Discount & Promotion Strategy: When Discounts Help, When They Train Bad Behavior, and How to Run Promotions That Don't Erode Your Pricing

Most founders treat discounts as a free conversion lever. A deal stalls? Offer 20% off. New customer hesitating? Black Friday code. Year-end push? Flash sale. The discount works once — that customer signs — and then six months later you've discovered three things: the discount became the expected price, you're now anchored 20% below your sticker, and the customers who paid full price two months ago found out and feel cheated. Discounts are easy short-term and toxic long-term when applied without a strategy.

A working discount strategy does specific work. It reserves discounts for situations where the discount is causal (genuinely changes the buying decision) and avoids them where the deal would have closed anyway. It protects pricing integrity (existing customers don't see surprise discounts to new prospects). It uses promotions as channels (Black Friday, annual prepay) without conditioning customers to wait. Done well, discounts close deals without eroding sticker price; done badly, discounts become the price.

This guide is the playbook for thinking about discounts deliberately — when they help, when they hurt, what types to use, the mechanics of running promotions, and the discipline that prevents discount-creep from destroying your pricing power.

What Done Looks Like

By end of the exercise:

A documented discount policy (who can give what / when)
A small set of approved discount types (annual prepay, multi-year, NPO, etc.)
An approved promotion calendar (Black Friday, anniversary, etc.)
Sales team aligned on negotiation discipline
Discount-by-segment analytics in CRM
Customer-fairness rules (existing customers honored)
Quarterly review of discount impact

This pairs with Pricing Strategy (the foundation), Pricing Page (where discounts surface), Pricing Packaging & Tier Design (tier-fit), Pricing Experiments (testing), Annual Contract Negotiation (where most B2B discounts happen), Free Trial vs Freemium (entry-mechanic), Sales Playbook (sales discount authority), Self-Serve vs Sales-Led (different discount cultures), Conversion Rate Optimization (testing), Reduce Churn (save-offer discounts), and Raise Prices (related: net pricing).

When Discounts Help vs When They Hurt

Not all discounts are equal. Get the framework before deploying any.

Help me distinguish good discounts from bad.

The honest framework:

**Good discounts** (typically positive):

| Discount type | Why it works |
|---|---|
| Annual prepay (15-20% off) | Locks in customer; cash up front; reduces churn |
| Multi-year (extra 5-10%) | Even longer commitment |
| Volume / scaling (10-30% on big seats) | Value scales with usage |
| Non-profit / education | Mission alignment; brand value |
| Founders / early customer | Genuine appreciation; small audience |
| Loyalty / renewal pricing-cap | Reduces existing-customer churn |
| First-time buyer for budget-constrained | Causal: changes the buying decision |

**Bad discounts** (typically negative):

| Discount type | Why it backfires |
|---|---|
| "Limited time" repeated quarterly | Trains customers to wait |
| "Sign by Friday" pressure | Creates resentment; rarely closes good deals |
| Discount to win against cheaper competitor | Race-to-bottom |
| Discount because customer asked | Sets precedent for asking |
| Holiday / Black Friday for B2B SaaS | B2B buyers don''t wait for holidays |
| Win-back at deeper discount than original | Conditions churn |
| Save-offer 50% off | Tells the leaver they were paying too much |

**The "would they have bought anyway?" test**:

For each discount you''re tempted to give, ask:
- Would this customer have signed at full price?
- Is the discount changing their decision, or just their payment?

If they''d have signed anyway: you just gave away revenue.
If the discount changes the decision: it might be worth it.

**The "what does this train customers to do?" test**:

For each promotion type, ask:
- Will customers expect this again?
- Does this train them to wait for discounts?

If the answer is yes: you''re damaging future pricing power.

**The "how does the existing customer feel?" test**:

For any new-customer promotion, ask:
- If a 6-month-old customer learned about this, would they feel cheated?

If yes: don''t do it, OR extend the same offer to existing customers, OR run a private offer (single-customer; no posting).

For my company:
- Discounts currently offered (audit)
- Customer-fairness audit
- The "are we training bad behavior?" check

Output:
1. The discount audit
2. The keep / kill recommendations
3. The customer-fairness fix list

The biggest unforced error: discounting because it''s easy. Sales rep needs to close; offer 20% off; deal closes. Rep takes credit. Customer becomes a low-margin account. The next deal stalls; same playbook; same discount; pattern repeats. Two years in, your effective pricing is 20% below sticker. The fix: discount discipline starts with "no" as the default.

