Forbes: Everyone's Talking About Vibe Coding—It's Time For Vibe Planning

Read the article on Forbes.com.

"Vibe coding" exploded this year.

Computer scientist Andrej Karpathy captured the ethos succinctly: Describe what you want in plain language, let AI generate most of the code and iterate conversationally. The concept is not only gaining traction among startups; large companies are leaning in, too. Leaders report a fast rise in AI-generated code across engineering teams, showing how quickly "prompt-to-product" is becoming normal.

While we haven't reached the point where businesses are run on "vibes," more and more companies are open to using agentic AI. According to a May 2025 PwC survey, 79% of senior business executives say that they've already deployed agentic AI in their business.

One key area where it may make sense to incorporate agentic AI is in integrated business planning (IBP) practices across various functions, including revenue, service levels, working capital, sustainability and brand. Think of this like "vibe" planning, where planning teams can leverage real-time data and an AI-automated IBP loop that allows them to get a feel for the market and make data-informed decisions with governance.

Vibe planning fuses three elements:

1. Sensing The Vibes (Weak Signals): These include social chatter, short-cycle demand pulses, store-level point-of-sale (POS) data, etc.

2. Structured Planning: One set of numbers is used to make decisions across demand, supply, finance, procurement and other business functions.

3. Agentic Execution With Guardrails: AI proposes scenarios, policies constrain actions and humans approve service level agreements (SLAs) with clear decisions.

The Five Stages Of The Vibe-Planning Loop

1. Sensing

An AI-enabled planning platform ingests multi-horizon signals (e.g., POS, orders, marketing calendars, social and weather) into a knowledge graph of products, locations, customers and suppliers. At this stage, AI agents continuously flag anomalies (e.g., a spike in 12-packs in Houston or a packaging lead-time risk in ASEAN).

2. Hypothesizing

AI agents propose a demand hypothesis (a 10% shift from vanilla cookies to a limited-edition SKU, for example), potential supply moves (e.g., alternate line plus weekend overtime) and financial implications (gross margin). Each hypothesis references the evidence and expected key performance indicator (KPI) lift or risk.

3. Simulating

AI agents can showcase potential scenarios such as constrained versus unconstrained supply, promotion versus baseline, carbon-aware routing versus fastest-ship, and also take into account sensitivities across service, inventory and earnings before interest and taxes (EBIT).

4. Deciding

Planners and stakeholders can review a decision brief that outlines set assumptions, confidence, trade-offs, KPI deltas and policy checks (e.g., capacity caps, minimum safety stock, sustainability thresholds). This document sets who approves what and determines a decision time frame, so choices don’t age out while markets move.

5. Acting And Learning

Approved scenarios become executable plans across supply, allocation, pricing and deployment. AI agents monitor the outcomes versus expectations, update priors and write back to a decision log for audit and post-game learnings.

With this structure, IBP with planners at the helm remains the core function, but AI agents can augment the process by accelerating scenario creation. As a result, planning teams can accelerate their creative speed without sacrificing the discipline of IBP.

What Makes Vibe Planning Different From Classic IBP

There are a few distinctions between vibe planning and traditional IBP practices. Vibe planning reduces the potential for siloed trade-offs because finance, commercial, supply chain and sustainability teams can see the same impact curve (e.g., service versus inventory versus emissions versus margin) before committing.

Vibe planning also moves the reconciliation stage from a monthly cadence to a real-time loop. IBP remains a single source of truth among cross-functional teams, but the reconciliation phase never stops as scenarios are updated continuously as signals change.

Vibe planning also removes manual planning in individual spreadsheets and instead replaces this with explainable agents that provide scenario narratives that cite evidence, not just numbers. As a result, each recommendation for decisions includes the "why this, why now?" context, allowing planners to make better decisions at a faster pace. To keep up with the speed of business changes, companies may want to consider incor

Key Practices To Incorporate Vibe Planning Into The IBP Process

For organizations that are ready to incorporate AI agents into their planning processes, here are some practices to keep in mind:

• Define your vibe signal library. Start with six to 10 high-value feeds. These can include POS, returns, promo calendars, weather warnings, supplier ETAs, competitor price watching or other relevant signals.
• Codify policies early. Before unleashing AI agents into the planning process, write the rules to provide the necessary parameters. This can include capacity, minimum order quantity (MOQ)/minimum order value (MOV), service tier priorities, carbon limits or cash constraints.
• Stand up decision types with SLAs. Some SLAs to consider include "Promo uplift re-plan," "Allocation flip during stockouts," "Packaging lead-time shock" and "Region-weather disruption." Assign approvers, clocks and escalation paths.
• Never let agents change "live" without policy and approval. The software world learned this the hard way; enterprises should avoid the same mistake in planning.
• Make explainability non-negotiable. Every agent recommendation must carry evidence, constraints and confidence.
• Rely on the decision log. Every big call or decision needs to capture the necessary context, assumptions, potential scenario deltas, approvals and realized outcomes. This information becomes your institutional memory (and audit trail).
• Start narrow; expand fast. Pilot this process with one category, one channel or one market. Prove the cycle time and KPI lift, then roll out horizontally to additional categories, channels and markets.
• Train your workforce for the craft. Just as AI coding still rewards engineers who truly understand code, AI planning rewards teams who understand demand/supply/finance mechanics. The work is real—and yes, cognitively taxing—so equip your employees with the necessary skills.

Vibe planning through agentic AI can bring quantifiable value across the business. Agents sense weak signals, translate them into decisions, simulate trade-offs and enforce decision SLAs so approvals keep pace with volatility. You keep the creativity and momentum of the agent era while protecting the enterprise with shared truth, policy and accountability.

Keep the plan tight, but let the vibes flow. That’s how modern operators win.