Kritter Introduction 
By - Arjun Som (CEO) 
arjun.som@kritter.in 
12-12-2014
Ad-tech – Business Case 
An ad-tech company running trade on top of its 
home grown, third party licensed, or open 
source technology. 
-- Why extend or home grow, technology 
-- Why RTB 
-- Why Programmatic 
-- Why buy ready products and not build 
-- Why Kritter 
12-12-2014
Business Case – Why 
Intellectual Property: Intellectual property is not ad-serving, RTB bidder, big data platform, Which most 
ad-tech companies claim. IP is when trade happens on these technologies, the learning your campaign 
managers get to isolate ROI techniques and provide scale to advertisers. 
Data: Inventory/supply, targeting, how granular should the targeting be, what creatives work, what time 
of day, what bid gets me the best traffic. That sliced by advertiser product offering. That sliced by 
advertiser feedback of the user in the flow, registered, engaged, bought, repeat buy etc.. Harnessing this 
data could lead to automating media planning and buying 
Programmatic: A system that incorporates media teams learning, provide a platform to automate 
redundant activities, moves them from low level to high level tasks of analyzing and acting 
RTB: All supply is moving to exchanges with only remnant staying with ad-networks. All demand moving 
to platform with access to multiple supply sources to give single point access to media buying. Making 
your media buying unbiased with multiple supply sources is the market direction. 
Vertical Integration: Basic Economics, 80% of ad dollars are for performance, and 20% for brand. 
Stabilizing and de-risking your supply is the natural progression. Giving your sales team a technology 
where they can say YES to every business enquiry, with ease of global scale and no risk of fill rates 
Supply sources: Apart from keywords, which is company owned supply. All the other supply is with 
exchanges, or privately sold by premium publishers. All non exclusive supply sources are opening up for 
media buyers to increase competition for their inventory which drives up their prices.
Business Case – Advantages 
• Access the best supply through exchanges 
• De-risk your supply, from regular eCPM and fill demands of 
publishers 
• Embed your media buying team’s learning into the technology 
bringing efficiency, fewer people can handle more business 
• Compete with global giants with a local approach, by acquiring 
technology they have and utilizing relationships you have 
• Avoid paying recurring fees against licensed technology, own it! 
• Increase relevance with your advertisers, by providing a full array of 
media buying capabilities 
• Plan the advertiser’s budget, instead of getting one assigned to you 
• Have device agnostic serving and buying technology
Kritter! 
12-12-2014
Founding Team 
Arjun Som: BE in Computer Science, GGSIPU. 
- Specializes in the set up of the direct sales, inside sales and online sales channels, customer acquisition, Org set up, and deep 
understanding of start up execution 
- Worked in various roles in the digital marketing space, including affiliate, agency, demand, supply and network, in companies like 
InMobi, Sapient, iGlobalMedia, and Indiaplaza 
- Leads Kritter’s business development, marketing, and over all PnL 
in.linkedin.com/in/arjunsom/ 
Kshitij Sooryavanshi: B.Tech., Computer Science, Indian Institute of Technology, Kanpur. 
- Specializes in core engineering, solving, machine learning, mathematical computing, algorithmic tasks, data mining, and predictive 
modeling 
- Worked in companies like, InMobi, Vizexperts, Apache design solutions (acquired by Ansys) 
- Has varied experience in internet startup product development 
- Leads Kritter’s product line development, managing Programmatic buying and selling platform 
in.linkedin.com/pub/kshitij-sooryavanshi/31/b43/b64 
Rohan Rai: B.Tech + M.Tech. (dual), Computer Science, Indian Institute of Technology, Delhi. 
- Specializes in data engineering, distributed data platform,big data and setting up data management platform 
- Experienced in start up domain having worked in early to large stages of InMobi, ZivaSoftware, Bizense, Qwest 
- Has varied horizontal experience in new age technology cutting across search, ops, dev ops, business rule management systems 
- Leads kritter’s data engineering 
in.linkedin.com/in/rohanrai 
Vikrant Chahar: B.Tech., Computer Science, Indian Institute of Technology, Guwahati. 
- Specializes in core engineering, solving ad serving, ad relevance, ranking, selection and market place dynamics 
- Worked in companies like, InMobi, Stratoshear, Globallogic, 
- Has varied experience in both software product and services 
- Leads Kritter’s product line development managing ad-serving 
in.linkedin.com/pub/vikrant-chahar/13/223/676 
12-12-2014
Kritter Business Model 
$ 
Adv Consumer 
12-12-2014 
Advertising / Technology Players 
Inventory 
$ $ 
Product 
and 
Services 
$
Kritter - Solution 
Kritter is a pure play product and services company in the digital advertising 
space. Kritter provides industry players with highly customizable products and 
flexible services to run their digital business. 
