Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
データサイエンスを活用した契約資産価値評価(LTV)とその改善 / Lifetime Valu...
Search
SAWA, Norihiko
February 25, 2022
Business
160
1
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
データサイエンスを活用した契約資産価値評価(LTV)とその改善 / Lifetime Value as Asset Pricing with Data Science
SAWA, Norihiko
February 25, 2022
More Decks by SAWA, Norihiko
See All by SAWA, Norihiko
OCR Optimized for Images Created by Print Typesetting
nakamods320yen
0
160
日本経済新聞社 デジタル人材採用案内 / Nikkei Digital Hiring
nakamods320yen
1
49k
AI making new content
nakamods320yen
0
8.4k
Firebase and React Native
nakamods320yen
2
770
Dive into E2D3
nakamods320yen
0
510
Other Decks in Business
See All in Business
事業部でのAI推進とAIネイティブレベルの話
fotoyuma
0
490
スマサテでの日々 -Sumasate Tour Deck-
sumasate
0
3.1k
20年続く長寿タイトルが、15年目にして売上を伸ばせた理由
gree_tech
PRO
0
810
株式会社ジグザグ_新規投資家向け資料_2026年7月
zig_zag
0
2.4k
PM Kansai#5 Sponsor Session
y0suke4mano
0
120
AI時代のプロダクトディスカバリー:名刺OCR機能β版開発の舞台裏 2026.02.17
kawashimayoshihisa
0
120
負債解消という仕事は儲かる
uproad3
6
4k
研修ガイドブック
nicrecruit
1
240
エンジニアの示唆に溢れている漫画『HUNTER×HUNTER』
shinyasaita
0
120
プレイド概要説明資料_2026/08
plaid
PRO
1
3.8k
Kellogg Magazine Summer 2026
tsogo817421
2
190
47グループ 会社紹介資料 / Company Introduction
47holdings
1
22k
Featured
See All Featured
The Organizational Zoo: Understanding Human Behavior Agility Through Metaphoric Constructive Conversations (based on the works of Arthur Shelley, Ph.D)
kimpetersen
PRO
0
430
Mobile First: as difficult as doing things right
swwweet
225
10k
