Decoy Pricing to Boost Protein Upsells: A/B Test Plan and Modeled Margin Lift

Why decoy pricing moves proteins on menus

Decoy pricing uses a clearly inferior or overpriced option to change perception of value. For operators, the goal is to shift a portion of customers from low-margin proteins to a targeted higher-margin protein while maintaining or growing average check.

Behavioral logic

Customers compare options rather than evaluate each price individually. A strategically placed decoy makes the target protein look like the best value even if it is not the lowest priced choice.

Where this fits in menu engineering

Use decoys after you have categorized proteins by profitability and popularity. This tactic is best when a target protein has competitive food cost but lower popularity than a cheaper low-margin alternative.

Three concrete decoy structures with plate pairings and numbers

• Asymmetric decoy: Menu lists Chicken Bowl $12, Target Pork Bowl $15, Decoy Steak Bowl $22. Costs: chicken $4, pork $6, steak $12. Contribution per plate: chicken $8, pork $9, steak $10. The decoy makes pork feel like the sensible upgrade because it is far cheaper than steak but visibly higher quality than chicken.
• Bundle decoy: Offer Chicken Plate $13, Chicken Plate + Side $16, Target Salmon Plate + Side $18. Costs: chicken $4, side $1.50, salmon $7. Contribution: chicken solo $9, chicken bundle $10.50, salmon bundle $11. This pairs the target with a perceived premium that is relatively small for you but meaningful for the guest.
• Comparative pairing: Present two columns: Base Bowl with Chicken $11 and Add-Protein column: Chicken +$0, Pork +$3, Lamb +$7. Costs: chicken $3.50, pork $5.50, lamb $11. The large jump to lamb turns Pork into the appealing mid-tier upgrade.

These examples prioritize a decoy that is either too expensive or too high cost for most guests so that the mid-tier target becomes the most attractive sensible choice.

Modeled margin and volume impact

Run simple per-cover contribution models before you change the menu. Example baseline and projected uplift:

Baseline mix per 1,000 covers: Chicken 60% (600), Pork 25% (250), Steak 15% (150). Contribution per plate from example 1: Chicken $8, Pork $9, Steak $10.

Baseline total contribution: (600 x 8) + (250 x 9) + (150 x 10) = 4,800 + 2,250 + 1,500 = $8,550. Contribution per cover $8.55.

Scenario with decoy nudging 8% of Chicken buyers to Pork and 2% of Steak buyers to Pork: New mix Chicken 50% (500), Pork 35% (350), Steak 15% (150). New contribution: (500 x 8) + (350 x 9) + (150 x 10) = 4,000 + 3,150 + 1,500 = $8,650. Contribution per cover $8.65.

Net change: +$100 on 1,000 covers or +$0.10 per cover. If average daily covers are 300, that is about $30 per day incremental contribution. If decoy increases average check by $1 through perceived upgrade, multiply accordingly.

Note: Small per-cover changes compound rapidly over month and year. Test effects often vary by daypart and guest segment so run segmented models.

A/B test plan operators can run next week

Primary and secondary metrics

Primary metric: percentage of covers choosing the target protein. Secondary metrics: average check, contribution per cover, units sold per protein, and guest satisfaction scores if tracked.

Sample size and duration example

Use a simple sample size estimate for a conversion change. Formula approximation: n = Z^2 * p*(1-p) / d^2. With Z=1.96 for 95% confidence, baseline p=0.15 target share, and detectable uplift d=0.05, n ≈ 200 covers per variant.

For restaurants, run the test until each variant has the required number of covers and you capture representative dayparts. That typically means 2 to 4 weeks depending on traffic.

Step by step implementation

1. Select two comparable periods or split traffic by day/terminal. Keep staffing and ops identical across variants.

2. Implement variant A with existing menu copy and variant B with decoy pricing and revised plate pairing. Update POS categories so you can track protein choices automatically.

3. Monitor metrics daily. If a variant creates operational stress such as increased prep time, pause and adjust before concluding the test.

4. Analyze: compare target protein share, average check, contribution per cover, and any change in return rate or waste. Run a simple proportion Z-test for conversion differences.

5. If positive and operationally feasible, roll to all outlets with a phased playbook including training notes for servers and cooks.

Operational considerations and guardrails

Keep portion sizes and prep steps consistent. Train staff on the language to present the upgrade. Track any change in ticket time or plate waste. Avoid confusing menus that increase order errors.

Test one variable at a time. If you change price and description simultaneously, you will not know which caused the effect.

Be ready to revert decoy pricing if it reduces guest satisfaction or if food cost volatility erodes the target margin.

If you would like the sample A/B test spreadsheet and contribution calculator I use with operators, message me and I will send the template and a quick walkthrough.

Quick checklist before you launch

Confirm POS tagging, staff brief, supply stability for the target protein, and that you have at least the minimum sample size capacity for the planned test duration.

This approach gives you a low-risk way to nudge guests toward higher-margin proteins while measuring real impact on margin and volume.

If you want help tailoring the decoy structure to your menu and food costs, reply with your current protein costs and prices and I will propose two decoy variants and an estimated margin lift.

Frequently Asked Questions

How large of a price difference should the decoy be?

Make the decoy noticeably higher than the target so the target seems like the sensible upgrade. In practice a decoy priced 30 to 60 percent above the target works well, but confirm with cost and perceived value for your customers.

Will decoy pricing upset regular guests?

Most guests respond to value framing. Keep the decoy realistic for your brand and avoid removing existing choices. Monitor guest feedback and rollback if you see negative comments or decreased satisfaction.

How do I know if the uplift is statistically significant?

Compare the change in target protein share using a proportion Z-test. If you reach the sample size for your detectable effect and the p-value is below 0.05, treat the uplift as statistically significant.

Can I run decoy tests during peak season?

You can, but ensure both variants see similar dayparts and traffic. Peak season can shorten test duration but may also amplify operational issues, so prepare staff and suppliers.