Kritter Products: Coded in 
-- DSP : Java 
-- Ad-server : Java 
-- Bidder: Java, Python, R, C++ 
-- Data platform (Big Data): Pig 
Kritter Services: 
-- Integrating above products into client’s stack 
-- Customizing Kritter products to fit client’s business needs 
-- Open up Kritter product modules, and integrate it to client’s technology 
-- Offshore outsourcing existing engineering tasks 
12-12-2014
Kritter – CRM Module 
A module to integrate your media buying 
technology to the advertiser’s CRM 
-- For automating media planning, and spending 
-- Controlling spends on real-time feedback 
-- Integrating LTV with media buying 
-- Increasing switching cost 
-- Aligning as a sales channel for long term ad 
spends 
12-12-2014
Data! (CRM Module) 
Marketing KPIs 
PC: 
Permutations and combinations sliced by billions of interactions. 
-- Registration 
-- Engagement and user quality 
-- Credit card sync 
-- Transaction 
Mobile: 
-- App Download 
-- Engagement and user quality 
-- Credit card sync 
-- Transaction 
Milestones 
- Ad view, to click, to visit, to 
Registration/Download to 
Transaction 
- Transaction to repeat buy 
Marketing parameters 
- LTV : Life time value (ARPU) 
- CPR/I: Cost per registration/Install 
- CPE: Cost per engagement 
- CPT: Cost per transacting customer 
- CPRU: Cost per repeat user 
Customer Touch points: 
-- Site/App 
-- ISP 
-- Device 
-- Creative 
-- On click landing page 
-- User flow 
Ad-tech system: 
Campaign: 
-- Budget 
-- Bid 
-- Ad, creative, landing page 
Targeting: 
-- Geo 
-- Placement 
-- Site 
-- Device 
-- OS 
-- ISP 
-- User/cookie 
Supply: 
-- Exchange 
-- Site/App 
-- Affiliate 
-- Aggregator 
Reporting: 
Trade reporting sliced by all the parameters
Kritter - USP 
-- Built a business model around enterprise ad-tech software 
-- Deploys products with source code on client’s stack 
-- Customizable products to fit business needs and client’s technology 
-- Experience in building high scale ad-tech 
-- Industry experience, to help streamline business customization and product 
roadmap 
-- Helps companies take new models to market with lightning speed 
-- All new technology requires a learning curve, specialized skill, and revenue risk to 
learn if it works. Kritter helps you de-risk! 
-- 80% of the software code has 20% of the trading impact, and 20% of the software 
code has 80% of the trading impact. Kritter products provide the 80% code right out of 
the box. Your engineers can focus on tweaking the 20% code impacting trade, using 
Kritter algorithms as a base model to improve upon. This keeps engineering team lean 
and high skilled. (Computational Mathematics/machine learning specialists and data 
scientists) 
-- Helps profit positive companies acquire technology keeping investments lean 
-- We are a start-up
K-DSP- Architecture 
Agency 
Agency 
Dev 
Adv 
12-12-2014 
- Campaign Set up 
- Targeting 
- Creatives 
- Bid & Budget 
Ad- Server 
-- App and Placement 
set up 
- Margin 
- eCPMFloor 
- Ad preferences 
Pub 1 
Pub 2 
Pub 3 
Pub N 
-- Event tracking 
-- Billing 
-- Internal and External Bid setting 
-- Ad and Site approval 
-- eCPM maximizer 
-- Selection and Filtering 
-- CPM, CPC, CPI billing 
Bidder 
Exng 1 
Exng 2 
Exng 3 
-- CPM to CPM bidding Exng N 
-- CPC to CPM bidding
Key Features - DSP 
Highlights: 
• Integrate multiple supply sources: Websites, Applications, Exchanges, Aggregator, SSPs 
• Campaign creation, budget management, bid management 
• Ad creation, IAB standard creative format uploading, targeting 
• Billing system to handle house clearing 
• Reporting: Demand, Supply and Network statistics, along with saved queries 
Key features: 
APIs: Integrate any third party campaigns management or supply management tools with APIS. Build self serve 
UI and integrate it to the back end with APIs 
Org Fitment: Customize the admin UI to superimpose your organization structure. This will allow you to set up 
access control across your ad-operations 
Margin Management: Set budgets on the demand side by having internal and external bids and budgets. Set 
payout margins for publishers and eCPM floors to manage supply buying costs 
Content Guidelines: Ad approval and site approval queue, whenever a new ad or site is created or existing 
ones are modified, this comes into the queue for approval. This ensures your content guidelines are maintained 
centrally, and your ad-ops team has a organized way of making sure what is going into the system to avoid irate 
advertisers and publishers 
Hygiene setting: 4 types of hygiene mapping in the system. Premium, Family safe, performance, mature. To 
make sure the right set of ads are going onto the right set of inventory, and to maintain system hygiene.