The B2B Discount Hierarchy: Most-to-Least Acceptable

Not all discounts are equal in B2B SaaS. Use this hierarchy.

Help me prioritize B2B SaaS discounts.

The hierarchy (best to worst):

**Tier 1: Annual prepay (almost always good)**

- 15-20% off list for paying upfront
- Customer commits 12 months; you get cash + reduced churn
- Low risk; high upside
- Per [annual-contract-negotiation](annual-contract-negotiation.md)

**Tier 2: Multi-year (excellent at enterprise)**

- 5-10% additional for 2-3 year commit
- Locks in for longer; reduces re-negotiation
- Use sparingly with smaller customers (commit risk)

**Tier 3: Volume / scale discounts**

- E.g., 10% off at 50+ seats; 20% at 100+
- Built-in to pricing tiers
- Predictable; transparent

**Tier 4: Mission-aligned (NPO, education)**

- 30-50% off (consistent industry practice)
- Brand value; mission alignment
- Verified eligibility (501(c)(3); .edu domain)

**Tier 5: Founder / early-customer discounts**

- "Founder pricing" for first 10-50 customers
- Locked for life or for X years
- Genuine appreciation
- Document carefully (set expiration)

**Tier 6: Sales-discretion discounts (use sparingly)**

- Rep authority to offer up to X%
- For specific situations: budget-constrained but high-fit
- Each one slightly erodes pricing
- Track per rep; review quarterly

**Tier 7: Save-offer discounts (avoid; if used, controlled)**

- Customer trying to cancel; offer 30% off to stay
- Risk: trains threats-to-cancel as discount-extraction
- Acceptable if: only first-time use; specific reason; clear duration
- Per [reduce-churn](reduce-churn.md)

**Tier 8: Time-pressure discounts (avoid for B2B)**

- "20% off if you sign by Friday"
- Rarely works for sophisticated buyers; creates resentment
- Sometimes OK for SMB where decision-cycles are short

**Tier 9: Race-to-bottom (don''t)**

- Beat competitor X''s price
- Anchors customer to lowest price; no path back

**Tier 10: General public-promo "Black Friday SaaS sale"**

- Trains customers to wait
- Anchors prices low for future buyers
- B2B almost never benefits

**The B2B SaaS rule**:

Use Tier 1-5 freely.
Use Tier 6-7 with discipline + review.
Avoid Tier 8-10 entirely.

For my product:
- Which tiers are in use today
- The "we''re using these but shouldn''t" list
- The "we should add" list

Output:
1. The current-state map
2. The keep / add / kill list
3. The discipline rules per tier

The biggest hierarchy mistake: mixing tiers without distinguishing. Treating all discounts as equivalent — annual prepay (smart) vs. "we''ll match competitor''s lower price" (toxic) — uses the same approval process and the same effect on sales-rep behavior. The fix: tier them; different approval per tier; different review cadence.

Annual Prepay — The Discount That Almost Always Works

Annual prepay is the discount you should default to. Get the mechanics right.

Help me design annual-prepay discounts.

The mechanics:

**Standard offer**:

- Monthly billing: $X/mo
- Annual prepay: $X × 10 (~17% off)

In other words: "pay for 10 months; get 12."

This produces:
- 17% effective discount (10/12)
- Cash upfront (12 months)
- Reduced churn (annual customers churn ~50% less)
- Reduced billing ops

**Why ~17% (10-month equivalent)?**

- Round numbers are easier
- Visibly "20% off" feels generous
- Per industry standard (Slack, Notion, Airtable, etc.)

Alternative: 20% off (pay for 9.6; get 12) — slightly more generous

**The "would-be churners" math**:

If 10% of monthly customers churn each month:
- After 12 months, ~28% of original cohort retained
- Annual customers can''t churn until renewal

The 17% discount typically pays for itself in churn-reduction alone, even before considering cash-flow benefits.

**Display on pricing page**:

Two common patterns:

**Pattern A: Toggle**

> Monthly | **Annual (Save 20%)**
> $79/mo or **$66/mo (billed annually)**

**Pattern B: Single price with note**

> $79/mo *(or $790/yr — save 17%)*

Pattern A converts better; Pattern B is simpler.

