September 16, 2026 · 9 min read

Writing review-request emails people actually answer

Most review-request 'best practice' has no real source. What verified data says about timing and subject lines, and the CAN-SPAM rule most businesses skip.

The request went out. A few days pass. Two replies come back out of forty, and you start wondering whether it was the subject line, the timing, or whether anyone reads business email at all anymore.

Some of that is fixable. Subject lines, sender identity, and send time all have real, if uneven, data behind them. But before getting into what works, it is worth naming what does not hold up — because most of what circulates as review-request "best practice" turns out, on inspection, to have no real source at all.

The timing rule everyone repeats and nobody sourced

"Send the review request within 24 to 48 hours." Or, depending which vendor blog you land on, "within 3 to 7 days." Both appear, stated as settled fact, across dozens of review-software sites. Neither traces to a named study, a disclosed sample size, or a publisher willing to put their own data behind it. It functions as an axiom, repeated because everyone else repeats it, not because anyone measured it.

The closest real research points the opposite direction from "there's a magic window." A 2025 Management Science paper studied a natural experiment on TripAdvisor, where a cohort of hotels began actively soliciting reviews through a certified-partner program while a matched control group did not (Gao, Wang, Ding & Guo, 2025). Soliciting hotels saw review volume rise 34.3% and average rating rise 0.151 points — but review length fell 16.9%, reviews became more generic, and unsolicited review volume from the same customer base fell 15.5%. The authors call it motivational crowding-out: asking suppresses the reviews people would have written on their own. Separately, peer-reviewed survey-methodology research on repeated email solicitation has found the same pattern outside the review context — re-contacting the same recipient too often lowers future response and raises attrition, while send-timing itself shows little measurable effect (Journal of the Academy of Marketing Science, 2023).

Google's own guidance is silent on the question in a telling way. The Google Business Profile content policy bans soliciting reviews in bulk, selectively asking only customers expected to leave positive feedback, offering incentives, and — as of the 2026 update — using in-store kiosks or shared tablets to collect reviews on the spot. It says nothing about how fast to ask or how many times. Google tells you what not to do and leaves cadence alone entirely.

So: no primary source establishes an optimal hour count, and the one real causal study on the topic says the bigger risk is asking too often, not too late. The practical version — ask once, tied to something concrete the customer will recognize (the job finished, the order arrived), rather than racing a clock nobody has actually validated.

What has real data behind it: subject lines

Two behavioral studies, both dated but real, measured personalization directly. Yes Lifecycle Marketing analyzed over 7 billion emails sent in Q2 2017 and found subject lines personalized with a name had a 21.2% open rate against 14.1% for non-personalized ones — roughly a 50% relative lift, with click-to-open 58% higher (PRNewswire, Yes Lifecycle Marketing, 2017). Experian's 2013 Email Market Study found personalized promotional emails had 29% higher unique open rates and 41% higher click rates than generic ones (Experian, 2014 release). Neither has a recent, transparent successor — I looked for a 2024–2026 study repeating this measurement and did not find one — but the direction is consistent across both, a decade apart.

For a baseline of what "normal" even looks like, Mailchimp's own send-log data (not a survey) puts the average open rate across all industries at 35.63% and average click rate at 2.62%, based on its own delivered campaigns (Mailchimp email marketing benchmarks). Compare your own numbers against that, not against a vendor's marketing page.

Sender identity is where the data thins out. It is widely asserted that the "From" name matters more than the subject line — Litmus's own blog makes that claim — but neither Litmus nor anyone else I could find publishes the underlying percentage with a disclosed sample. Specific figures circulating online (a name-based sender boosting opens by 27%, or 19%, or 45% of subscribers naming sender as their top factor) trace to no real methodology in any case I checked. Treat the direction as reasonable and the specific numbers as unverified.

What is verifiable is a structural point, not a statistic: an email arriving from a generic no-reply address reads differently than one from a name the recipient recognizes, with a reply path that actually reaches someone — the business's own name in the From field, and Reply-To pointed at the business's real inbox, rather than a shared platform address.

Send time: weekday mornings, with one caveat

Two large vendor datasets, neither specific to review-request email, converge on a similar window. MailerLite's own analysis of 2,138,817 campaigns sent December 2024 through November 2025 found peak open rates on weekday mornings (9–11am recipient-local), with one exception: Friday at 6pm scored highest of the week overall, the one slot where morning-favoring opens and evening-favoring clicks line up (MailerLite, best time to send email). Omnisend's analysis, described as roughly 26 billion emails though without a disclosed date range, found Tuesday through Thursday, 9–11am, performed best, with Monday weakest (Omnisend, best time to send email).

Neither company segments its data by message type, so there is no primary source specifically for review-request timing. Weekday mornings are a reasonable default drawn from general email behavior, not a guarantee for this particular kind of message.

The legal question almost nobody asks

It is easy to assume a review request, sent to someone who already bought something, counts as a "transactional or relationship" message and sits outside most of CAN-SPAM. The FTC's own compliance guide warns against exactly that assumption: "You shouldn't assume that any message you send to recipients who have an ongoing commercial relationship with you... are transactional or relationship messages" (FTC, CAN-SPAM Act: A Compliance Guide for Business).

The statute defines that exemption narrowly, covering only messages that complete an agreed transaction, deliver warranty or recall information, notify of a change to an existing account or membership, or relate to employment (15 U.S.C. § 7702). A review request does not do any of those things — it promotes the business's own reputation, which is what makes a message "commercial" under the Act's own definition. Under the FTC's primary-purpose test, that puts most review-request email in the commercial bucket, with the full compliance list attached — not because the FTC has ruled on review requests by name, but because they do not fit any of the exempted categories.

This is general information, not legal advice. Rules vary and change — check with a lawyer about your own program.

Source for all six: 15 U.S.C. § 7704 and the FTC's compliance guide linked above. The maximum penalty is $53,088 per individual email in violation (90 FR 5580, effective 17 January 2025 — unchanged into 2026, since the government-wide inflation adjustment was cancelled that year). "Per email" is the detail that matters: in a review-request program sending to hundreds of customers, the exposure scales with volume, not incidents.

This is also why a review-request tool built compliance-first attaches a real sender name, a working reply path, and a one-click opt-out to every message by default. ReviewHero ships with all three built in rather than left for a business to bolt on later.

What we could not verify

A few numbers show up constantly in this space and none of them held up to checking:

• "9x more reviews from automated requests", commonly attributed to Yotpo — not found anywhere on yotpo.com; no dataset behind it.
• "Reminder emails collect 35% more reviews", commonly attributed to Trustpilot — checked Trustpilot's own blog post and its transparency report directly; neither contains this figure.
• "Response rates drop 60–70% after two weeks", again attributed to Yotpo — same problem, unfindable on Yotpo's own site.
• Specific sender-name lift percentages (27%, 19%, 45%) — repeated across marketing blogs, traceable to no named study in any case checked.

None of these are cited above, and none should be repeated as fact.

What to do next

1. Send the request once, tied to a concrete trigger — the job finished, the order arrived — rather than chasing a specific hour count no study actually established.
2. Put the business's real name in the From field and set Reply-To to an inbox a person actually reads.
3. If you send a reminder, send exactly one, and give the first message real time to be seen before the second goes out.
4. Default to a weekday morning send; treat evening sends as something to test, not assume.
5. Check your email program against the CAN-SPAM list above — sender accuracy, physical address, a working opt-out — even if the message feels transactional. Building it correctly once costs less than one complaint later.