Episode 25: Budget, bidding aur A/B testing
Daily vs lifetime budget, bid strategies, learning phase, aur A/B testing ka sahi tareeka.
Neha poochti hai: “Ad A achha hai ya Ad B?” Uski dost kehti hai A, uska bhai kehta hai B.
Marketing mein raay nahi, test jeetta hai. Customers ko vote karne do, unke clicks aur orders se.
Daily vs lifetime budget
| Type | Kaise | Kab |
|---|---|---|
| Daily | Roz average ₹X kharch | Hamesha chalne wale campaigns |
| Lifetime | Poore period ka total, Meta baantega | Fixed sale, launch (start–end date) |
Campaign budget (Advantage campaign budget) mein Meta khud ad sets ke beech paisa baant-ta hai. Ad set budget mein aap control karte ho. Chhote account mein dono chal sakte hain; test ke liye ad set budget zyada saaf hai.
Bid strategies, simple bhasha mein
- Highest volume (default): budget mein jitne zyada results. Shuru yahin se.
- Cost per result goal: “ek order ~₹250 ke aas-paas chahiye”. Thoda data aane ke baad.
- Bid cap: auction mein max bid. Advanced, chhote budget ke liye risky.
- ROAS goal: revenue ÷ spend ka target. Website + purchase tracking ke saath.
StyleStreet: pehle highest volume, 2–3 hafte baad cost goal try.
Learning phase: ad ko seekhne do
Naya ad set ya bada change ke baad Meta “learning” mein jaata hai. Is waqt results upar-neeche hote hain.
- Meta ki guidance: ek ad set ko ek hafte mein lagbhag 50 optimization events chahiye stable hone ke liye
- Learning mein baar-baar edit mat karo. Har bada edit learning restart karta hai
- Kam budget mein 50 purchases mushkil hain, isliye kabhi-kabhi “messages” jaise upar wale event par optimize karte hain
A/B testing: ek baar mein ek cheez
| Test karo | Variable A | Variable B |
|---|---|---|
| Hook | “10 baar dhoya…” | “POV: 8:45 AM…” |
| Offer | COD + exchange | Free scrunchie |
| Audience | Interests | Broad |
| Format | Reel | Carousel |
Meta Ads Manager mein built-in A/B test feature bhi hai jo audience ko barabar baant-ta hai.
Winner kab declare karein?
2 din, 200 rupaye ke baad winner chunna = sikka uchhaalna.
- Kam se kam 5–7 din (weekday + weekend dono)
- Har version par itna spend ki kuch results aayein (jaise 15–20+ messages/orders har side)
- Fark bada aur lagataar ho, ek din ka nahi
Agar dono barabar rahein?
Toh woh variable shayad itna matter nahi karta. Jo sasta/aasaan hai woh rakho, aur agla bada variable test karo (jaise offer ya hook).
StyleStreet ki real problem
Har din badlav, kabhi stable nahi
Neha roz subah ads dekhti: kal ₹400 per order, aaj ₹600? Turant audience badal deti, budget aadha, naya creative.
Har edit se learning reset. 3 hafte tak ad set kabhi stable nahi hua, aur Neha ko laga “ads kaam nahi karte”.
Practical example: Hook test: wash test vs POV
Same audience, same offer, same budget (₹150/day each), 7 din:
| Ad | Spend | CTR | WhatsApp chats | Orders | Cost/order |
|---|---|---|---|---|---|
| A: “POV 8:45 AM” | ₹1,050 | 0.9% | 14 | 3 | ₹350 |
| B: “10 baar dhoya” | ₹1,050 | 1.6% | 26 | 7 | ₹150 |
Winner: B. Agla test: B ke saath do alag offers.
Aapka mission: Test plan banao
StyleStreet ke liye agle 4 hafte:
- Budget type choose karo (daily/lifetime) aur reason likho
- Shuruaati bid strategy decide karo
- 4 hafte ke liye 4 tests ki list (ek baar mein ek variable)
- Har test ka hypothesis likho: “Mujhe lagta hai B behtar hoga kyunki…”
- Winner declare karne ke rules likho (din, spend, results)
- Test results ka sheet format banao
Quiz
-
A/B test ka sabse important rule?
- A. Zyada se zyada cheezein ek saath badlo
- B. Ek baar mein sirf ek variable badlo
- C. Sirf weekend par test
- D. Budget alag rakho
Answer dekho
Sahi jawab: B. Ek baar mein sirf ek variable badlo
Tabhi pata chalega ki fark kis cheez se aaya.
-
Learning phase mein baar-baar edit karne se?
- A. Results jaldi aate hain
- B. Learning restart hoti hai aur results unstable rehte hain
- C. Kuch fark nahi
- D. Ad band ho jaata hai
Answer dekho
Sahi jawab: B. Learning restart hoti hai aur results unstable rehte hain
System ko seekhne ka time do.
-
Naye account ke liye shuruaati bid strategy?
- A. Bid cap
- B. Highest volume
- C. ROAS goal
- D. Manual CPM
Answer dekho
Sahi jawab: B. Highest volume
Pehle data aane do, phir cost goals lagao.
-
Ad B ka cost/order ₹150, Ad A ka ₹350 (7 din, barabar spend). Kya karein?
- A. Dono band
- B. B ko winner maan kar scale karo aur agla test
- C. A ko budget do
- D. Aur 1 saal wait
Answer dekho
Sahi jawab: B. B ko winner maan kar scale karo aur agla test
Saaf aur lagataar fark hai. Winner rakho, aur test aage badhao.
Client project mein output
Project file mein ab Budget & Testing Plan:
- Budget type + bid strategy
- 4-week test roadmap
- Winner rules
- Pehla result: “10 baar dhoya” hook winner (₹150/order)
Golden rule: Raay se nahi, test se decide karo. Ek baar mein ek cheez, kaafi data, aur learning ko saans lene do.