Believing is Seeing
oripsolob
1
200
Sharpening the Axe: The Primacy of Toolmaking
bcantrill
46
3k
Data-driven link building: lessons from a $708K investment (BrightonSEO talk)
szymonslowik
1
1.3k
The World Runs on Bad Software
bkeepers
PRO
72
12k
ReactJS: Keep Simple. Everything can be a component!
pedronauck
666
130k
Primal Persuasion: How to Engage the Brain for Learning That Lasts
tmiket
0
430
Design of three-dimensional binary manipulators for pick-and-place task avoiding obstacles (IECON2024)
konakalab
0
550
GraphQLの誤解/rethinking-graphql
sonatard
75
12k
HU Berlin: Industrial-Strength Natural Language Processing with spaCy and Prodigy
inesmontani
PRO
0
670
New Earth Scene 8
popppiees
3
2.5k
Transcript
ຊܦࡁ৽ฉࣾΠϊϕʔγϣϯɾϥϘᖒل σʔλαΠΤϯεΛ׆༻ͨ͠ ܖࢿ࢈ՁධՁʢ-57ʣͱͦͷվળ %BUBCSJDLTXFCJOBS"*׆༻ʹΑΔϏδωε՝ղܾγϦʔζ$-7ɾ$IVSOੳ 1
ϓϩμΫτϚωʔδϟʔɺݚڀһ ᖒل 2 ݄ץ w ܦࡁχϡʔεͷ༗ྉߪಡαʔϏε w ສਓͷ༗ྉߪಡऀ w ֹ݄
ԁʢεϚϗ՝ۚͰ ԁʣ w ץҎདྷɺܖ৳ͼଓ͚͍ͯΔ w ͞ΒͳΔ͕ٻΊΒΕΔ ܦిࢠ൛ w ຊܦࡁ৽ฉࣾೖࣾ w ։ൃԽɺσʔλδϟʔφϦζϜ w ػցֶशʹΑΔهࣄਪનɺݟग़͠ಈըͷࣗ ಈੜɺը૾ղੳ݁ՌʹΑΔΞϓϦͷ69 ্ͳͲ w σʔλαΠΤϯεΛͬͨࣄۀઓུͷཱҊ w ܖܗଶͷݟ͠ʹΑΔࣄۀՁ্
"HFOEB 3 w ࣗݾհ w $-7 -57 w ܖࢿ࢈ՁධՁ w
ࣄۀՁධՁ w ײԠੳʹΑΔɺࢪࡦ͝ͱͷܦࡁޮՌͷࢼࢉ w $IVSOੳ w -57ͷվળํ๏
ܖࢿ࢈ՁධՁʢ$-7ʣ 4
◼︎ ਓͷސ٬͔ΒಘΒΕΔকདྷ ΩϟογϡϑϩʔͷׂҾݱࡏՁ ◼︎ ӈਤ୯७ԽͷͨΊɺؒͷ ଘ$-7 $VTUPNFS-JGFUJNF 7BMVF ◼︎
ച্͚ͩͰͳ͘ɺ֫ಘඅ༻ͱҡ࣋ අ༻Λߟྀ ◼︎ ຖ݄ͷղՄೳੑʹΑΓ$BTI JOPVU fl PXచݮ͢Δ 5 ֹ݄ʹଘΛ͔͚ͨͷͷ૯͕ɺ ͦͷސ٬͔ΒকདྷಘΒΕΔΩϟογϡϑϩʔ ࢿ࢈ՁධՁ͕Մೳ αϒεΫܖ ۚ༥ࢿ࢈ͦͷͷ ࠂએඅΛ֫ಘඅ༻ɺγεςϜ։ൃඅΛҡ࣋අ༻ͱͯ͠ො՝ɻ ͰׂҾɻղʢΞϓϦ՝ۚʣΛͱͯ͠ܭࢉɻ ԁ
αϒεΫͷ ࣄۀՁධ Ձʢ-57ʣ ܖͷ$-7ΛͤࣄۀՁʢ-57ʣʹͳΔ <>$6450.&34"4 "44&54+063/"-0' */5&3"$5*7& ."3,&5*/(70-6.& /6.#&38*/5&3
