The Forty-Year Arc of Business Software

To see where operations is going, you have to see where it’s been. Business software has moved through four distinct eras in forty years, each one absorbing the last. Understanding the arc is the only way to understand why this moment is different.

Era One: Digitize the Filing Cabinet (1985-2005)

The first era was about getting paper off desks. ACT! shipped in 1987. QuickBooks shipped in 1992. Peachtree was already running in accounting departments. The job of the software was simple: take what was on paper, put it on a screen, make it searchable.

The software was a container. A human did all the work, and the software just remembered what the human did. The cashier still rang up the sale. The bookkeeper still entered the invoice. The salesperson still made the call. The software just held the record afterward.

Era One ended when the records started to need to talk to each other.

Era Two: Connect the Filing Cabinets (2005-2020)

Salesforce, founded in 1999, became the dominant force of Era Two by selling the cloud-based CRM as the connective tissue of a business. HubSpot launched in 2006 with the same premise for marketing. The vertical CRMs followed for the trades: Jobber in 2011, ServiceTitan in 2012, Housecall Pro in 2013. ServiceTitan in particular did not just connect the filing cabinets for HVAC and plumbing. It became the operating system the entire industry now runs on.

The job of the software got bigger. Link the records. Surface the pipeline. Trigger the reminder. Show the dashboard.

But the human was still the engine. The CRM told you what to do. You still had to do it. Era Two replaced the filing cabinet with a screen, and replaced the rolodex with a contact list, but it did not replace the work. It just made the work easier to see.

Era Three: Automate Between the Cabinets (2012-2024)

Zapier launched in 2012. Integromat (now Make) launched the same year. Workato in 2013. A thin layer of glue grew up between every SaaS application in the world. If-this-then-that. When a form is submitted, send a Slack message. When a deal closes, create an invoice.

This was useful at the edges and brittle at the core. The automation layer could move data between systems on a schedule. It could not decide anything. It could not handle the weird stuff. It could not run a business. It could only run a workflow.

Era Three did something else, too. It conditioned an entire generation of business owners to think automation meant if-this-then-that. Triggers and actions. Rigid pipes. That mental model, more than the tools themselves, is the thing Era Four is breaking. When an owner today hears “we automate your operations,” they picture a Zap. What is actually arriving is something the Zap could never be.

Era Four: Operate the Cabinets Autonomously (2024-?)

The date matters. ChatGPT launched in November 2022. GPT-4 followed in March 2023. By 2024, the first production-grade agent frameworks (LangChain, CrewAI, the early Agentforce previews) had crossed the line from research demo to deployable system. That is the floor of Era Four. Not the day the technology was invented, but the day it became something you could put in front of a paying customer.

For the first time in the history of business software, the software can do the work. Not just hold the record of the work. Not just remind a human to do the work. Not just move data between workflows. Actually do it. Decide. Route. Escalate. Execute. Close the loop. Learn from what went wrong.

Each of the prior three eras absorbed the era before it. Era Four absorbs the dashboard, the integrator, and the human at the screen. All three.

That is why this moment is different. The first three eras improved the office. The fourth one ends it.

The rest of the manifesto is what gets built in its place.