D2CX COHORT 9 · LIVE LAB · 4 HOURS
THE SECOND ORDER LAB
A 4-HOUR INVESTIGATION INTO ONE MISSING CUSTOMER.
BEFORE WE START: 1. Keep chat open, you'll type more than I talk. 2. Duplicate your workbook 3. Have your best-performing ad open in a tab.
01
FOLLOW UP FOR REPEAT
THIS PROTEIN TUB REACHED HER 21 DAYS AGO. IT IS STILL ALMOST FULL. Yesterday, the brand sent her:
“RUNNING OUT? REORDER TODAY — 20% OFF.”
CHAT: What might actually be happening in her life right now from the pre
A protein tub · 1 kg · ₹2,199 · first order · delivered in 4 days
second order lens? One line. 02
GROUND RULE
THESE WORDS ARE ON LEAVE FOR 4 HOURS.
RETENTION
CRM
LIFECYCLE
ENGAGEMENT
If you catch yourself using one, translate it into what the CUSTOMER is doing. “We'll improve engagement” → “she should open the tub twice a day and notice something by week 3.”
03
CASE FILE 001
FICTIONAL
PRIYA, 29, GURGAON. THE BRAND'S DASHBOARD SAYS:
SAW: Instagram ad: “glass skin in 14 days.” BOUGHT: ₹899 vitamin-C serum · first order · COD.
CAC ₹412 ROAS 2.2
DELIVERY: 6 days. USED IT: we don't know.
“CAMPAIGN SUCCESSFUL.”
ORDERED AGAIN:
never.
Every one of you owns a few hundred Priyas.
04
POLL 01 · COMMIT TO A VERDICT
WHY DID SHE NEVER RETURN? A — She forgot the brand
B — The serum didn't work
C — It worked, she didn't notice
VOTE. D — No discount, no deal
E — She never actually used it
F — Rival / marketplace took her
G — Wrong customer from day one
05
ACTIVITY · 12 MIN
UNLIMITED WHATSAPP CREDITS. BRING ONE OF THEM BACK. PRIYA Promised glass skin in 14 days. Real results need 90.
ROHAN Bought protein at 50% off. Has never paid full price for anything.
MEERA
ARJUN
Loved your kids' snack. Blinkit now has a lookalike brand,
Bought a kurta on Diwali. Liked it. Simply forgot your
₹40 less, 10 min away.
brand's name.
PICK ONE. TYPE THE EXACT MESSAGE. 3 MINUTES. 06
THE REVEAL
A MESSAGE IS A COURIER. CAN FIX
CANNOT FIX
PRODUCT
MEMORY
PROMISE
FRICTION
PERSON TIMING
One of four customers was saveable by message. Most retention budgets are spent messaging the other three. 07
FIRST ORDERS ARE BOUGHT. SECOND ORDERS ARE BUILT. WORKBOOK P1: REWRITE THIS IN YOUR OWN WORDS.
08
CHAT STORM · YOUR OWN MONEY
A BRAND YOU BOUGHT ONCE AND NEVER AGAIN. FORMAT: BRAND — ONE WORD WHY. mama earth Beauty by bie Kwality walls bodywise RAS oils lip Project shades Nat Habit roadster shoes mama earthDove roll ons Savana good bug Earth Rhythm Nutrabox D'You Juicy Chemistry Cetaphil
PRODUCT
PERSON Campa cola Minimalist sunscreen Amul protein milkshake
TIMING
MEMORY
ACCESS
Amul protein lassi GoodBug good bug
Your answers will be sorted live into these five columns. Then YOU name the columns. 09
FRAMEWORK 01 · YOU BUILT THIS
THE FIVE BREAKS. 1
2
3
4
5
PRODUCT
PERSON
TIMING
MEMORY
ACCESS
Wrong customer
Need hasn't returned
acquired
— or moment passed
It didn't earn a return
Liked it. Forgot it.
Returning was harder than switching
LIGHTNING POLL: gym in January · Diwali kurta in August · crushed COD box · “that peanut butter… what was it called?” · out of stock on reorder day WB P3: RATE YOUR BRAND 1–5 ON EACH. CIRCLE THE WEAKEST.
10
CASE FILE 001 · NEW EVIDENCE
FICTIONAL
LISTEN LIKE A DETECTIVE, NOT A MARKETER. The ad said 14 DAYS. The leaflet inside — which nobody reads — says 8–12 WEEKS, twice daily.
Delivery: 6 days. The box arrived WARM. Gurgaon, June.
She used it 5–6 times over two weeks. Day 13: mirror, nothing.
Day 20: the bottle slides behind the moisturiser.
