TapCub is live — analytics, insights and live chat, free on one platform
Growth & Marketing

Customer Appreciation: Specific, Timely and Checked Against a Holdout

Customer appreciation works when it is specific, timely and judged by repeat behaviour, not compliments. Triggers and timing, gestures and their limits, and a worked holdout example showing a promising result that is not yet proven.

Sophia Bennett Published Nov 19, 2026 11 min read Growth & Marketing
Sample data: 1,000 customers reach a third order; 500 randomly receive a note and early access, 500 do not. 135 vs 110 repurchase (+22.7% relative, 25 orders, $800 margin against $700 cost). The gap gives z about 1.84, p about 0.066: promising, not proven.Sample data

Customer appreciation is recognising something a specific customer did and thanking them, without asking for a purchase. It works when specific, timely and proportionate, and you only know it worked by comparing repeat purchases with a random holdout. Below: triggers and timing, gestures and gift limits, and a worked example that looks promising but is not yet proven.

What you will learn

  • Name what the customer did and send close to the event; a line that fits everyone is a newsletter
  • Judge it by repeat purchases against a random holdout, not by compliments or replies
  • A 5-point gap on 500 per group was not significant (p about 0.066); size the holdout before you start

What is customer appreciation, and how is it different from a promotion?

Customer appreciation is recognising something a specific customer did, such as a fifth order, a renewal, a referral or a patient wait, and saying thank you in a way that shows you noticed. A promotion asks the customer to buy something. Appreciation comes first, and it works because it is not a request.

The difference shows up in what triggers it and what you expect back:

AppreciationPromotionLoyalty program
TriggerSomething the customer didYour calendar or stockPoints or tier reached
Message"Here is what we noticed""Here is an offer""Here is your reward"
What you expectGoodwill; possibly a repeatA purchase nowRepeat purchases over time
How to judge itRepeat purchase against a comparison groupSales during the offerProgramme margin and repeat rate

The thesis of this guide: appreciation should be specific, timely and judged by whether it changes repeat behaviour. Warm words are not a result. The wider set of retention tactics is in customer retention examples, B2B reward schemes are in B2B loyalty programs and the account-management side in customer success strategy. This article is about the thank-you itself and how to check it.

What makes appreciation feel real?

Three properties separate a thank-you people remember from one they delete:

  1. Specific. It names the thing: the order count, the year, the issue they were patient with. A generic message could have gone to anyone.
  2. Timely. It arrives close to the event, while the customer still remembers it. A thank-you for a renewal that arrives after the next invoice reads as billing.
  3. Proportionate. The gesture matches the relationship. A lavish gift after a first small order feels odd; silence after three years feels worse.
GenericSpecific
"Thank you for being a valued customer.""Your fifth order shipped today. Thank you for coming back."
"We appreciate your business.""Thanks for staying with us through the migration last month; it took longer than we said."
"Happy anniversary!""Two years ago today you set up your first project. You have since added 14 teammates."
"You've earned a reward.""You referred two colleagues this quarter. Here is early access to the feature you asked about."

If a line could be sent unchanged to every customer, it is a newsletter, not appreciation.

Whom should you thank, and when?

Pick triggers that are real events in your data, so each message can be specific:

TriggerWindow to sendExample gesture
Second order (the first repeat)Within 2 days of deliveryShort note and a tip
Fifth order or a spending milestoneWithin 2 daysEarly access, small gift
Customer anniversaryThe day or the weekPersonal note from a named person
Referral that convertsWithin 2 daysThanks from the person who got the referral, plus a reward
Review or feedback givenWithin a dayReply that says what you changed
Long or difficult support case resolvedAfter resolutionApology and a follow-up from a person
RenewalBefore the next invoiceThanks with what has improved since they joined

These windows are suggestions; there is no research-backed number. What matters is that the gesture lands while the event is fresh.

What can you give, and what should you check first?

GestureTypical costWatch out for
Personal note or short callTimeMust come from a real person who knows the account
Early access or a feature previewLowDo not promise what you cannot ship
Small giftLow to mediumBusiness gifts have tax limits; recipient policies
Discount or creditMarginIt becomes a promotion; do not name it appreciation
Event or dinnerHighCheck recipient rules on gifts and hospitality
Public recognitionLowAsk permission first

For business gifts in the United States, the IRS limits the deduction to $25 per recipient per year; see the Gifts section of Publication 463. Many companies also have their own rules about what employees may accept, so a gift can create a problem for your customer rather than a thank-you. If the thank-you is an email that also promotes something, the CAN-SPAM requirements apply, including a working opt-out; the FTC guide notes there is no exception for business-to-business email.

GestureCostWatch out for
Personal note or callTimeMust come from someone who knows the account
Early accessLowDo not promise what you cannot ship
Small giftLow to mediumTax limits; recipient policies
Discount or creditMarginIt is a promotion, not appreciation
Event or dinnerHighRecipient rules on hospitality

US business-gift deduction limit: $25 per recipient per year (IRS Publication 463).

Did the thank-you change repeat behaviour?

The only way to know is to compare thanked customers with similar customers who were not thanked, chosen at random. Comparing recipients with everyone else is misleading, because you probably chose the best customers to thank.

Sample data: 1,000 customers reach their third order. You assign 500 at random to receive a handwritten note and early access to a new feature, and 500 to a holdout who receive nothing special. Sixty days later:

GroupCustomersRepurchased within 60 daysRepeat rate
Thanked50013527.0%
Holdout50011022.0%

The difference is 5.0 points, a relative lift of 5.0 ÷ 22.0 = 22.7%, and 25 extra repurchases (135 − 110). If an order is worth $80 at a 40% margin, each is worth $32, so the extra margin is 25 × $32 = $800. The notes cost $1.40 each, 500 × $1.40 = $700. Net: $800 − $700 = $100, a return of 100 ÷ 700 = 14.3%.