**The "auto-renew" question**:

Most SaaS auto-renew annual contracts. This is fine if:
- 30-90 day advance notice email
- Cancellation easy
- Per [Stripe Customer Portal](stripe-payments.md) self-serve

If cancellation is painful: customers complain; reputation suffers.

**Annual-to-monthly downgrade requests**:

When a customer wants to downgrade from annual to monthly mid-cycle:
- Honor existing annual (no refund); switch to monthly at renewal
- OR: offer prorated refund minus discount-claw-back ("you got 20% off; we deduct that from refund")

Decide policy; document clearly.

**The exception: per-seat annual**:

For per-seat plans, annual gets tricky:
- What if they need to add seats mid-year? (Add at monthly rate; or pro-rate annual)
- What if they need to remove seats? (No refund typically; honor through renewal)

Document policy; train sales.

**The "do annuals make sense for us?" check**:

Annual prepay works best when:
- ACV justifies it ($500+/yr typically)
- Buyer has budget approval cadence
- Self-serve: clean discount mechanic
- Sales-led: standard practice

Skips:
- Free / starter tier
- Trials
- Per-API-call usage-based pricing (annual hard to bill)

For my pricing:
- Annual offer presence
- Discount level
- Renewal policy

Output:
1. The annual-prepay offer
2. The renewal policy
3. The display strategy

The biggest annual-prepay mistake: not offering it at all. Many indie SaaS leave the easy 20-30% conversion lift on the table. Customers who want the lower price prefer annual; customers who don''t are unaffected. The downside is minimal; the upside is real.

The Promotion Calendar: Run Promos Without Damage

Some promos are worth running. Schedule them deliberately.

Help me build a promotion calendar.

The legitimate promotion windows:

**Anniversary / company milestones**:

- 1-year anniversary
- 100th customer / 1000th customer
- Major funding milestone (carefully)

Acceptable: limited-window discount tied to genuine event.

**Product launches**:

- New tier / feature launch
- Beta-to-GA transition
- Major version release

Acceptable: limited-time launch pricing for early adopters.

**Customer events**:

- Annual customer conference
- Webinar series finale

Acceptable: tied to event participation.

**Industry events**:

- Conference sponsorships → conference-specific discount code
- Per [conference-and-event-marketing](../3-distribute/conference-and-event-marketing.md)

Acceptable: limited audience; conference attribution.

**Holiday / seasonal (B2B context)**:

- End-of-fiscal-year (December)
- New-fiscal-year kickoff (January)
- Per [annual-contract-negotiation](annual-contract-negotiation.md)

Acceptable: B2B fiscal cycles align.

**Avoid**:

- Black Friday / Cyber Monday (consumer cadence; not for B2B)
- Random "summer sale"
- "Just trying to hit quarterly numbers"

**The 1-2x per year rule**:

Don''t run more than 1-2 customer-facing promotions per year. More than that:
- Trains customers to wait
- Existing customers feel cheated
- Erodes pricing power

Sales-led discounts (negotiated) can happen more often; public promos should be rare.

**The promotion mechanics**:

For each promotion:
- Specific code (`LAUNCH2026`, `CONFA`)
- Specific window (start/end dates)
- Specific audience (public / event / private)
- Specific reason (tied to event / launch)
- Existing-customer treatment (extend or not)

**The "fairness to existing customers" rule**:

When you announce a public promotion:
- Email existing customers: "We''re running X promotion; here''s how it affects you"
- If beneficial to extend: extend
- If not: explain (often: "this is for new customers; here''s why")

Customers respect honesty. They resent surprise discounts to others.

**Track impact**:

- Conversions per promotion code
- Revenue from promotion-eligible customers
- Existing-customer churn during promotion (do they feel betrayed?)

For my plans:
- Year-1 calendar
- Existing-customer communication
- Track + measure

Output:
1. The 12-month promotion calendar
2. The mechanics per promo
3. The fairness rules

The biggest promotion mistake: running promotions reactively. "Sales is short this quarter; let''s do a flash sale." Reactive promos signal desperation; they train customers to wait; they erode pricing. Plan promotions a year out; tie them to genuine events; resist the reactive flash-sale temptation.

Sales-Discretion Discounts and Approval Authority

For sales-led motions, individual discounts happen. Build the policy.