Day 45: the brand speaks for the first time — “20% OFF! STOCK RUNNING OUT!”
11
FRAMEWORK 02 · BREAKOUT · PAIRS · 8 MIN
FIND THE FIRST BROKEN LINK.
PROMISE 1
→
EXPERIENCE 2
→
VALUE FELT 3
→
MEMORY 4
→
EASY RETURN 5
For each link: INTACT or BROKEN — with evidence from the case. Mark the FIRST break. Everything downstream inherits the failure.
WB P4: PRIYA'S CHAIN, THEN YOUR OWN.
12
OWN-BRAND AUDIT · PAIRS · 10 MIN
LIVE · COHORT BRAND
OPEN YOUR OWN AD. LET'S SEE. A — THE PROMISE
B — THE REALITY
C — THE GAP
What your ad or product page makes a
What actually happens, order to Day 7.
Where expectation exceeds experience.
STRANGER expect. One honest
Delivery, unboxing, first use, first visible
Your partner reads your ad as a stranger
sentence — what it implies
result.
and tells you what THEY expect.
13
POLL 03 · GUESS BEFORE REVEAL
LIVE · COHORT BRAND
WHICH ENTRY PRODUCT BUILT THE BEST RETURNERS? A
B
C
[HERO / GIFT SKU]
[TRIAL / MINI SKU]
[BORING CORE SKU]
The “smart funnel” product
Full-size. Nobody's favourite ad.
The bestseller — largest share of new customers
VOTE, THEN DEFEND IT IN CHAT — ONE LINE. VOTES LOCK IN 90 SECONDS.
The founder is in this room. They will reveal the real numbers. · Fallback if founder absent: anonymised case (next slide).
14
TSO NEWSLETTER >
SUBSCRIBE
ONE RETENTION STORY. EVERY WEEK.
THE SECOND ORDER WEEKLY NEWSLETTER
BREAK 10 MINUTES
PUZZLE WHILE YOU'RE GONE: what share of Indian D2C repeat orders ride on a discount? Answer when you're back.
17
FRAMEWORK 03 · CALCULATE YOURS
VERIFIED
HER CLOCK VS YOUR CLOCK. CLOCK 1 — HER NEED
CLOCK 2 — YOUR MESSAGES
Pack size ÷ realistic daily use
When do you FIRST ask for a reorder?
= days until the need returns.
GAP = Clock 2 − Clock 1.
(Or: the occasion, the season, the growth stage.)
Early = interruption. Late = forgotten.
COUNTRY DELIGHT: a business physically built on milk's 24-hour clock — daily subscription, edit till 11:59 PM, at the door
by 7:30 AM. Cross-sell starts only after the habit exists.
TYPE YOUR GAP IN CHAT AS A NUMBER OF DAYS. · NO NATURAL CLOCK? WRITE THE OCCASION.
16
FRAMEWORK 04 · POLL 04 · MATCHER
FIVE WAYS A CUSTOMER COMES BACK. RUNS OUT
STEPS UP
COMES AROUND
SPREADS
BECOMES A HABIT
replenishment
next product, next need
occasion, season
another household member
ritual, identity
MATCH: milk · vitamin-C serum · festive kurta · kids' snack box · whey protein · scented-candle gift · dog food · everyday perfume
THEN IN CHAT: YOUR BRAND'S WAY BACK, TWO WORDS. · WB P7
18
CALCULATION · POLL 05
YOUR DASHBOARD IS LYING TO YOU. TWICE. THE LIE “Repeat rate: 14%” — counted across ALL customers,
DON'T JUDGE A BATSMAN ON TWO BALLS.
including one who bought yesterday.
THE HONEST QUESTION
A 20-day-old customer on a 90-day product isn't LOST.
Of 100 customers whose clock has fully passed (≥ 1.5 ×
She's BUILDING.
your product clock), how many bought again?
COMPUTE BOTH NUMBERS ON THE SAMPLE TABLE. POST THEM. · WB P8: YOUR FORMULA, YOUR BOUND.
19
RANK VOTE + ₹ MATH
THREE REPEAT CUSTOMERS WALK IN. RANK THEM. A
B
C
THE REFILLER
THE CODE HUNTER
THE INVISIBLE LOYALIST
Full price, on her own, month after
Repeats ONLY when the 40% code
month.
arrives.
Loves you — buys you on Amazon / quick commerce, where you never see her.
THEN THE MATH: ₹1,000 AOV · 55% gross margin · what does B's repeat earn after 40% off + shipping? Type the number.
