20171026 APT case interview workshop note

Case Interview Introduction: Ken
• B.S.機械 M.S.商研所
• case interview: A mile high view
◦ what & why
‣ analytical skills
‣ logic structure quant skills
◦ types
‣ mkt sizing
‣ strategy /operation case
• average interview structure
◦ getting to know u (10min)
‣ show your personality (humor is good)
‣ show some enthusiasm for the company
‣ explain why consulting n why this company
‣ don't be put off if the interview dives Right into the case and skip this section
◦ case (30min)
◦ Q n A (5min)
‣ ur chance to learn more about the company (it should always b 2 way street)
‣ never say u have no question
‣ be sure to leverage this opportunity
• reality: what they r looking 4
◦ how u think
◦ quantitative skill
◦ composure professional demeanor
◦ communication skills
• debunk A few myth
◦ not expected to solve A complete case: process
◦ advanced calculating is not required
◦ no wrong ans, every ans has value
• few example
◦ is the credit card promotion profitable 花旗信用卡電影票打折
◦ should tesla change pricing strategy 特斯拉在台灣的定價偏高
◦ ANA is considering to add A new PE type seats. what issue does it need to consider 是否增加新的倉等
◦ should uber enter food delivery industry
• why is it so hard
◦ industry background
◦ short time
◦ identify the key question
• Framework
◦ at the minimum know the basic
‣ profit=total revenue-total cost
‣ revenue=price*quantity
‣ cost=fixed cost+quantity*variable cost
◦ even better
‣ be able to draw supply n demand curve
‣ A through review of an income balance sheet
◦ tread carefully
‣ u want to use A framework that applies to the question you are being asked and that you r familiar with
• A few tips
◦ be honest
◦ listen, clarify the question. pause n think
◦ identify A Framework or structure
◦ ask question: interviewer may not know what kind of info u need
◦ think out loud:write down your finding n assumption. don't miss any possibility. think of every possibility. there's no limit
◦ remember there's no Right answer

Live demo: Ken n Tony
• Case: gobuy 飲料店,是一個 premium brand,想要拓展學生客群,欲在校園店推出學生證折扣5塊的促銷活動
• 重複問題,提出問題:是否願意承擔虧錢的風險
• 目標:提升mkt share 不要虧錢
• strategy: 比較pre pos的差距
• pre
◦ mkt share:
‣ 台北地區的有50間店有10間店為校園店
‣ Revenue(以月為單位)
• 一例一休 營業25天
• 一天約500杯
• 熱銷品(50%.$50.2種)/其他飲料(50%.$60.約20種)%
‣ Cost
• fixed
• variable
◦ profit rate: 一般飲料90%,但這個是premium brand 所以是熱銷品60%一般品70%
• pos
◦ 兩類產品都有折扣嗎?
‣ 只做熱銷品折價
◦ 為什麼有這種構想?之前有做過活動嗎?如果有的話,有相關銷售的成長數字嗎?
‣ 上班族的campaign 熱銷品成長20%
◦ 競品效益
‣ 其他類下降5%
◦ hypothesis 學生對價錢會更敏感,熱銷品+40%其他類-10%
◦ 學生:社會人士的比率
‣ 7:3(社會人士會持平)
◦ 學生購買熱銷品:一般品的比率
‣ 6:4
• margin 以天為單位
◦ pre
‣ 學生
• 熱銷
◦ 500*0.7*0.6*(50*60%)=6.3k
• 一般
◦ 500*0.7*0.4*(60*70%)=5.6k
◦ pos
‣ 學生
• 熱銷
◦ 500*0.7*0.6*1.4*(50*60%-5)=7.35
• 一般
◦ 500*0.7*0.4*0.9*(60*70%)=5.04
◦ Sum 一天增加0.4k 一個月增加0.4*25=10k
• 總結
◦ mkt share↑
◦ 一個月 10k ↑
◦ 幾點注意
‣ premium brand的形象是否會因為折價而有影響
‣ 注意assumption是否符合預期: 學生因為價格下降導致來客數的兩倍成長,若assumption不符合預期,那麽margin也會相應的變化
‣ 注意競食作用,campus store的折價活動會不會搶奪其他premium store的顧客
◦ Thus, 可以實施campaign, 但要隨時注意觀察campaign的情況
• comment
◦ 邊問邊展現business sense
◦ 把握好架構,省去計算fixed Cost
◦ think out loud 講出每個數字的含義
◦ 可以回去算保守的assumption,即如果學生的來客數成長和上班族的來客數成長相同,那麽margin會怎麼變化
prior to each interview
• research the company
• current employee
• study ur resume
• don't panic

Q&A
• Case會不會練太多
◦ Framework只是輔助,現實情況更複雜,要更靈活更flexible
• 如果interviewer否定你的數字或是assumption怎麽辦?
◦ 主要還是要看你怎麼解決問題,所以要問對的問題
◦ think out loud,邊解邊確認你的數字是不是和客戶的實際狀況相同
• 台北辦公室的規模
◦ 24-25人,東京:9人
• 有沒有用比較新的統計方式、machine learning
◦ 長期合作
◦ 有machine learning,建模型
◦ test n learn: 模型是根據過去的數據去預測未來,但是更重要的是要test n learn ,不斷更新model
• high level 的假設
◦ 要給,展現你思考的全面性以及structure
◦ 但是也要展現效率,告知interviewer先看針對某一塊dive in
• 台北辦公室如何拓展客群
◦ 有業務代表
• mckinsey vs apt
◦ mckinsey適合從0到1,hypothesis driven
◦ APT適合日常營運相關,需要數據作證
• 會不會特定鑽研某個產業?怎麼培養產業敏感度?
◦ 用客戶的資料去做出決策
◦ 長期合作
◦ 新聞
◦ 多看case
Like what you read? Give Peilin Huang a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.