K- ad-server - Architecture 
Campaign 
Ad 
Creative 
Targeting 
Predictive 
Analytics 
Ranking 
Algo 
Rich 
Media 
Serving 
Feedback 
Bidder 
Summary 
Fraud 
Detection 
Post 
Impression 
Marketing 
KPI 
Billing 
Site/App 
Exchange 
Filters and 
Floors 
Predictive 
Analytics 
Clearing 
House 
Revenue 
Max 
Rich 
Media 
Market 
Place 
Margins 
Manual 
Overrides 
DMP
Key Features – Ad-server 
Highlights: 
• Mobile and PC ad-serving :Banner, text, and, rich media ad delivery on exchanges and direct inventory 
• Post impression tracking, attribution right up to conversion 
• Conversion based pricing and selection, targeting manipulation on the fly for maximizing ROI 
• Fraud detection and mitigation 
• Connect multiple supply Sources, publishers, exchanges, aggregators, etc… 
Key features: 
Selection: Hundreds of advertisers, with hundreds of campaigns, thousands of ads, thousands of creatives and 
millions of permutation and combinations with targeting and placement preferences. The ad-server 
accommodates and automates. 
Filtering: Advertisers want to be on some inventory and don't on some, they would like some categories and 
not like some, they choose granular targeting to ensure measurability across thousands of ads. Supply partners 
have brand protection clauses, anti competition ad-preferences, floor pricing, and house ads. The ad-server 
honors and automates all. 
Ranking: Advertisers want acquisitions, the publisher wants eCPM and fill, while the network wants to make 
profits from all interactions. The ad-server has a unique ranking algorithm that automates optimization.
K-Bidder - Architecture 
Ad-server 
Exchange - 
1 
Exchange - 
2 
Exchange - 
n 
Algorithms 
Adapter Integratio 
ns 
Data 
Summaries 
Campaign 
Budget 
Campaign 
Details 
Post 
Impression 
Ad-serve 
Logs 
Conversion 
DMP 
Win 
notification 
Supply 
Forecasting 
Price 
Landscape 
Offline 
Bidding 
CTR 
Prediction 
Online Bid 
Adjustment 
Risk 
Managemen 
t
Key Features 
Highlights: 
• RTB bidding across any number of exchanges 
• Offline bid feedback provided to the online bidding agent to minimize server costs 
• Single point management across hundreds of advertisers, thousands of campaigns, thousands of ads, and 
multiple exchanges. Keeping ROI and profits positive 
Key features: 
• Profit Maximization: Advertiser enters the max bid he is willing to pay for targeted traffic, while the 
exchange bid request keeps modifying over time. The Bidder component ensures it bids a value lower than 
the max bid, and for which traffic is available. This provides scale to and ensures profits for the network 
• Supply Forecasting: Forecasts traffic against targeting granularity and combinations. Learning from 
interactions over time. 
• Price Landscape: Predicts the price of the inventory, learning from interactions. 
• CTR Prediction: Our bidder allows for your advertisers to bid on CPC while the bidder predicts CTR and 
bids on CPM on the exchanges. 
• Risk management: Supply through exchanges is ever changing, interactions once positive could go into 
heavy losses in a matter of seconds. Kritter’s unique financial risk module, monitors and mitigates this.
K-DP - Architecture 
Ad-server 
Supply 
Products 
Demand 
Products 
Advertiser Publisher 
Transport 
Raw Data Store (Distributed Data System) 
Transport 
Processing 
Fan Outs 
HDF 
S 
MySQL 
Propri 
ety 
Hive 
Reporting API 
Reporting Analytics Feedback 
DMPs, Data 
Enricher, 
Data 
Augmenters 
Machine 
Learning
Key Features 
Highlights: 
• Integrate with multiple logical boxes, like ad-serving, bidder, buy side, and sell side products. To transport 
and store data of various dimensions and granularity 
• Fan outs/APIs: Pipe this data in to various front end products, like analytics, interface based query system, 
customer reporting. Pipe processed information into machine learning systems to empower various 
algorithms. 
