"It costs five times more to acquire a new customer than to keep an existing one" is one of the most repeated numbers in marketing — and, like several stats this site has already checked, nobody can point to where it actually comes from. What is real and verifiable: a 1990 Harvard Business Review study that started the wider conversation about retention economics, a 2000 peer-reviewed paper that pushed back on it, and current 2025–2026 benchmark data on exactly which post-purchase messages — order confirmations, shipping updates, review requests, cross-sells — actually earn money.
This piece covers what happens after "thank you for your order": the real economics of keeping a customer versus finding a new one, what repeat buyers are actually worth, and a message-by-message breakdown of what to automate first — with the stats that didn't survive a check flagged clearly instead of repeated as fact.
Key takeaways
- The "5x cheaper to retain than acquire" number has no traceable source — multipliers range from 3x to 30x depending on which unsourced blog you land on. What is real: a named 1990 Harvard Business Review study, and current data showing repeat customers are genuinely worth more.
- Returning shoppers convert roughly 75% more often and spend about 16% more per order than first-time visitors (Barilliance, 1.3B sessions) — and at the vendor level, repeat customers can be 21% of a store's customers but 44% of its revenue (Gorgias, 12,000+ merchants).
- Shipping-confirmation emails post the highest open rate of any automated send measured (62.67%) and $3.08 revenue per email — beating even order confirmations. Cross-sell emails open reasonably well (42%) but earn the least per send ($0.95) of the core post-purchase types (Omnisend).
- Automated flows are disproportionate earners: Omnisend's 2025 data shows automations are 2% of sends but 30% of email revenue; Klaviyo's shows flows earn roughly 18x the revenue per recipient of one-off campaigns.
- Reviews genuinely move purchase decisions — 270% higher purchase likelihood with 5 reviews versus none (Northwestern's Spiegel Research Center) — but the flashy WhatsApp-specific "cart recovery" multipliers and brand-name examples circulating online couldn't be traced to any real source.
Is retention really cheaper and more profitable than acquisition?
The number gets repeated with total confidence: acquiring a new customer costs five times more than retaining one you already have. We looked for where it comes from — the way we've checked other viral marketing stats — and found no traceable study behind it. It's attributed inconsistently to Bain, Harvard Business Review, Philip Kotler and Peppers & Rogers, and the multiplier itself shifts depending on which site you land on: 3x here, 7x there, 25x, even 30x elsewhere. No two sources agree on a number, a sample, or a methodology.
What is real is the study that started the wider conversation: Frederick Reichheld and W. Earl Sasser Jr.'s "Zero Defections: Quality Comes to Services," published in Harvard Business Review in September–October 1990. Built from Bain & Company's own client work across banking, insurance and auto-service businesses, it argued that even a small cut in how many customers a company loses each year has an outsized effect on profit. Bain's own current summary of that research states the headline finding plainly: retaining just 5% more customers can increase profits by almost 100%.
- Bain's own page illustrates the finding with MBNA America: over eight years the credit-card issuer brought its annual customer defection rate down to about 5% — roughly half the ~10% industry average at the time — and, in Bain's own words, "profits have increased sixteenfold."
- The exact multiplier shifts depending on which secondary source repeats it — some cite a 25%–85% range, others 25%–95%, others 25%–100% — and no version traces to a specific, quotable page in Reichheld's follow-up book, The Loyalty Effect.
- A peer-reviewed pushback exists too: Werner Reinartz and V. Kumar's "On the Profitability of Long-Life Customers in a Noncontractual Setting" (Journal of Marketing, 2000) tested the idea against a real retail/catalog dataset and found that simply keeping a customer longer doesn't reliably make them more profitable — in businesses without a subscription or contract, revenue growth over the relationship matters more than tenure alone.
almost 100%
the profit increase Bain & Company's own current summary attributes to a 5-percentage-point improvement in customer retention — the origin of the widely (and often loosely) cited "Reichheld stat."
Source: Bain & Company, summary of Reichheld & Sasser, "Zero Defections: Quality Comes to Services," Harvard Business Review, Sept–Oct 1990
Read together, the honest summary is: retention is a real, well-documented lever with named case studies behind it — not a myth — but the exact multiplier is more folklore than fact, and "keep them longer" only pays off if you also grow what they buy.
Do repeat customers actually spend more?
