Gains:
- Kampanya mekaniğini (indirim, bundle, kupon, eşik) hedef ve bütçeyle uyumlu biçimde yapay zeka desteğiyle tasarlayabilme
- Ability to set up pre-campaign forecasting and post-campaign measurement (uplift, ROI, cannibalization) in draft form
- Yapay zekanın önerdiği kampanya iddialarının mevzuat ve dürüst reklam ilkeleriyle doğrulanması gerektiğini kavrayabilme
Kampanya, perakendenin en görünür silahıdır: doğru kurulduğunda ciroyu, trafiği ve müşteri kazanımını yükseltir; yanlış kurulduğunda marjı yer, zaten alacak müşteriye para dağıtır ve hiçbir kalıcı fayda bırakmaz. In this unit, you will use artificial intelligence as an assistant that designs campaign mechanics, predicts its impact, and measures its aftermath. Ama kampanyanın hedefini, bütçesini ve iddialarının doğruluğunu siz belirlersiniz.
Basic concepts:
Campaign mechanics: How the offer works: percentage discount, "buy 2 pay 1", threshold discount (20% over 150 TL), bundle, coupon, points.
Uplift (artış): Kampanyanın, kampanya olmasaydı gerçekleşecek satışın üzerine eklediği ek satıştır. This is what you really want to measure; toplam satış değil, kampanyanın yarattığı fark.
Cannibalization: The sale of the campaign product occurs by reducing the sales of other products of the same business. It must be deducted to see the net gain.
ROI (Yatırım Getirisi): Kampanyaya harcanan her lira karşılığında elde edilen ek kâr.
First the target, then the mechanic
Every campaign should answer one question: What do you want?
target
Proper mechanics
Attention
increase traffic
Remarkable main product discount (loss leader)
Margin products should be sold alongside
enlarge basket
Threshold discount, bundle
Threshold slightly above average basket
melt stock
Markdown, "last chance"
Establishing misleading urgency
new customer
First shopping coupon
New customer only
loyalty
Points are specific to the member
Shouldn't be public
Hint: "Let's have a discount" is not a goal. First, "traffic, cart, or inventory?" ask; mechanics comes after that.
Budget and breakeven
A campaign is an investment; Consider up front how much margin you expect to sacrifice and what you expect to gain. The AI calculates you the break-even point: “With this discount, we reduce the margin by this much; sales must increase by this much for the profit to break even.” If Uplift does not exceed this threshold, the campaign loses money.
Measurement: what happened when the campaign ended?
The success of the campaign is measured by three questions:
- How much was the uplift? Compare the campaign period to the similar normal period.
- Was there cannibalization? Did the campaign product outsell its full-priced counterpart?
- Is the ROI positive? Is the margin sacrificed less than the additional profit gained?
AI quickly establishes these comparisons from the data, but you verify the comparison period and assumptions.
Legislation and honest advertising
Campaign communications are subject to the law:
- Fake reference price prohibited: To say "699 instead of 999" the product must actually be sold at 999.
- Stock adequacy: There must be reasonable stock of the product announced as a "campaign"; Otherwise it's misleading.
- Clarity of conditions: Conditions such as "on selected products", "limited in stocks" should be stated in a legible manner.
- Fake urgency: Never-ending "last 2 hour" timers are misleading.
Beware: AI may exaggerate marketing copy or add unsubstantiated claims to make it attractive. Verify each campaign claim with actual condition and stock.
Campaign design step by step
- Clarify the goal: Traffic / basket / stock / earnings.
- Set up the mechanics: Targeted offer, threshold, duration.
- Estimate impact: Expected uplift, margin impact, breakeven.
- Produce content: Draft message, title, channel text.
- Filter + measure: Verify claims, measure post-campaign uplift/ROI/cannibalization.
mini cases
Case 1 — Basket enlargement with threshold discount: The average basket of an online store is 210 TL. Artificial intelligence suggests "free shipping over 250 TL" and shows that this can push the cart over the threshold. After the campaign, the average basket increases to 244 TL, and the increase in shipping costs is more than covered by the additional turnover; ROI is positive.
Case 2 — Cannibalization is caught: A market offers a 30% discount on brand A detergent; Sales triple, the team is happy. Artificial intelligence analysis reveals that sales of full-priced B and C brand detergents have fallen by 40%. Net category profit actually decreased. In the next campaign, the discount is distributed to the entire category and cannibalization is reduced.