6 w ্ه؆қࣜɻղܖ͔Βͷܦա݄σϞάϥʹґଘ͢Δ w ࣮ࡍʹίϗʔτ͝ͱʹ$-7Λܭࢉ͠͠߹Θ͍ͤͯΔ ࣄۀՁ -57 = ܖ "316 ܧଓظؒ × × ԁ ສਓ ϲ݄ ղ͕ຖ݄ҰఆͰͷ࣌ ݄࣌ ฏۉސ٬୯Ձ $-7k = T ∑ t=1 $BTI'MPXk,t (1 + i)t -57 = N ∑ k=1 $-7k ίϗʔτ͝ͱͷ$-7ɺ ݄ ʹίϗʔτ ͔ΒಘΒΕΔΩϟογϡϑϩʔɺ ׂҾ $-7k $BTI'MPXk,t t k i $-7
◼︎ কདྷͷऩӹΛߟྀͨ͠ࢦඪ ◼︎ αϒεΫࣄۀͰղ ʢ⁶ܧଓظؒʣ૿ݮͷকདྷ ऩӹͷӨڹ͕େ͖͍ ◼︎ ख़ظʹֻ͔ࠩͬͨ͠ࣄۀ Ͱɺղվળ͕ࣄۀ ͷ伴ΛѲΔ
◼︎ ސ٬୯ՁΛࢉೖ͢Δ͜ͱ Ͱɺ୯Ձ૿ͷҙ͚ࣝΛ ߦ͏ ◼︎ কདྷͷࣄۀͷ࢟ʹ͍ͭͯͷ ༧ଌਫ਼͕ߴ·Δ 7 ࣋ଓతͳʹɺղͷվળ͕͔ܽͤͳ͍ -57Λҙࣝ͢Δ ॏཁੑ ݄ ݄ ݄ ݄ ݄ ສ ສ ສ ສ ສ ສ ສ w্هΑΓɺݶք࣌ͷܖ ͱͳΔ w࣮ࡍʹղਃࠐ݄ʹΑͬͯҟͳΔ 4VCTmax = "DR w 4VCTt+1 − 4VCTt = "DR − w ⋅ 4VCTt ݄ ͷ༗ྉܖ ݄࣍ਃࠐʢҰఆͱ͢Δʣ ݄࣍ղʢҰఆͱ͢Δʣ 4VCTt t "DR w ղͱؒਃࠐ͝ͱͷܖͷਪҠ <>-JGFUJNF7BMVFܦిࢠ൛Ͱͷࣄ ۀඪͷཧ
ײԠੳ ɹ$IVSOੳΛ༻͍ͨࢪࡦධՁ 8
$IVSOੳͷछྨ 9 ղͷਪҠɺଐੑ͝ ͱͷղͷूܭ ࢪࡦͷɺղʹ༩͑ ΔӨڹͷௐࠪ ఆٛͨ͠είΞʹجͮ ͍ͨࢪࡦͷޮՌͷਪܭ ղͷѲͷͨΊͷूܭɻ ղࢭΛاਤ࣮͠ࢪͨ͠ࢪࡦ͕
࣮ࡍʹղࢭͨ͠ྔΛଌΔɻͦ ͷࢪࡦʹͨͬͨϢʔβʔͷɺ࣮ ࢪલޙͷσʔλΛ༻͍Δɻ ࣮ࡍͷղߦಈʹݱΕΔ·Ͱɺ ϲ݄ఔͭඞཁ͕͋Δɻ "#ςετείΞੳΛߦ͏ ͷ͕ཧ͕ͩɺίετ͕͔͔Δͷ ͕ɻ ղͱ૬ؔͷ͋ΔΤϯήʔδϝ ϯτείΞΛఆٛ͠ɺࢪࡦޮՌ ɺΤϯήʔδϝϯτείΞࣗମ ͦͷมͷ૿ݮΛݟͯஅ͢ Δɻ ܦిࢠ൛ͷείΞ๚ස Ӿཡهࣄຊ͕มɻ ࠓճ͓͢͠Δ͜ͱ
1SPEVDU5FDIͷࢪࡦɺࣄۀߩݙΛධՁ͠ʹ͍͘ എܠײԠੳͱࢪࡦධՁ 10 ࣄۀՁ (LTV) ܧଓ݄ F V ๚ස ফඅهࣄຊ
ඇࣗൃత ղࢭ Net ARPPU ηοτߪೖ ൢചखྉ ܦIDܾࡁൺ ిࢠ൛ϓϥϯൺ ՝ۚऀ ৽نདྷ๚ऀ ࠶དྷ๚ऀ ແྉମݧਃࠐ ՝ۚ ίϯςϯπϚʔέ ϓϩϞʔγϣϯ ϓϩμΫτ نͷ֦େ ސ٬֫ಘ ղࢭ ൺֱత༰қ ൺֱత༰қ ࠔ ࣄۀߩݙͷ ධՁ త νʔϜ ^ ^ ^ ܦిࢠ൛αϒεΫࣄۀʢݸਓʣͷ,1*πϦʔ ॳճ՝ۚ ηοτ୯Ձ MAU
◼︎ සൟʹདྷ๚͠ଟ͘ͷهࣄΛ ಡΉϢʔβʔͷղ ͍ ◼︎ ΤϯήʔδϝϯτείΞΛ ఆΊɺΤϯήʔδ֊ڃ͝ͱ ʹϢʔβʔΛྨ͠Ѳ ◼︎ -JHIUϢʔβʔʹΤϯήʔ
δ͍ͨͩ͘͜ͱ͕ඞཁ 11 είΞʹج͍ͮͨϢʔβʔྨΛߦ͍ɺ.JEEMFҎ্ͷϢʔβʔͷׂ ߹Λ,1*ͷҰͭʹઃఆɻ w ΤϯήʔδϝϯτείΞ Λఆٛ w ''SFRVFODZʢ๚සʣ w 77PMVNFʢফඅهࣄຊʣ w ͝ͱͷΛجʹϢʔβʔΛྨ F V log2 (F V + 1) ײԠੳͱࢪࡦධՁ $IVSOੳͱ Τϯήʔδϝϯτ είΞ 0 30,000 60,000 90,000 120,000 g0 g1 g2 g3 g4 g5 g6 g7 g8 g09 g10 g11 g12 g13 ֊ڃ͝ͱͷਓͱղ log2 (F V + 1) -JHIU .JEEMF -PZBM 4VQFS-PZBM <>ܦిࢠ൛Λʹಋ͍ͨσʔλ׆ ༻ज़ɻ%9Λ્ΉͭͷนͱΓӽ͑ํ