Set aside the multiplier debate and ask a simpler question: does a second-time buyer actually behave differently than a first-time one? Two independent datasets say yes, with real numbers behind them — though both come from a single vendor's own customer base rather than an independent study, so treat the direction as solid and the exact percentages as one company's honest read, not a universal law.
| Metric | Value | Source |
|---|---|---|
| Returning-visitor conversion lift | ~75% higher than new visitors | Barilliance (1.3B sessions, 2016–17) |
| Returning-visitor spend per transaction | ~16% higher than new visitors | Barilliance |
| Repeat customers' share of a store's customers | 21% | Gorgias (12,000+ merchants) |
| Repeat customers' share of revenue | 44% | Gorgias |
| Repeat customers' share of orders | 46% | Gorgias |
One number worth retiring: the claim that "a store's top 10% of customers spend 3x more per order, and the top 1% spend 5x more" is repeated on dozens of e-commerce blogs, credited to an RJMetrics report from 2015. The report itself is gone — RJMetrics was folded into Adobe/Magento in 2016–17 — but its own press release survives, and it says something different: the top 1% of customers were 18x more valuable than average, and the top quartile (25%, not 10%) spent nearly 6x more in their first 30 days. Two different metrics and two different customer slices got flattened into one tidy, wrong-sounding stat somewhere along the way.
What should a post-purchase flow actually send?
Omnisend's email benchmark data — built from 2025 send activity across its customer base — breaks out performance by exactly this kind of automated message, and the pattern is worth noticing before you build anything.
| Message type | Open rate | Click rate | Revenue per email |
|---|---|---|---|
| Shipping confirmation | 62.67% | 16.01% | $3.08 |
| Order confirmation | 57.91% | 8.36% | $2.88 |
| Customer feedback | 49.17% | 4.29% | $1.14 |
| Order follow-up | 47.70% | 4.12% | $1.75 |
| Cross-sell | 42.09% | 3.02% | $0.95 |
The two highest-earning messages — shipping confirmation, order confirmation — are also the two with nothing to sell; they're pure utility. Cross-sell, the one message explicitly asking for a second purchase, earns the least per send of the five, even though it opens reasonably well. That's not a reason to skip it — $0.95 per email at scale is still real money — it's a reason not to lead with the pitch: earn trust with the utility messages first.
Why do automated flows outperform one-off campaigns by so much?
Two different vendors, two different years of data, the same conclusion: a message triggered by what a customer just did earns several multiples of a scheduled blast sent to everyone.
- Omnisend's 2026 report, covering full-year 2025 activity, tells a similar story: automations made up just 2% of all sends but drove 30% of total email-attributed revenue, earning $2.87 per send against $0.18 for scheduled campaigns — 24% higher open rates and 19x higher conversion rates than one-off campaigns.
- Klaviyo's own click-rate figures make the same point more granularly: flows average a 5.58% click rate versus 1.69% for campaigns, and the top 10% of flow-running accounts reach revenue per recipient as high as $7.79.
18x
roughly how much higher the revenue-per-recipient of an automated flow is versus a one-off campaign send, per Klaviyo's 2026 benchmark report — flows generate about 41% of total email revenue from just 5.3% of all sends.
Source: Klaviyo, 2026 Email Marketing Benchmarks by Industry
In oToK, this is a triggered automation, not a scheduled send: an order created, paid, refunded, cancelled or fulfilled in a connected Shopify, WooCommerce or Wix store can start the flow the moment it happens.
Does asking for a review actually pay off?
Reviews aren't just nice to have — they measurably change buying behaviour. In a survey PowerReviews fielded in March 2023, 99.75% of shoppers said they read reviews at least sometimes when shopping online, 91% said they do so always or regularly, and 93% said reviews impact whether they buy.
- The same research found purchase likelihood peaks at a 4.0–4.7 star average, not a perfect 5.0 — a flawless rating reads as less credible than a good-but-real one.
- On timing, the most detailed guidance comes from PowerReviews' own practitioner recommendations — not a published study, so treat it as informed advice rather than data: roughly 21 days after delivery for durable goods like electronics or appliances, 14 days for soft goods, and 7 days for anything seasonal.
270%
greater purchase likelihood for a product with five reviews compared with the same product with none, per a study from Northwestern University's Spiegel Research Center, run with PowerReviews.
Source: Spiegel Research Center at Northwestern University x PowerReviews, "How Online Reviews Influence Sales" (2017)
A couple of specific-sounding review-request numbers didn't survive a check and are worth naming so you don't repeat them: a claimed 68%/28%/4% breakdown of which request (first, second, third ask) earns the review turned out to describe how eventual reviewers were won, not an overall response rate; and a widely-repeated "70% of reviews come from post-purchase emails" statistic traces to a source that, when opened, doesn't contain that number at all. In oToK, a review request is one more branch on the same automation that already fires when an order is marked fulfilled — no separate tool needed to time it a week or two after delivery.
Email or WhatsApp for the post-purchase message?
WhatsApp is the obvious channel to reach for — in Israel specifically, 99% of the population uses it, ahead of YouTube (97%) and Facebook (88%), per the Israel Internet Association's 2025 index, making it the single most-used digital platform in the country. But the flashy stats attached to "WhatsApp for post-purchase" claims mostly don't survive a check.