• Monitoring and alerting: Create various rules and processes to alert business and technology teams on 
anomalies and fluctuations. 
• Processing: Create custom rules to process the raw data, integrate external DMPs to enrich data. The 
processing will keep in mind the varied outcome of the data needed for the business and technology 
Key features: 
• Big Data: This platform transitions with your business scale with seamless big data/distributed data 
infrastructure. Works on RDBMS, and when your data hits inflection point, can se ported into a Hadoop 
infrastructure without any coding 
• DMPs: Integrate external or home grown DMPs like bluekai, lotame, factual, etc to enrich ad-serving and 
bidding algorithms 
• Customize: The DP can be customized to fit your business models like, hyper local, lat long, retargeting, 
UDID based serving, cookie technology
Kritter Pricing 
Product Pricing: One-time price for a life-time license 
• Business training and handover 
• Technical training and handover - Documentation 
30 days: Acceptance period – Free 
30 days: Warranty period - Free 
Services: Charged on time spent 
• White labeling the product 
• Integrating the product with existing technology 
• Customizing client’s business model into the product 
Maintenance: 
• Reactive Maintenance: Send an email to Kritter support and pay hourly charges for debugging (In case of client 
having an existing technology team, wanting to manage tech operations) 
• Proactive Maintenance: Monthly fixed charges to manage 24/7 tech operations and debugging of the software (In 
case of client having no technology team, wanting to outsource full tech operations 
Enhancements: 
• Enhance patch: Kritter will keep releasing patches to enhance its software and will notify the customer of the new 
features. Customer can pay a one time patch update price to acquire new features 
• Enhancer retainer: Post delivery if the customer is looking at regular enhancement of the software. A pod will be 
built within Kritter to manage the road map of all developments, having a product manager, project manager, and 
engineers (based on work load). Patches included
Kritter Product- Infrastructure cost 150mn Requests/day
Kritter Product- Infrastructure cost 1.5bn Requests/day
Kritter Vs. Appnexus (Rough Estimate)
Pricing Models 
Product Acquire: 
-- License and customization Price 
split by deployment stages 
-- Engineering handover 
-- Acceptance and warranty 
period 
-- Maintenance as required by 
client 
Who? 
- Clients with Engineering team 
- Clients with existing tech 
- Clients looking to extend 
business and tech 
Product Hire & Acquire: 
-- License and customization Price 
split into monthly installments 
-- Once installments are 
completely paid, the license is 
handed over 
-- Engineering handover 
-- Acceptance and warranty 
period 
-- Maintenance fee incorporated 
into installments 
Who? 
- Clients with no engg team 
- Clients with no tech 
- Clients looking to move 
business from third party to 
their own 
- Clients looking to change 
business model through 
technology 
- Clients trying tests with tech 
and business model 
Specific tech module 
Acquire: 
-- Part payment for Integration 
and deployment into stack 
-- Engineering handover 
-- Acceptance and warranty 
period 
-- Rest of the payment 
-- Maintenance as required by 
client 
Who? 
- Client’s looking to acquire 
specific modules, like, RTB, 
Fraud det, Billing, Algorithms, 
etc… 
- Clients with Engineering team 
- Clients with existing tech 
- Clients looking to extend 
business and tech
Thank You 
www.kritter.in 
Enabling technology for advertising 
business 
12-12-2014

Kritter introduction - technology player

  • 1.
    Kritter Introduction By- Arjun Som (CEO) arjun.som@kritter.in 12-12-2014
  • 2.
    Ad-tech – BusinessCase An ad-tech company running trade on top of its home grown, third party licensed, or open source technology. -- Why extend or home grow, technology -- Why RTB -- Why Programmatic -- Why buy ready products and not build -- Why Kritter 12-12-2014
  • 3.
