013, Emulex in 2015, and they attempted Qualcomm in 2017.

That acquisition pattern starts to look a lot like Salesforce over the years. Is Salesforce in a mature industry?

Perhaps it would help to classify the type of business Salesforce is actually in. "Cloud software" or "software as a service" feels entirely too broad - I don't think Shopify or Figma would consider Salesforce to be a direct competitor. Salesforce have a large customer relationship management (CRM) business, which is used by most large sales teams for tracking pipeline and deals. Tableau is a data analysis tool for understanding and visualizing data, and MuleSoft is the connective glue that pulls all these systems together. Informatica is in a similar vein, focusing on data integration, quality, and governance. All of those pieces of software have the same buyer: the chief operating officer.

Consider any type of finance person you've ever worked with. They live in Excel. There are Excel e-sports competitions. If the company issues a Mac and insists that Google Docs is a corporate standard, they will run shadow IT just to make Excel work. You cannot pry it out of their hands. Microsoft will be a forgotten company one day and we will still be using Excel on the moon. Salesforce is the same way with enterprise sales teams: they might use Hubspot or some other tool for a bit when the company's growing, but they all convert to Salesforce at a certain size and scale. It is the toolkit for the sales team.

I think Marc Benioff's approaching this stage in Salesforce's life in the same way as Hock Tan with Broadcom. "Software" might still be a growing market, but when it comes to software that can be sold to a Fortune 500 COO, the industry has matured. If you're not from Salesforce, Microsoft, or one of a dwindling number of large software businesses, you're not getting in the door to talk about your business process software.

What does that mean for enterprise software plays? For one, it raises the bar for the type of vision needed to be successful. You can't simply "do a business process better," because most of the people executing those processes just think of Salesforce as part of the business process. Nor can you really compete on cost, because that's a race to the bottom, and it's often more cost-effective to simply hire Salesforce groupies rather than the type of people who are going to want to tinker with new tools. COOs are rarely measured on Salesforce implementations; they're measured on how much their sales teams are selling. The software is an afterthought.

You've instead got to re-think the way business is done, which is why I think Salesforce bought Slack in the first place. Slack was thinking differently about how people would approach work, including how they might retrieve information out of customized workflows - something that could threaten Salesforce's business process approach in the long run. Better to just buy them before they become a real problem.

Eventually there will be a new class of businesses that finds a new way of work, or who sees the old ways of working and thinking as somehow antithetical to their philosophy. But it's unlikely to happen before Marc Benioff retires, and then it'll be someone else's problem.