ํ๏ײԠੳͱࢪࡦධՁ w ΤϯήʔδϝϯτείΞ ͷଟՉ͕ղ͢͠ ͞Λද͢ w ֊ڃͷ͕Ұ༷ͩͱԾఆ͢Δͱɺ' 7͕૿ݮ ͨ࣌͠ʹ ͷ͕εϥΠυ͢ΔྔΛ
ܭࢉͰ͖Δ F V log2 (F V + 1) ࢪࡦͷ' 7վળྔΛݟࠐΉ͜ͱͰ $IVSOվળɺച্ߩݙΛਪఆ͢Δ 12 ༷ʑͳಛྔͷ͏ͪɺղͷӨ ڹ͕େ͖͍ͷ'ͱ7 w ֊ڃ͝ͱͷղɺ֓Ͷ୯ௐݮগͱ ͳΔ w ͕૿͑ΔͱɺҰఆ͕࣍ͷ֊ڃʹεϥΠυ͢Δ log2 (F V + 1) F V ࠜڌײԠੳ 10% ར༻ਓͷׂ߹ ๚ස × 1.2 ফඅهࣄຊ × 1.1 0 30,000 60,000 90,000 120,000 g0 g1 g2 g3 g4 g5 g6 g7 g8 g09 g10 g11 g12 g13 ֊ڃ͝ͱͷਓʢݱࡏͱ26.9%վળʣͱղ log2 (F V + 1) 1ؒͷച্૿ 0.15ԯԁ
-57վળҊͷྫࣔ 13
◼︎ ػೳར༻͝ͱʹɺސ٬ͷΤϯ ήʔδϝϯτ͕ҟͳΔ ◼︎ ͞ΒʹΤϯήʔδ͞ΕΔΑ͏ ʹɺػೳΛਪનදࣔ͢Δ Έ ◼︎ ᶃػೳ͝ͱͷΤϯήʔδଟՉ ᶄϢʔβʔ͕ػೳΛ͍࢝Ί
Δ֬ͦΕͧΕΛਪఆ͢ΔϞ σϧΛ࡞ ◼︎ %BUBCSJDLT্ͰϞσϧΛߏங த 14 ʢࣄྫհʣ'JOBODJBM5JNFTͰػೳͷಡऀΤϯήʔδϝ ϯτʹ༩͑ΔӨڹͱɺͦͷػೳͷར༻։࢝֬Λܭࢉ͠ػೳ ͷਪનදࣔΛߦͬͨ <>)PXXFDBMDVMBUFUIF/FYU#FTU "DUJPOGPS'5SFBEFST ͞ΒͳΔΤϯήʔδͷͨΊ ͷػೳͷਪન /FYU#FTU "DUJPOʢ/#"ʣ
·ͱΊ 15 -57ܭࢉ ܖͷࢿ࢈ՁධՁ $IVSOੳΛͬͨ ࢪࡦධՁ -57վળʹ ػցֶशΛ -57Λ͏͜ͱͰɺࡒձܭʴܖ
Ͱݟ͑ͮΒ͍ظతͳ ੑධՁઓུΛཱͯ͘͢ͳ Δɻ ղʹ૬ؔͷߴ͍Τϯήʔδϝ ϯτείΞΛఆٛͨ͠ɻ ͦͷมͰ͋Δ๚සফඅه ࣄຊΛ,1*ʹ͍ͯ͠Δɻ ײԠੳʹΑΔࣄۀߩݙͷఆྔ ධՁߦ͍ͬͯΔɻ վળ͖͢υϥΠόʔ͕ಛఆ ͞Ε͍ͯΕσʔλαΠΤϯε XFCΤϯδχΞϦιʔεΛ ߟ͑͘͢ͳΔɻ
ࢀߟจݙ 16
ࢀߟจݙ 17 w <>4VOJM(VQUB%POBME3-FINBOO $6450.&34"4"44&54 +063/"-0'*/5&3"$5*7&."3,&5*/(70-6.&/6.#&3 8*/5&3IUUQTXXXHTCDPMVNCJBFEVNZHTCGBDVMUZSFTFBSDI QVC fi
MFTHVQUB@DVTUPNFSTQEG BDDFTTFE'FCSVBSZ w <>ᖒل -JGFUJNF7BMVFܦిࢠ൛ͰͷࣄۀඪͷཧIUUQT IBDLOJLLFJDPNCMPHBEWFOU@MUW BDDFTTFE'FCSVBSZ w <>8FC୲ऀϑΥʔϥϜ ܦిࢠ൛Λʹಋ͍ͨσʔλ׆༻ज़ɻ%9 Λ્ΉͭͷนͱΓӽ͑ํIUUQTXFCUBOJNQSFTTDPKQF BDDFTTFE'FCSVBSZ w <>(FPSHF,BTUSJOBLJT )PXXFDBMDVMBUFUIF/FYU#FTU"DUJPOGPS '5SFBEFSTIUUQTNFEJVNDPNGUQSPEVDUUFDIOPMPHZIPXXF DBMDVMBUFUIFOFYUCFTUBDUJPOGPSGUSFBEFSTFEBCB BDDFTTFE 'FCSVBSZ