Case 3 — Avoiding misleading claims: A textile retailer wants an "entire store up to 70% off" announcement, but only 5 items are 70% off. Artificial intelligence marks this in the ethical filter as "the condition is unclear, misleading". The text clarifies as "up to 70% off on selected products, 20-40% on others"; The risk of consumer complaints is prevented.
Campaign calendar and fatigue
Campaigns don't live in a vacuum; They sit on a campaign calendar. The store that constantly discounts trains the customer to "expect the sale": he stops buying at full price, waiting for the next offer. This is called promotion addiction and will permanently drag down margin. Bu yüzden kampanya sıklığını ve derinliğini bilinçli planlayın; Announcing a "big sale" every week makes the discount meaningless.
Benzer bir tuzak kampanya yorgunluğudur (promotion fatigue): aynı müşteriye çok sık, çok benzer teklif gitmesi. İlk mesaj ilgi çeker, beşincisi göz ardı edilir, onuncusu rahatsız eder ve abonelikten çıkışa yol açar. Give your campaign calendar to artificial intelligence and ask "Which segments are we touching and how many times this month? Is there a risk of fatigue?" you may ask. This way you keep the communication load balanced.
İpucu: Kampanyayı sadece "ne kadar sattırdı" ile değil, "markaya ne yaptı" ile de değerlendirin. Kısa vadede ciro getiren ağır indirim, uzun vadede fiyat algısını ve marjı aşındırıyorsa net kazanç olmayabilir.
Weak prompt / Strong prompt
Weak prompt:
Give me a campaign idea.
No target, budget, product, restriction; cliché returns.
Powerful prompt:
Your role: campaigns and promotions specialist.Context: My goal is to grow the average basket. Average basket is 180 TL, my gross margin is ~38%. There is overstock in the winter category. Task: (1) Suggest 2 campaign mechanics that fit the target (threshold/bundle), (2) expected basket effect for each and uplift required for breakeven, (3) draft a short announcement text. Rule: Do not use fake reference price, vague condition or fake urgency. Write the conditions clearly. Show account.
Copiable prompt templates
1) Mechanical recommendation according to the target
My goal is [traffic/basket/stock/earnings]. My margin is [X%], average basket is [Y] TL. Suggest 2-3 campaign mechanics suitable for this goal and write down the advantages/risks of each. Give it in a table.
2) Campaign breakeven estimate
I apply [mechanics]: [detail]. Calculate the margin impact and give the sales uplift required to break even. Explain what happens if the uplift falls below this threshold. Show account.
3) Post-campaign measurement
Below are the normal period sales during and before the campaign period. (1) calculate the uplift, (2) is there cannibalization in other products, (3) is the ROI positive according to the sacrificed margin? Show step by step.Data: [paste]
4) Campaign text ethics audit
Aşağıdaki kampanya duyurusunu denetle: sahte referans fiyat,belirsiz koşul, stok yanıltması, sahte aciliyet veya kanıtsıziddia var mı? Riskleri işaretle, dürüst alternatif öner.Metin: [yapıştır]
Common mistakes
- Targetless campaign: Saying "Let's make a discount"; The mechanic must exit the target.
- Sadece toplam satışa bakmak: Uplift ve kanibalizasyon ölçülmezse başarı yanılsaması olur.
- Başabaşı hesaplamamak: Feda edilen marjı kapatacak satış artışını bilmemek.
- Sahte referans/aciliyet: Yanıltıcı ve yasa dışıdır.
- Herkese açık sadakat teklifi: Üyeye özel avantajı herkese vermek anlamı boşaltır.
- İddiaları doğrulamamak: Yapay zekanın cazip metnini kontrolsüz yayınlamak.
In summary
A good campaign sets the goal first, the mechanics second; limits the budget to breakeven and measures the result by uplift, cannibalization and ROI. AI speeds up mechanical recommendation, impact estimation, text drafting, and measurement calculation. You are responsible for the accuracy of the claims and regulatory compliance.
Application task
Clarify the goal for an upcoming need (basket expansion or stock thinning). Remove 2 mechanics with "1) Mechanic recommendation by target", select one and run "2) Campaign breakeven prediction". Produce the announcement text and pass it through "4) Campaign text ethical audit".
checklist
- [ ] I set the campaign goal clearly.
- [ ] I chose the mechanic based on the goal.
- [ ] I calculated the lift required to break even.
- [ ] I evaluated the risk of cannibalization.
- [ ] I wrote the claims and conditions honestly/clearly.
- [ ] My post-campaign uplift and ROI measurement plan is ready.