Most behavioral-econ posts treat persuasion techniques like spells. Cast them at the user, watch conversion go up. After running hundreds of A/B tests across SaaS pricing, onboarding, and lead-gen flows, I can tell you that's not how it works. Some persuasion techniques are real and consistently move metrics. Some only work in specific conditions. And some make things measurably worse when you misapply them.
Door-in-the-face — make a big ask first, get rejected, then make the real ask — is one of the techniques that *does* survive A/B testing. But not under every condition. This post covers seven real-world examples of door-in-the-face from the kind of experiments I run and review, plus the failure modes I see most often.
What door-in-the-face actually is (in one sentence)
You ask for something deliberately too big. The other person says no. You then ask for the smaller thing you actually wanted, and they're more likely to say yes than if you'd just asked for the smaller thing directly.
The mechanism is reciprocity: the asker has "conceded" by retreating to a smaller request, and the responder feels social pressure to also concede by saying yes. The original Cialdini research found roughly 50% compliance for the small ask when preceded by a rejected large ask, versus around 17% when the small ask came first.
Those numbers are old and from a specific context (volunteering hours with juvenile offenders). I've seen the effect replicate in some commercial contexts and fail in others. The replication conditions are what this post is about.
Example 1: Enterprise pricing with a "platinum" tier that nobody buys
The clearest commercial application I've seen is pricing pages.
A B2B SaaS company runs A/B tests on its pricing page. One variant shows three tiers (Starter / Pro / Business). The other variant shows four tiers (Starter / Pro / Business / Enterprise) where "Enterprise" is priced 3-4× the Business tier and has a "Contact us" button.
Conversion on the Business tier — the one the company actually wants people to buy — tends to be 10-25% higher with the Enterprise tier visible, even though almost no one clicks Enterprise.
The Enterprise tier is functioning as the door-in-the-face large ask. Visitors see it, mentally reject it, and the Business tier feels reasonable by comparison. The technique works here because the contrast is anchored on a price they were already considering.
It stops working — I've watched this happen — when the Enterprise tier is priced absurdly higher (10×+). Then visitors mentally bucket the whole company as "not for me" and bounce instead of converting on Business.
Example 2: Nonprofit fundraising with the "alternative ask"
I know an analytics team that ran this test for a national nonprofit.
Variant A asked donors to commit to a recurring $50/month gift. Variant B asked donors to commit to a one-time $1,000 gift, and after they declined, a follow-up screen asked for the recurring $50/month.
The completion rate on the $50/month was meaningfully higher in Variant B — somewhere in the 15-25% lift range, based on what was reported back. The conversation flow felt heavier (more drop-off at the $1,000 ask), but the people who made it to the second ask converted at much higher rates than people who saw the smaller ask first.
This is door-in-the-face working exactly as the original research predicted. The reason it replicates in fundraising is the same reason it replicated in the original volunteering experiment: the asker is perceived as making a real concession.
Example 3: Sales outreach with a "best case" first ask
Practitioner version of door-in-the-face I've seen in outbound sales: a sales rep emails a CMO with a deliberately big ask — "would you be open to a 45-minute strategic discussion next week with our CEO?" — knowing the answer will be no.
The follow-up is "completely understand — would 15 minutes with one of our solution engineers next week work?"
Reply rates on the 15-minute ask are systematically higher when preceded by the 45-minute ask than when the 15-minute ask is the first email. The catch: the technique stops working in the same campaign if the rep does it three times in a row. Recipients pattern-match it as a sales script.
I'd estimate the effect persists for somewhere between one and three rounds, then it's gone. Beyond that, the only thing left is the perceived effort to write you back.
Example 4: Trial-extension upsells in SaaS
Here's an experimental application I think is underused: SaaS trial-extension flows.
A user's free trial is about to expire. The team is testing two variants:
• Variant A: "Your trial ends in 3 days. Upgrade to Pro for $49/month to keep your data."
• Variant B: "Your trial ends in 3 days. Upgrade to Business ($199/month) for full team access, or stay on Pro at $49/month."
In several SaaS contexts I've seen, Variant B converts the $49 Pro plan at a higher rate than Variant A — because the $199 Business plan is the implicit "big ask" that makes the $49 feel like the reasonable retreat.