    Business Case –Why Intellectual Property: Intellectual property is not ad-serving, RTB bidder, big data platform, Which most ad-tech companies claim. IP is when trade happens on these technologies, the learning your campaign managers get to isolate ROI techniques and provide scale to advertisers. Data: Inventory/supply, targeting, how granular should the targeting be, what creatives work, what time of day, what bid gets me the best traffic. That sliced by advertiser product offering. That sliced by advertiser feedback of the user in the flow, registered, engaged, bought, repeat buy etc.. Harnessing this data could lead to automating media planning and buying Programmatic: A system that incorporates media teams learning, provide a platform to automate redundant activities, moves them from low level to high level tasks of analyzing and acting RTB: All supply is moving to exchanges with only remnant staying with ad-networks. All demand moving to platform with access to multiple supply sources to give single point access to media buying. Making your media buying unbiased with multiple supply sources is the market direction. Vertical Integration: Basic Economics, 80% of ad dollars are for performance, and 20% for brand. Stabilizing and de-risking your supply is the natural progression. Giving your sales team a technology where they can say YES to every business enquiry, with ease of global scale and no risk of fill rates Supply sources: Apart from keywords, which is company owned supply. All the other supply is with exchanges, or privately sold by premium publishers. All non exclusive supply sources are opening up for media buyers to increase competition for their inventory which drives up their prices.
  • 4.
    Business Case –Advantages • Access the best supply through exchanges • De-risk your supply, from regular eCPM and fill demands of publishers • Embed your media buying team’s learning into the technology bringing efficiency, fewer people can handle more business • Compete with global giants with a local approach, by acquiring technology they have and utilizing relationships you have • Avoid paying recurring fees against licensed technology, own it! • Increase relevance with your advertisers, by providing a full array of media buying capabilities • Plan the advertiser’s budget, instead of getting one assigned to you • Have device agnostic serving and buying technology
  • 5.
  • 6.
    Founding Team ArjunSom: BE in Computer Science, GGSIPU. - Specializes in the set up of the direct sales, inside sales and online sales channels, customer acquisition, Org set up, and deep understanding of start up execution - Worked in various roles in the digital marketing space, including affiliate, agency, demand, supply and network, in companies like InMobi, Sapient, iGlobalMedia, and Indiaplaza - Leads Kritter’s business development, marketing, and over all PnL in.linkedin.com/in/arjunsom/ Kshitij Sooryavanshi: B.Tech., Computer Science, Indian Institute of Technology, Kanpur. - Specializes in core engineering, solving, machine learning, mathematical computing, algorithmic tasks, data mining, and predictive modeling - Worked in companies like, InMobi, Vizexperts, Apache design solutions (acquired by Ansys) - Has varied experience in internet startup product development - Leads Kritter’s product line development, managing Programmatic buying and selling platform in.linkedin.com/pub/kshitij-sooryavanshi/31/b43/b64 Rohan Rai: B.Tech + M.Tech. (dual), Computer Science, Indian Institute of Technology, Delhi. - Specializes in data engineering, distributed data platform,big data and setting up data management platform - Experienced in start up domain having worked in early to large stages of InMobi, ZivaSoftware, Bizense, Qwest - Has varied horizontal experience in new age technology cutting across search, ops, dev ops, business rule management systems - Leads kritter’s data engineering in.linkedin.com/in/rohanrai Vikrant Chahar: B.Tech., Computer Science, Indian Institute of Technology, Guwahati. - Specializes in core engineering, solving ad serving, ad relevance, ranking, selection and market place dynamics - Worked in companies like, InMobi, Stratoshear, Globallogic, - Has varied experience in both software product and services - Leads Kritter’s product line development managing ad-serving in.linkedin.com/pub/vikrant-chahar/13/223/676 12-12-2014
  • 7.
    Kritter Business Model $ Adv Consumer 12-12-2014 Advertising / Technology Players Inventory $ $ Product and Services $
  • 8.
    Kritter - Solution Kritter is a pure play product and services company in the digital advertising space. Kritter provides industry players with highly customizable products and flexible services to run their digital business. Kritter Products: Coded in -- DSP : Java -- Ad-server : Java -- Bidder: Java, Python, R, C++ -- Data platform (Big Data): Pig Kritter Services: -- Integrating above products into client’s stack -- Customizing Kritter products to fit client’s business needs -- Open up Kritter product modules, and integrate it to client’s technology -- Offshore outsourcing existing engineering tasks 12-12-2014
  • 9.
    Kritter – CRMModule A module to integrate your media buying technology to the advertiser’s CRM -- For automating media planning, and spending -- Controlling spends on real-time feedback -- Integrating LTV with media buying -- Increasing switching cost -- Aligning as a sales channel for long term ad spends 12-12-2014
  • 10.