20
FINAL CHALLENGE · BRIEFING
PLAUSIBLE IS NOT PROVEN. 1. OBSERVED — only what the data shows
5. FIRST TEST — 14 days, cheap, reversible
2. HYPOTHESES — max 3, ranked
6. DO NOT DO YET — tempting but premature
3. MOST LIKELY BREAK + chain link
7. OWNER — which function
4. EVIDENCE STILL NEEDED — gettable in a week SCORING: evidence + · “we can't know yet, here's how we'd find out” + · confident guessing −
21
20:00 ROOMS ARE LIVE. BOX 1 IS SACRED: OBSERVED ONLY.
22
VERDICTS · 90 SECONDS PER ROOM
DEFEND YOUR DIAGNOSIS. 90 SEC
One spokesperson. Sheet on screen. No preamble.
CHALLENGE
Other rooms attack via chat — evidence only.
SCORE
Evidence earns. “We don't know yet” earns. Swagger costs.
HOT SEAT AFTER VERDICTS: [COHORT BRAND] gets the room's 7-box treatment — founder answers with facts only.
23
SOLO · QUIET 15 · WB P10–11
THE ONLY CASE THAT MATTERS: YOURS. 1. Why should she return? One sentence. Banned words: quality, delight, community, discount — unless made specific.
2. My most likely FIRST break — which of five, which link, what evidence.
3. My 14-day experiment against that break — small, cheap, reversible. With a stop rule.
4. The ONE metric that says it worked. · 5. One question for 5 real customers this week.
6. One thing I will STOP doing — usually a flow you're running right now.
7. Owner + a 30-minute weekly review slot. Day, time, name.
24
PAIRS · 5 MIN · WB P12
SAY IT OUT LOUD. NOW IT'S REAL. READ
Your return-reason sentence + your 14-day experiment. Out loud.
ONE HARD QUESTION
Partner asks exactly one. Example: “How will you KNOW in 14 days?”
DATE
Exchange numbers. Day-14 check-in goes in both calendars. Now.
THEN: POST YOUR ONE EXPERIMENT IN CHAT. ALL OF YOU. AT ONCE.
25
9 BRANDS . AN ACCELERATOR PROGRAM.
SAME METRIC. DIFFERENT REPEAT. Category
SPP-60 O1→O2 O2→O3 O3→O4
Med. days
90d+ tail
Repeat on code
Fresh staples
37.8% 45.1% 69.7% 81.4%
1d
—
—
Protein F&B
22.9% 28.2% 46.6% 56.5%
27d
10%
84.8%
16.2% 27.0% 10.0%
38d
24%
—
Athleisure
~15.6% 14.1% 25.8% 20.5%
19d
12%
86.8%
Kurtas
9.8%
11.6% 18.5% 34.5%
11d
13%
63.0%
Swimwear
7.1%
11.0% 16.8% 26.5%
16d
30%
32.6%
Jewellery
~10.1% 9.4%
43%
42%
0–30d
~12%
~all
Socks
5.7%
16.2% 24.0%
21d
32%
43.6%
Auto accesories
—
9.0%
25
CASE FILE 001 · RESOLUTION
CASE CLOSED
THE FIRST LINK SNAPPED BEFORE SHE PAID.
The ad promised 14 days for a 90-day product. The warm box, the silence, the Day-45 discount — they only confirmed a decision that was already forming.
THE BRAND DIDN'T LOSE HER AT RETENTION. IT LOST HER AT ACQUISITION — AND BILLED THE LOSS TO THE CRM BUDGET.
First orders are bought. Second orders are built. 26
WHAT WILL YOU TELL YOUR TEAM ON MONDAY? ONE SENTENCE. TYPE IT NOW. Not “it was good.” The sentence that changes what they do this week.
THE SECOND ORDER · D2CX COHORT 9
27
THE SECOND ORDER
14-DAY FIELD TEST
DON'T HIRE ME YET. GO RUN THE EXPERIMENT YOU JUST DESIGNED.
14 DAYS. AFTER 14 DAYS
01
TEST
02
WATCH
Track the one number that would prove or disprove it.
03
TALK
Speak to 5 customers. Don't sell. Find evidence.
04
COME BACK
Only if the evidence still doesn't explain the second purchase.
Run the one experiment you committed to today.
IF YOU STILL CAN'T ANSWER:
“Why aren't enough of my first-time customers becoming second-time customers?” You probably don't need another campaign.
YOU NEED A DIAGNOSIS.
THE SECOND ORDER.
SCAN THE QR / LEAVE YOUR DETAILS.
We investigate the evidence with you.
We start with the evidence.
EVIDENCE BEFORE INTERVENTION.
INVESTIGATE →
NEXT