Before declaring success, check whether the difference could be noise. The pooled repeat rate is 245 ÷ 1,000 = 24.5%; the standard error of the gap is the square root of 0.245 × 0.755 × (1 ÷ 500 + 1 ÷ 500), about 0.0272; z = 0.05 ÷ 0.0272 ≈ 1.84, which corresponds to p of about 0.066 (the large-sample test is described in the NIST handbook). A 95% interval for the gap runs from about −0.3 to +10.3 points. In plain words: this looks promising, and it is not proven. To detect a true 5-point gap reliably (80% power, 5% significance) you would need roughly 1,158 customers per group. The honest conclusion is "worth repeating with a bigger holdout", not "appreciation pays 14%".

app.tapcub.com

Repurchase within 60 days

sample data, % of 500 customers
Thanked27.0%
Holdout22.0%
Sample data
Sample data: 1,000 customers reach a third order; 500 randomly receive a note and early access, 500 do not. 135 vs 110 repurchase (+22.7% relative, 25 orders, $800 margin against $700 cost). The gap gives z about 1.84, p about 0.066: promising, not proven.

What should you measure?

  • Repeat purchase in a fixed window, thanked versus holdout.
  • Time to next order. A faster return is an effect even when the rate is equal.
  • Support contacts and cancellations in the following weeks.
  • Referrals that follow.
  • Cost per customer, including the time of the person writing.
  • Opt-outs and complaints after the message. A thank-you that triggers unsubscribes is not appreciation, and you want to see that early.
  • Replies that say something specific. A customer who writes back about the detail you mentioned shows the message landed, though it is a weaker signal than repeat purchase.

Decide the window and the holdout before you send anything; looking afterwards for a window that works is the same error as stopping a test early.

How does this differ for business customers?

In B2B the person you thank is not always the person who pays. Thank the user who made the product work, the buyer who signed and the champion who argued for you, each for their own contribution. Keep gestures within what their employer allows, prefer things of use (access, an introduction, a result written up with permission) over objects, and tie thanks to outcomes the account has reached, such as a milestone in their own use.

What goes wrong with customer appreciation?

  • Sending the same message to everyone. It stops being appreciation.
  • Thanking late. The moment has passed.
  • Disguising a discount. If it asks for a purchase, call it an offer.
  • Judging by replies and compliments. Compare repeat behaviour with a holdout.
  • Choosing recipients by hand and comparing with the rest. You selected the best customers.
  • Declaring victory on a borderline result. The example above was not significant.
  • Overlooking gift rules. Tax limits and recipient policies apply.

How do you plan and check this in TapCub?

Take the third-order thank-you. TapCub does not send emails or gifts, and it does not split customers into test groups. It identifies who qualifies and measures what they do next:

  1. Call tapcub.login when customers sign in, so they appear in All-platform analytics › Users › People with a day-by-day timeline. Anonymous visitors have no profile and cannot be segmented.
  2. In All-platform analytics › Users › Segments & tags, build a rule segment for people who reached their third order, or use the RFM view with 10 groups, from champions to those at risk, to see who is slipping.
  3. Randomly assign the list to two groups outside TapCub, then paste each group's IDs into a static segment: one for thanked, one for holdout.
  4. In All-platform analytics › Behavior › Behavior analysis, bring each segment into a retention or funnel analysis for the repurchase event and compare the 60-day return of the two groups.
  5. In Live chat › Canned replies, keep a short, specific thank-you that agents can insert with / and edit before sending. Chat sees the visitor's history, so the message can name what they did.
  6. Costs, margins and the thank-you messages themselves live outside TapCub.

FAQ

What is customer appreciation?

It is recognising something a specific customer did and thanking them in a way that shows you noticed, such as a repeat order, a referral, a renewal or a patient wait. It is different from a promotion because it does not ask for a purchase.

How do you show customer appreciation?

Be specific, timely and proportionate: a personal note that names what they did, early access to something they asked about, a short call from someone who knows the account. Choose triggers from real events in your data and send close to the event.

Does customer appreciation increase repeat purchases?

It can, but you only know by comparing thanked customers with a random holdout. In the sample, 27.0% versus 22.0% looked promising but gave p of about 0.066, so it was not proven; larger groups would be needed.

How much should you spend on customer appreciation?

Spend in proportion to the customer's value and the relationship, and compare the cost with the extra margin from repeat purchases. In the sample, $700 of notes returned $800 of margin. For business gifts, the US deduction limit is $25 per recipient per year.

What are good customer appreciation ideas for B2B customers?

Personal notes to the user, buyer and champion for their own contributions, early access or an introduction, a result written up with permission and recognition at milestones in their usage. Keep within what their employer permits and avoid gifts that look like inducements.

Sources

Was this article helpful?

We use your answers to fix and expand this page.

Sophia Bennett

Product Adoption Consultant

Explores customer onboarding, live conversations, and self-service, turning common usage questions into practical guidance and improvements.

Find who to thank, and check what changed

Identified customers, RFM segments, static segments for a holdout and retention by segment.

See your data clearly. Find your growth.

Every click, backed by data. Install one line of code and see your first numbers in a minute.

No credit card · Free plans for analytics and chat · Cookieless analytics