Overview
The Pricing Agent finds margin opportunities for you by reviewing your book, comparing your pricing against a peer set, and surfacing specific items and accounts where you have room to raise price. You stay in control and can accept, reject, or modify each recommendation.
What the agent surfaces
The agent generates up to 100 recommendations per rep, each pointing to a specific item on a specific customer where your pricing may sit below a comparable peer set. Every recommendation shows where you have room to move price up without overexposing yourself.
How the agent decides what to recommend
The agent looks at a peer set of similar customers, grouped by volume tier and price tier, and compares your pricing for each item against that group. Where you are significantly below the peer average, it flags the opportunity and suggests an increase.
If your distributor sends customer segments through the ERP integration, the agent uses those segments to define the peer set instead. Over time, the agent also factors in each customer's price sensitivity based on ordering behavior, so it gets smarter about how aggressive to be.
To start, the agent leaves a customer's top 5 to 10 highest-volume items alone. Those items carry sharper pricing on purpose, since they help hold margin across the rest of the catalog, and reps push back when a tool suggests touching them. So the agent works lower in the basket first, and eases into top-item recommendations over time.
Cost change recommendations
Alongside peer-based opportunities, the agent surfaces recommendations tied to cost changes:
When a product's cost goes up, it can suggest holding your gross profit dollars while adjusting margin.
When a product's cost goes down, it can suggest holding your last price to capture the spread.
Each recommendation's rationale tells you whether it is a margin opportunity or a cost change, so you always know why it is flagged.
Tip: when you reprice after a cost change, use round up or random round up to land on a clean number. (See: How Peer Range Benchmarking Works.)
How to use the Pricing Agent
The agent does the analysis. You make the call on every recommendation. Here is the flow from a completed run to a live price change.
1. Open a completed run. The agent runs on a set cadence and lets you know when a new set of recommendations is ready to review.
2. Read the rationale before you act. Every run shows what the agent looked at and how it got there, so you never accept a number blind. Each recommendation names the item and customer, flags whether it is a margin opportunity or a cost change, and shows how your price compares to the peer set. Ask two things: do you buy the reasoning, and is the size of the change right for that customer?
3. Decide on each recommendation. You have three choices on every line:
Accept: take the agent's suggested price as is. The change applies as a price override on that item for that customer.
Reject: dismiss the recommendation and keep your current price. Nothing changes. Reject when you know something the agent does not, for example a customer relationship reason to hold the price where it is.
Override: set your own price instead of the agent's. Use this when you agree there is room to move but want a different number, for example landing partway between your current price and the suggested one.
4. Work faster with bulk actions and quick rounding. You can go through recommendations one at a time, or select several at once to accept or reject them together. When you accept or override and want a clean number, use round up or random round up to land on a natural-looking price.
5. Review everything at first. Treat every recommendation as something you review yourself rather than something that applies on its own. As the agent learns which kinds of recommendations you consistently accept, you will be able to let more of those through with less hands-on review and spend your time on the ones that need real judgment.
6. Publish to apply your changes. Your decisions do not take effect until you publish. When you hit Publish, Pepper asks you to confirm, showing how many recommendations you accepted and how many you rejected. Once you confirm:
Accepted changes overwrite the current prices for those customers.
Rejected recommendations are recorded as feedback and removed from the list, which helps the agent learn what you turn down.
Until you publish, nothing changes for your customers.