A commonly cited figure claims WhatsApp order/cart messages recover 45–60% of the time versus 5–10% for email, with named brand examples seeing up to 3x higher recovery after switching. Tracing the number through its own citation chain, it fell apart at every step: the page doesn't say what gets quoted from it, its own footnoted source doesn't contain the comparison at all, and no case study for any of the named brands could be located. Treat it as marketing folklore, not evidence.
| Message category | What it's for | Requires opt-in? |
|---|---|---|
| Utility | Order/account updates tied to a specific transaction — must not promote, upsell or include offers | Yes |
| Marketing | Promotions, offers, cross-sell, win-back | Yes |
| Authentication | One-time codes and verification | Yes |
Since June 2023, Meta's WhatsApp Business Platform requires every message template to be filed under Authentication, Utility or Marketing — and a template mixing utility content with any promotional language gets reclassified (and priced) as Marketing. That's the real-world reason a shipping update and a "here's 10% off your next order" message can't be the same template: file the update as Utility, and send the cross-sell separately as Marketing, with its own opt-in already collected. The honest framing, without an invented multiplier: WhatsApp's near-universal reach fits the utility side of post-purchase — updates people are actively waiting for — while email's larger content real estate suits the review request and cross-sell messages that need more than a line of text. In oToK, a post-purchase automation can route each message to whichever channel fits and fall back to the other per contact, rather than forcing one channel for the whole sequence.
What should you actually build, in what order?
- 1Automate the utility messages first: order confirmation and shipping/delivery updates — they already earn the most per send in the benchmark data above, and nobody objects to getting them.
- 2Add a review request triggered when the order is marked fulfilled, timed a week or two later depending on the product — durable goods can wait closer to three weeks.
- 3Layer in a cross-sell or replenishment message after that — it earns the least per send of the core message types, so it belongs after trust is built, not as the first thing a new customer hears from you.
- 4Route by channel, not by habit: WhatsApp for the update someone is actively waiting for, email for the review ask and the cross-sell that needs more space.
- 5Measure per message, not per flow: the benchmark data above shows more than a 3x gap in revenue-per-email between message types in the same sequence — know which one in your own flow is actually earning.
In oToK, this whole sequence is one automation: order created, paid, fulfilled and refunded are all real triggers straight from a connected Shopify, WooCommerce or Wix store, and each contact's page shows exactly how much revenue traces back to it.
Frequently asked questions
- Is it actually cheaper to retain a customer than acquire a new one?
- The specific "5x cheaper" figure repeated everywhere has no traceable source — it's attributed inconsistently to Bain, HBR and various authors with multipliers ranging from 3x to 30x. What is real and well-documented is Reichheld & Sasser's 1990 Harvard Business Review study, whose case studies (including Bain's own MBNA example, where profits reportedly rose sixteenfold as defection rates fell) show retention driving outsized profit. A peer-reviewed 2000 study in the Journal of Marketing adds a caveat: in businesses without a subscription, simply keeping a customer longer doesn't guarantee they're more profitable — you also have to grow what they buy.
- Do repeat customers really spend more than first-time buyers?
- Real data says yes, though the exact size varies by source: Barilliance's analysis of 1.3 billion e-commerce sessions found returning visitors convert about 75% more often and spend roughly 16% more per transaction than new visitors, and Gorgias' data across 12,000+ merchants found repeat customers are 21% of a typical store's customers but 44% of its revenue. The specific "top 10% spend 3x, top 1% spend 5x" figure that circulates widely doesn't trace back cleanly to real data, though — skip that one.
- How many days after delivery should I ask for a review?
- There's no single published study pinning down an exact day, but PowerReviews' practitioner guidance (based on its own review-request volume, not a controlled study) suggests roughly a week for seasonal items, two weeks for soft goods, and closer to three weeks for durable goods like electronics — enough time for the customer to actually use the product before you ask.
- Should post-purchase messages go on WhatsApp or email?
- Both, for different jobs. WhatsApp fits the utility messages people are actively waiting for — order and shipping updates — and reaches a near-universal audience in markets like Israel, where 99% of the population uses it. Email fits messages that need more space to make their case, like a review request or a cross-sell. Widely-quoted multipliers claiming WhatsApp recovers dramatically more orders than email couldn't be verified to any real source, so treat channel choice as fit-for-purpose, not a race with an invented winner.
- What's the first post-purchase automation I should build?
- Start with the messages already earning the most per send in benchmark data: order confirmation and shipping updates. They're pure utility, nobody unsubscribes over them, and only after that should you add a timed review request and, further down the sequence, a cross-sell or replenishment message.
