In partnership with
Sound familiar?
Over 4 million people have had the same lightbulb moment.
Morning Brew is a free daily newsletter that breaks down what's happening in business, finance, and tech — clearly, quickly, and with enough personality to make it the best email in your inbox.
No yelling. No filler. Just the news, finally making sense.
GM to the Top 1% ☕
Ask ten sales leaders if their org "uses AI" and nine will say yes without blinking. Ask how many are running actual agentic workflows, agents that take multi-step action without a human clicking approve at every stage, and the room goes quiet fast.
That gap is not a rumor. It is now a measured number, and it is bigger than most people running the number expected.
💡 THE 63-POINT GAP
Recent surveys put AI usage in sales orgs at roughly 87%. Ask the same orgs how many are running true agentic workflows, meaning an agent that plans and executes multiple steps on its own, and the number drops to about 24%. Only around a third of AI sales initiatives are hitting the ROI targets they were funded on.
The instinct is to blame the models. The data says otherwise. The single most cited reason initiatives stall is poor data quality, named by roughly 53% of respondents. Not model capability. Not lack of ambition. Dirty, incomplete, inconsistently entered CRM data that no agent can reason over reliably.
This should change how you think about your own AI rollout. The bottleneck was never "which tool." It was whether your pipeline data is clean enough for a tool, any tool, to trust. An agent given bad inputs does not fail gracefully. It acts confidently on garbage, which is worse than not acting at all.
The 63-point gap between "using AI" and "running AI agentically" is sitting in your own CRM fields right now, and it is more fixable than a model upgrade.
🔧 THE HYGIENE-FIRST ROLLOUT
Four steps before you add another agent to your stack.
1. Audit one field, not the whole CRM: Pick the single field your next-step automation depends on most (close date, next step, stage) and check what percentage is stale or blank. That number is your real adoption ceiling.
2. Fix the input before the output: Do not buy a better agent to fix bad forecasting. Fix mandatory field hygiene first. A cheap agent on clean data beats an expensive agent on dirty data every time.
3. Assign one owner per field: Vague ownership is why hygiene rots. Name one person accountable for one field's accuracy, weekly, with a number attached.
4. Re-test agentic scope monthly: As hygiene improves, expand what the agent is trusted to do unsupervised. Do not grant full autonomy on day one; earn it field by field.
🎯 THIS WEEK'S HOMEWORK
Pull your own pipeline and check the close-date field on every open deal. Count how many are stale, meaning unchanged for over 30 days with no note explaining why. That count is your personal version of the 63-point gap.
❓ QUESTION OF THE DAY
What is the one CRM field where bad data is quietly capping what AI can safely do on your team?
Reply with the field name. Fixing it is worth more than your next tool purchase.
See you tomorrow.
Edward
Founder, Morning Sales
P.S. Half the manual's 500 prompts exist because clean inputs beat clever tools, and most of them are written to force you to fill in the specific detail a lazy CRM update skips. 27 dollars: https://store.edwardgorbis.com