A recent conversation with a credit union CEO reinforced a pattern I continue to see across mid-market organizations: most companies don’t struggle with buying technology. They struggle with connecting activity to outcomes. Technology investments are measured at the extremes. On one end: adoption metrics, login rates, feature usage. On the other end: business outcomes like revenue growth and cost reduction. The problem is the gap between them. This is where the domino effect comes in. Technology ROI isn’t a straight line from implementation to revenue. It’s a chain reaction. Small dominoes represent daily user actions. Medium dominoes represent operational improvements. Large dominoes represent process effectiveness. The largest dominoes are business outcomes. Most organizations only measure the first domino and the last domino. They miss the middle—where ROI chains actually break.

Four enterprise platforms made nearly identical bets in one week. ServiceNow positioned itself as “the AI platform for business transformation.” monday.com rebranded as an “AI Work Platform.” HubSpot called itself “the agentic customer platform.” Salesforce continued pushing Agentforce autonomous agents.

All promising the same thing: AI agents that plan, coordinate, and execute work without human intervention.

But here’s what nobody’s saying: if your systems are broken, AI will make them worse.

AI agents require clean data, defined processes, and clear governance—exactly what system drift destroys. If your sales team tracks real pipeline in Excel because they don’t trust your CRM, adding AI agents won’t fix that. The AI learns from bad data and makes confidently wrong predictions.

Mid-market companies need to fix their foundations first. Then deploy AI from a position of strength, not desperation.