Help me design sales-discount authority.

The structure:

**Discount tiers and approval**:

| Discount level | Approver |
|---|---|
| 0-10% | SDR / AE without approval |
| 10-20% | Sales manager approval |
| 20-30% | Sales VP approval |
| >30% | CEO + finance review |

This prevents creep; rep can''t self-approve big discounts.

**The "give-get" rule**:

Reps should always trade discount for something:
- 10% off → pay annually
- 15% off → 2-year commit
- 20% off → multi-year + case study + reference

Discount alone is bad; discount in exchange for something is fine.

**The "discount narrative"**:

For every discount, the rep documents:
- Reason (why this customer needs it)
- Trade (what we got in exchange)
- Expected impact (close rate uplift)

Reviewed monthly. Patterns emerge:
- "We discount to win against [competitor]" → consider product / positioning
- "We discount for budget-constrained" → consider cheaper tier

**The "discount tracking by rep"**:

In CRM, track:
- Avg discount per rep
- # discounts per rep
- Closed-won at discount vs. full price

Reps with high discount-rates often haven''t learned to negotiate value. Coach them; don''t just praise the wins.

**The "discount-on-discount" trap**:

Year 1: customer signs at 15% off
Year 2: renewal at full list — they balk; offer 10% off (still discount)
Year 3: 5% off
Year 4: full list — they''re mad at the increase

Each year, the "increase" feels like a price hike to them, even though it''s just discount-removal.

The fix: at signing, document expected progression: "15% Y1; 10% Y2; full at Y3 — confirmed?"

Or: keep the discount level static across renewals; raise other tiers; their relative value grows.

**The "competitive deal" exception**:

When you need to win a deal against a cheaper competitor:
- Don''t match price (race to bottom)
- Differentiate on value
- If discount is needed: smallest possible; documented as competitive
- Track: competitive losses don''t systematically discount

**The audit trail**:

For every discount > 10%:
- CRM record with reason
- Approval chain documented
- Customer-facing reason ("annual prepay" / "non-profit" etc.)
- Quarterly review

For my sales:
- Current discount authority
- Average discount per rep
- The reform plan

Output:
1. The discount approval matrix
2. The give-get rules
3. The tracking + audit plan

The biggest sales-discretion mistake: unlimited rep discount authority. Reps optimize for closing; closing is easier with discounts; pricing erodes silently. The fix: cap rep authority; require approval beyond cap; track aggressively.

Customer Fairness — The Existing-Customer Question

Discounts to new customers can damage the relationship with existing ones. Manage carefully.

Help me handle customer fairness.

The risk:

- Customer X signed 6 months ago at $99/mo
- New customer Y signs today at $79/mo (Black Friday)
- X finds out; X is angry; X may churn

This is a real, recurring problem.

**The solutions**:

**1. Don''t run public discounts (best)**

If discounts only happen via sales negotiation (private, per-deal), no fairness issue. Different customers got different terms; that''s normal.

**2. Extend public discounts to existing customers**

When you announce a 20% off promotion:
- Email existing customers offering same discount on next renewal
- Or credit them
- Customer feels: "company is fair; I''ll stay"

**3. Honest disclosure**

If you ran a one-time promotion that some existing customers missed:
- Acknowledge it
- Explain why (tied to specific event)
- Offer good-will credit for top customers

**4. Anniversary pricing**

Existing customers get a guaranteed renewal discount as anniversary perk. Reduces churn. Acknowledges loyalty.

**5. Customer-tier-specific promos**

Promotions tied to existing-customer benefit:
- "Add a seat at 50% off" (only for current customers)
- "Annual prepay credit if you upgrade now" (only for current)

These reward loyalty without alienating.

**The "don''t price-discriminate publicly" rule**:

Public price-discrimination (different prices on website depending on customer cohort) damages trust:
- New visitors see lower price than logged-in customers
- Logged-in customers see "upsell to lower-tier"

Avoid. Pricing-page should show same prices to everyone.

**The "leaked discount" recovery playbook**:

When existing customer learns about new-customer discount:

1. Acknowledge: "Yes, we ran X promotion."
2. Explain: "It was tied to [reason]."
3. Offer (if appropriate): "Here''s how we''ll make it right: [credit / next-renewal-pricing / etc.]"
4. Don''t deny / dismiss.

Most customers appreciate honesty more than the dollars.