This only works when the Business plan is visibly bigger and the user has at least some plausible reason to consider it (so it doesn't feel artificial). If you're a single-user-only product and you flash "Business — $199/month — 10 user seats" at someone, they roll their eyes. The contrast needs to be plausibly relevant.
Example 5: Survey-completion uplift
A small but reliable effect: in user-research surveys, asking respondents to commit to "a 30-minute follow-up call next week" first, then asking them to "just complete this 5-minute survey" instead, gets meaningfully higher completion than asking for the 5-minute survey directly.
I've seen this used in product-team research workflows where they actually want both — the survey AND a follow-up call. The reframing as "if not the call, will you take the survey?" lifts both completion rates.
Example 6: Hiring negotiations (from the candidate side)
Door-in-the-face shows up in salary negotiations, intentionally or not. A candidate asks for $200K base. The hiring manager says they can't do that, but offers $175K with a sign-on bonus. The candidate would have happily taken $160K base going in.
This is the version of the technique that makes both sides uncomfortable because the "big ask" is real (the candidate would take $200K if offered), so neither side is doing the polite-fiction version. But the structure — anchor high, retreat to a smaller number, both parties feel like they've conceded — is identical to the textbook.
The failure mode in salary negotiation specifically: anchor wildly higher than your market and the hiring manager doesn't engage at all. They just pull the offer. The mechanism requires a plausibly serious first ask.
Example 7: Email-list opt-in with a "free trial" first
Newsletter signups follow the same logic. The double-ask variant — "want our 7-day email course?" first, and if declined, "just our weekly newsletter?" — outperforms a single "subscribe to our newsletter" CTA in several lead-gen contexts I've seen tested.
Worth noting: I've also seen this fail spectacularly when the "big ask" was a paid product. "Buy our $200 course; or, just subscribe to the newsletter" tanked newsletter signups vs. a direct newsletter ask, because visitors mentally categorized the brand as commercial.
When door-in-the-face backfires
After watching the technique work and fail across dozens of test contexts, three failure modes recur:
1. The big ask isn't plausible. If the user can tell the big ask was theater — that you knew they'd never say yes — the reciprocity mechanism doesn't fire. The retreat to the small ask doesn't feel like a concession; it feels like a sales script. Pricing tiers priced 10× higher than relevant break this way. So do sales outreach scripts that overuse the pattern.
2. The user doesn't see the big ask as related to the small ask. Reciprocity only works when the small ask is perceptually a retreat from the big ask. "Will you donate $1,000?" → "$50?" works because both are donations. "Will you upgrade to Business?" → "Subscribe to our newsletter" doesn't, because they're different categories.
3. The audience pattern-matches it. B2B audiences in particular have seen the technique. If your prospect has read the same blog posts you have, they'll see the structure and lose interest. The technique works better with audiences that haven't been over-exposed — and audiences that haven't been over-exposed are increasingly rare in 2026.
My operating rules
If I'm advising a team about using door-in-the-face in a marketing test:
1. The big ask has to be real. Either you'd take it if offered, or the user would plausibly consider it. No theater. 2. The retreat has to be in the same category. Pricing → pricing. Donation → donation. Survey → survey. 3. The audience has to be naive to the pattern. B2C consumers and one-off donation contexts work best. Enterprise B2B buyers with sales-veteran instincts work worst. 4. One use per relationship. Burn it once, the second use is obvious.
If you can satisfy all four, door-in-the-face replicates as a real conversion lift in my experience. Outside those conditions, it's noise.
The bigger point about persuasion in experiments
Door-in-the-face is unusual among behavioral-econ persuasion techniques in that it survives A/B testing in commercial contexts. A lot of others don't.
When I'm reviewing test ideas with founders or growth teams, my filter isn't "is this technique well-known in behavioral econ literature?" It's "does the mechanism actually fire under the conditions of this particular test?" The four rules above are how I check.
The wider lesson — and the one I keep returning to in CRO work — is that persuasion techniques are real, but they're conditional. The mechanism either fires or it doesn't, depending on context. The job of an experimentation leader is to know the conditions, not just the techniques.