Angry Little Tree

A friend of mine is currently trapped in enterprise software hell. His company purchased the El Grande Enterprise Call Center Management System With Rustproofing And Chrome Trim, and what they got was a Pinto: corrupted customer accounts, endlessly backlogged customer calls, and a rising river of complaints that’s turning the holidays into a nightmare.

It’s a question that gets asked so often it’s a cliché: why is enterprise software so universally wretched? Karen McGrane once called it “The Software That UX Forgot” – filled with sharp corners and painful user experience and processes that don’t match most organization’s real workflow.

After talking to my friend and revisiting some of my own repressed memories, I think it goes even deeper. One of the distinguishing characteristics of the “Enterprise” software category, is deep ties into the actual business process of the organization rolling it out. Billing? Call center management? Inventory control? All deeply connected to an org’s actual ability to do work and make money. In a media/publishing environment, article production and similar content workflows are part of the core business as well.

If a redesigned web site is a facelift, migrating to new “enterprise software” is open-heart surgery. Fear becomes a key driver for many stakeholders because so much is on the line and – sadly – the vultures often start circling. By vultures, I mean Enterprise Sales Dudes.

Many years ago, I worked for a nonprofit organization that needed someone to maintain the long-neglected web site. That someone was young, enthusiastic me. I arrived on day one, was assigned my cubicle, and curiously prarie-dogged over the walls to look at the sweltering, server-filled corner office where “The Consultants” worked.

What were The Consultants doing? Working on “The ERP,” specifically something called “SAP.” Young grubling that I was, I had no idea what that meant and asked for clarification. Three years earlier, I was told, the organization had been struggling to keep its call center organized, its conference infrastructure running smoothly, and so on. It had grown to the point where old systems weren’t working well, and they needed newer tools to support their big plans. Newer, bigger tools.

Enterprise tools.

A friend of the organization’s President ran a consultancy in the city, and wanted to help. Magnimoniously, he offered to donate – donate! – an expensive software license for the high-end enterprise software they needed. It would require some customization, of course, to fit a few of their specific needs. He had some excellent developers the organization might hire to make a few of those tweaks before rolling out the new system, of course.

Three years later, the corner office was packed with six of those developers, working away at hourly rates on the heavily-customized Enterprise Software The Organization Needed. They’d evolved into the defacto IT department, since the SAP ERP project was intended to be the digital heart and circulatory system of the entire organization. Tentative rollouts of different parts of the system happened over the year that I spent with the org, disrupting the call center one month and conference registration the next. I’m sure that the staff eventually learned to work with the new system’s idiosyncrases, and I’m sure the bugs and hiccups were eventually ironed out. The organization saved a bunch of money, though, because that software license? Totally gratis.

It’s always been a strong memory, though: my first experience with “Enterprise Software” tasted suspiciously like a modern version of Stone Soup. Those of us who work in open source have all seen similar stories: Free software! To run your business! With just a few customizations from our team of experts…

The stakes with enterprise software are so high that perfectly smooth transitions are extremely rare. The business drivers for the migrations are such that feature lists, uptime assurances, and carefully managed specs suck the air out of the room before things like “User Experience” and “Staff Workflow” can be considered. And the budgets of the projects – due to their size and clear connection to actual business revenue streams – means that they draw opportunistic vendors like flies.

Is there a way to fix it? At the small tactical scale, sure. There are lots of individual steps that those of us who are involved with the projects can do to incrementally improve things. Shifting away from waterfall to agile, integrating real prototypes and user testing early, treating clients like humans instead of billable-hour piñatas… All are important. But it’s hard not to think, sometimes, that the category of Enterprise Software is always fated to be at least a bit of a clusterfuck. If it weren’t, I suppose everyone would be doing it.