**The "would I be okay with this customer talking about my discount publicly?" test**:

If a customer who got a great deal mentioned it on Twitter: would other customers feel cheated?
- If yes: too generous; existing customers lose trust
- If no: deal was appropriately scoped

For my discounts:
- Audit fairness gaps
- The recovery plan if needed
- The communication template

Output:
1. The fairness audit
2. The fixes
3. The communication templates

The biggest fairness mistake: silent price-discrimination assumed nobody will notice. Customers talk; especially in B2B (community, conferences, online). When they discover the difference, they remember it for years. The fix: assume every discount becomes public; design accordingly.

Save-Offer Discounts: When to Use, When They Backfire

When a customer threatens to cancel, the save-offer discount is tempting. Use carefully.

Help me handle save-offer discounts.

The setup:

Customer hits "cancel" → cancellation flow asks why → customer cites cost → offer "stay for 30% off?"

**The argument FOR save-offers**:

- Lifetime value > immediate revenue
- Reduces churn metric
- Some customers genuinely budget-constrained

**The arguments AGAINST (often outweighs)**:

- Trains threats-to-cancel as discount-extraction
- Existing-customer fairness destroyed
- The customer who paid full price for 18 months and didn''t threaten to leave got nothing
- Often the customer churns 3 months later anyway (price wasn''t the real issue)

**The conditional play**:

Save-offers can work when:

- First-time-only (one save-offer per customer ever)
- Time-limited (X months only, then back to full)
- Behavior-conditional (must use product; not pause + use again)
- Reason-conditional (genuine budget cut; not standard request)

Document carefully. Track save-offer recipients separately.

**Better alternatives to discount save-offers**:

1. **Pause subscription** (vs cancel) — customer stays in DB; resume later
2. **Downgrade to cheaper tier** — keeps them; less revenue but real
3. **Dedicated check-in call** — find the real reason
4. **Address the actual problem** (most cancellations aren''t price; they''re value)

A 1-on-1 call from a CSM solves more cancellations than 30% off.

**The "what does this churn say?" question**:

Per [reduce-churn](reduce-churn.md):
- High churn = product / value / fit problem
- Discounting can''t fix that; just delays inevitable

If save-offer rate is climbing, that''s a signal of bigger issue, not a need for more discounts.

**The "save-offer sustainability" math**:

If 20% of customers churn → save-offer 30% off to half of them:
- 10% saved at 30% off
- Net: revenue down 3% on full base; new customers also see lower expectations

Compare to: investing the same effort in onboarding / activation / value:
- Lower churn fundamentally
- No pricing erosion
- Compounding effect

The discount is a band-aid; the fix is upstream.

For my flow:
- Current save-offer policy
- Cancellation reasons audit
- Alternatives prioritized

Output:
1. The save-offer policy
2. The non-discount alternatives
3. The "fix the upstream cause" plan

The biggest save-offer mistake: defaulting to discount-on-cancellation. It''s reactive; it doesn''t address why; it trains gaming. The right move: cancellation → human conversation → diagnose real issue → solve OR allow graceful exit OR pause. Discount is a tool of last resort, not first.

Discount-Free Selling — The Best Outcome

The best discount strategy is requiring no discounts to close.

Help me sell without discounts.

The principles:

**1. Anchor on value, not price**

Discount discussions happen when value is unclear. If buyer feels strong value: price isn''t the question.

Sales-rep training: lead with outcomes, not features. Per [sales-demo-calls](sales-demo-calls.md).

**2. Qualify hard**

The deals that need discounts are often deals that shouldn''t close at any price. Qualify aggressively:

- Budget?
- Need urgent?
- Authority to sign?

If "no" to budget → not your customer right now.

**3. Walk-away willingness**

Reps who can walk away close at full price. Reps who can''t walk away discount.

Train walk-away: "It sounds like our pricing is a hurdle; we''d love to keep in touch when budget aligns."

Some deals close 6 months later at full price.

**4. Premium positioning**

Lower-priced products discount; premium products don''t.

If you''ve positioned as the premium choice (per [competitive-positioning](../1-position/competitive-positioning.md)), discounting confuses positioning.

**5. Strong references**

When customer is uncertain: case studies + reference calls (per [customer-references](customer-references.md)) close deals without discount. Social proof beats price discount.

