Price Approval Workflows for Retail Teams

Automated pricing rules are powerful. They monitor competitors, apply your strategy logic and generate price recommendations in real time. But the question every retailer faces after setting up their first rule is: should these changes go live automatically, or should someone review them first?

If you are building your pricing automation strategy, our complete guide to dynamic pricing covers the strategic framework. This post focuses on the operational layer - the approval workflows that sit between your pricing rules and your live storefront.

The short answer is that most retailers should not fully automate price changes. Here is why, and what to do instead.

Why full auto-push is risky

On paper, automatic price updates sound ideal. Your rules detect a competitor price change at 2 AM, adjust your price by 3 AM and your storefront is competitive before anyone wakes up.

In practice, three things go wrong regularly:

1. Bad competitor data. Competitor prices are scraped from external websites. Sometimes the scraper picks up a sale price instead of the regular price. Sometimes a competitor lists the wrong price by mistake. Sometimes a product page changes format and the scraper returns a cached value from three weeks ago. If your rule auto-pushes based on bad data, you could drop a $200 product to $40 before anyone notices.

2. Rule interaction effects. When multiple rules apply to related products, they can create cascading changes that individually look small but collectively shift your category pricing in unintended ways. A 3% drop on Product A triggers a related rule on Product B, which affects Product C. Each change is within bounds, but the net effect is a 12% margin erosion across the category.

3. Competitor pricing errors. Competitors make mistakes. If a major competitor accidentally prices a $500 appliance at $50 and your rule matches them, you could sell inventory at a massive loss before the competitor corrects their error. This happens more often than you would expect.

These are not hypothetical scenarios. Every retailer that has run fully automated pricing for more than six months has at least one story about an auto-push that went wrong.

The L1/L2 approval model

The most effective approval structure we see in practice is a two-tier model:

L1: Pricing analyst review

The pricing analyst is your first line of review. They understand the rules, the competitive landscape and the product categories. L1 review covers:

• Sanity checking price changes that exceed the auto-push threshold
• Verifying that the competitor data driving the change is accurate
• Approving changes that are within normal operating parameters
• Escalating changes that are unusual, high-value or potentially sensitive

In most retail organisations, the L1 reviewer handles 80-90% of queued price changes. The review takes seconds per item because the pricing engine surfaces the rule logic, the competitor data point and the projected margin impact in a single view.

L2: Category manager approval

The category manager gets involved when a price change meets certain escalation criteria:

• The change exceeds a defined dollar or percentage threshold
• The product is in a protected category (premium brands, MAP-regulated products)
• The change would move the product's price to the lowest or highest in the market
• The rule has flagged a potential data quality issue

L2 review is less frequent but higher stakes. The category manager brings commercial context that the pricing analyst may not have - upcoming promotions, vendor negotiations, seasonal strategy shifts.

Setting your auto-push thresholds

The threshold for auto-push versus manual review depends on your risk tolerance, your average order value and your team's capacity to review changes.

Here is a framework that works for most mid-market retailers:

Percentage-based thresholds

• Under 3% change: Auto-push. These are minor adjustments that keep you competitive without materially affecting margins. At $100, a 3% change is $3 - unlikely to cause meaningful harm even if the data is slightly off.
• 3% to 5% change: L1 review. The pricing analyst checks the competitor data and approves or rejects. Most of these will be approved within minutes.
• Over 5% change: L2 review. The category manager evaluates the commercial implications before the change goes live.

Dollar-based thresholds

For high-value products, percentage thresholds alone are not sufficient. A 4% change on a $2,000 product is $80 - that deserves more scrutiny than a 4% change on a $20 product.

• Change under $5: Auto-push regardless of percentage
• Change $5 to $20: L1 review
• Change over $20: L2 review

Combined approach

The most robust configurations use both percentage and dollar thresholds. A change must pass both checks to auto-push:

• Auto-push only if the change is under 3% and under $5
• L1 review if either threshold is exceeded
• L2 review if the change is over 5% or over $20

This prevents a 2% change on a $1,000 product ($20) from auto-pushing, which a percentage-only threshold would allow.

Configuring approval workflows in PryceScan

PryceScan's pricing engine supports configurable approval workflows at the rule level. When you create or edit a rule, the Approval section lets you:

1. Set the auto-push threshold - define the percentage and dollar limits for automatic execution
2. Assign L1 reviewers - select team members who receive queued changes for first-level review
3. Assign L2 reviewers - select category managers or senior staff for escalated changes
4. Configure notifications - choose whether reviewers get email alerts, in-app notifications or both
5. Set review deadlines - define how long a change can sit in the queue before it expires (prevents stale recommendations from being approved days later when market conditions have changed)

Review queue workflow

When a price change exceeds the auto-push threshold, it enters the review queue with full context:

• The product name and current price
• The recommended new price and the percentage change
• The rule that generated the recommendation
• The competitor data point that triggered it
• The projected margin at the new price
• A timestamp showing when the competitor data was collected

The reviewer can approve, reject or modify the recommendation. Modified recommendations are logged separately so you can track how often manual overrides occur and whether they improve outcomes.

What happens without approval workflows

Retailers who skip approval workflows typically encounter one of these outcomes within the first quarter:

The margin erosion problem. Without review, rules optimise for competitiveness at the expense of profitability. Small, individually reasonable price drops accumulate into significant margin compression. By the time the monthly P&L reveals the issue, weeks of suboptimal pricing have already occurred.

The trust problem. When a pricing error goes live - and it will - the team loses confidence in the automation. Instead of fixing the approval workflow, they turn off the rules entirely. The business reverts to manual pricing, having wasted the setup investment.

The customer trust problem. Customers who see wild price swings lose confidence in your pricing. A product that was $149 yesterday, $89 today and $139 tomorrow looks like a pricing error, even if each price was technically correct based on competitor movements.

Scaling your approval process

As your rule coverage grows, the volume of queued changes increases. Here is how to scale without adding headcount:

Tighten your rules. Rules with tighter bounds generate fewer changes that need review. If 60% of your queued changes are being approved without modification, your auto-push threshold is probably too conservative. Widen it slightly and monitor the results.

Use category-level thresholds. Commodity products can have more aggressive auto-push thresholds than premium products. There is no reason to apply the same 3% threshold to a $12 cable and a $1,200 laptop.

Review in batches. Rather than reviewing changes one at a time as they arrive, encourage your L1 reviewers to batch their reviews into two sessions per day - morning and afternoon. This is more efficient and provides a natural rhythm for the team.

Track override rates. If an L1 reviewer approves 98% of queued changes without modification, the threshold is too conservative for that category. If they reject or modify 30%, the rule itself may need adjustment.

Connecting the pieces

Approval workflows are one part of a broader pricing operations framework. If you are still deciding between rules-based and AI-powered pricing, our comparison guide explains why rules are the right starting point for most mid-market retailers. If you have not yet built your first rule, our step-by-step tutorial walks you through the entire process.

And if you are concerned about the legal and ethical dimensions of automated pricing, our guide to dynamic pricing legality and ethics covers the regulatory landscape and best practices for transparency.

The goal is not to remove humans from pricing decisions. It is to focus human attention where it adds the most value - on strategic decisions, unusual situations and high-stakes changes - while automation handles the routine adjustments that keep you competitive every day.