**6. Pilot / proof of value**

Some deals need a smaller commitment to reduce risk:
- 90-day pilot at smaller scope
- Same per-unit price as production
- No discount; less commitment

**7. Better tier-fit**

Sometimes "need a discount" means "need a different tier":
- Cheaper tier with fewer features
- Per-seat instead of flat
- Pay-as-you-go vs commitment

Tier-fit > discount.

**The "discount ratio" KPI**:

Track:
- % of deals closed at full list price
- Avg discount on discounted deals

Goals:
- Self-serve: 95%+ at list (any deviation is a bug)
- Sales-led SMB: 80%+ at list
- Sales-led mid-market: 60-70% at list (some flex expected)
- Enterprise: 50%+ at list (negotiation expected; just keep it bounded)

If %-at-list is dropping: investigate.

For my sales:
- Current %-at-list
- Discount-free playbook gaps
- The training plan

Output:
1. The %-at-list baseline
2. The discount-free plays
3. The KPI target

The biggest discount-avoidance mistake: treating "discount" as a sales tool rather than as a strategic option. Discounts are sometimes correct; they should never be reflexive. The discount-free playbook (qualify, anchor value, references, tier-fit) closes more deals at higher margins than the discount playbook ever will.

Avoid Common Pitfalls

Recognizable failure patterns.

The discount mistake checklist.

**Mistake 1: Discount as default sales tool**
- Rep''s first move; not last resort
- Fix: train discount discipline

**Mistake 2: Public Black Friday for B2B**
- Trains customers to wait
- Fix: B2B promo cadence; tied to events

**Mistake 3: Save-offer reflex**
- Cancellation → 30% off
- Fix: human conversation; address real issue

**Mistake 4: Existing-customer betrayal**
- New customer gets discount; existing finds out
- Fix: fairness rules; transparent extension

**Mistake 5: Race-to-bottom against competitor**
- Match competitor''s lower price
- Fix: differentiate on value; walk away

**Mistake 6: No discount approval levels**
- Reps self-approve unlimited
- Fix: tiered approval matrix

**Mistake 7: Discount creep over time**
- Year 1: 10%; Year 2: 15%; Year 3: 20%; can''t recover
- Fix: track + audit; reset annually

**Mistake 8: Renewal-time deeper discounts**
- Signaling annual price hikes
- Fix: stable discount levels through renewals

**Mistake 9: Promo-rich pricing page**
- Always-on discount banners
- Fix: clean pricing page; promos are exceptional

**Mistake 10: Discount over fixing value problem**
- Customers leaving because of value; discount band-aid
- Fix: address upstream

**The quality checklist**:

- [ ] Documented discount policy
- [ ] Tiered approval matrix
- [ ] Annual prepay offered
- [ ] Mission-aligned discounts (NPO/edu) clear
- [ ] Save-offer policy with conditions
- [ ] Existing-customer fairness rules
- [ ] Sales-rep training on walk-away
- [ ] Track: % at list price
- [ ] Track: avg discount per rep
- [ ] Quarterly discount review

For my approach:
- Audit
- Top 3 fixes

Output:
1. Audit results
2. Top 3 fixes
3. The "v2 discount policy" plan

The single most-common mistake: never thinking about discounts as a system. Each discount is decided ad-hoc; rep has authority; reasons aren''t documented; trends aren''t reviewed. The result: pricing-power erodes silently. The fix: a policy, an approval matrix, a quarterly review. Discounts become deliberate, not reflexive.

What "Done" Looks Like

A working discount strategy in 2026 has:

Documented policy (who can give what, when, why)
Annual-prepay default discount (15-20% off list)
Mission-aligned discounts (NPO / education) clear
Tiered approval authority for sales discounts
Save-offer rules with conditions
Existing-customer fairness honored
1-2 public promotions per year max
Sales team trained on walk-away
KPIs tracked (% at list; avg discount)
Quarterly discount review
"No" as the default answer to "can we discount this?"

The hidden cost of weak discount strategy: pricing power evaporates one deal at a time. A 5% discount this quarter; a 10% next; a 15% the quarter after. Each individual discount is tactically defensible. The cumulative effect: net price 30% below sticker, customers anchored low, raise-prices conversations impossible. Pricing is fragile; discount discipline is its preservation. Treat every discount like it''s permanent (because effectively it often is).