    Data! (CRM Module) Marketing KPIs PC: Permutations and combinations sliced by billions of interactions. -- Registration -- Engagement and user quality -- Credit card sync -- Transaction Mobile: -- App Download -- Engagement and user quality -- Credit card sync -- Transaction Milestones - Ad view, to click, to visit, to Registration/Download to Transaction - Transaction to repeat buy Marketing parameters - LTV : Life time value (ARPU) - CPR/I: Cost per registration/Install - CPE: Cost per engagement - CPT: Cost per transacting customer - CPRU: Cost per repeat user Customer Touch points: -- Site/App -- ISP -- Device -- Creative -- On click landing page -- User flow Ad-tech system: Campaign: -- Budget -- Bid -- Ad, creative, landing page Targeting: -- Geo -- Placement -- Site -- Device -- OS -- ISP -- User/cookie Supply: -- Exchange -- Site/App -- Affiliate -- Aggregator Reporting: Trade reporting sliced by all the parameters
  • 11.
    Kritter - USP -- Built a business model around enterprise ad-tech software -- Deploys products with source code on client’s stack -- Customizable products to fit business needs and client’s technology -- Experience in building high scale ad-tech -- Industry experience, to help streamline business customization and product roadmap -- Helps companies take new models to market with lightning speed -- All new technology requires a learning curve, specialized skill, and revenue risk to learn if it works. Kritter helps you de-risk! -- 80% of the software code has 20% of the trading impact, and 20% of the software code has 80% of the trading impact. Kritter products provide the 80% code right out of the box. Your engineers can focus on tweaking the 20% code impacting trade, using Kritter algorithms as a base model to improve upon. This keeps engineering team lean and high skilled. (Computational Mathematics/machine learning specialists and data scientists) -- Helps profit positive companies acquire technology keeping investments lean -- We are a start-up
  • 12.
    K-DSP- Architecture Agency Agency Dev Adv 12-12-2014 - Campaign Set up - Targeting - Creatives - Bid & Budget Ad- Server -- App and Placement set up - Margin - eCPMFloor - Ad preferences Pub 1 Pub 2 Pub 3 Pub N -- Event tracking -- Billing -- Internal and External Bid setting -- Ad and Site approval -- eCPM maximizer -- Selection and Filtering -- CPM, CPC, CPI billing Bidder Exng 1 Exng 2 Exng 3 -- CPM to CPM bidding Exng N -- CPC to CPM bidding
  • 13.
    Key Features -DSP Highlights: • Integrate multiple supply sources: Websites, Applications, Exchanges, Aggregator, SSPs • Campaign creation, budget management, bid management • Ad creation, IAB standard creative format uploading, targeting • Billing system to handle house clearing • Reporting: Demand, Supply and Network statistics, along with saved queries Key features: APIs: Integrate any third party campaigns management or supply management tools with APIS. Build self serve UI and integrate it to the back end with APIs Org Fitment: Customize the admin UI to superimpose your organization structure. This will allow you to set up access control across your ad-operations Margin Management: Set budgets on the demand side by having internal and external bids and budgets. Set payout margins for publishers and eCPM floors to manage supply buying costs Content Guidelines: Ad approval and site approval queue, whenever a new ad or site is created or existing ones are modified, this comes into the queue for approval. This ensures your content guidelines are maintained centrally, and your ad-ops team has a organized way of making sure what is going into the system to avoid irate advertisers and publishers Hygiene setting: 4 types of hygiene mapping in the system. Premium, Family safe, performance, mature. To make sure the right set of ads are going onto the right set of inventory, and to maintain system hygiene.
  • 14.
    K- ad-server -Architecture Campaign Ad Creative Targeting Predictive Analytics Ranking Algo Rich Media Serving Feedback Bidder Summary Fraud Detection Post Impression Marketing KPI Billing Site/App Exchange Filters and Floors Predictive Analytics Clearing House Revenue Max Rich Media Market Place Margins Manual Overrides DMP
  • 15.
    Key Features –Ad-server Highlights: • Mobile and PC ad-serving :Banner, text, and, rich media ad delivery on exchanges and direct inventory • Post impression tracking, attribution right up to conversion • Conversion based pricing and selection, targeting manipulation on the fly for maximizing ROI • Fraud detection and mitigation • Connect multiple supply Sources, publishers, exchanges, aggregators, etc… Key features: Selection: Hundreds of advertisers, with hundreds of campaigns, thousands of ads, thousands of creatives and millions of permutation and combinations with targeting and placement preferences. The ad-server accommodates and automates. Filtering: Advertisers want to be on some inventory and don't on some, they would like some categories and not like some, they choose granular targeting to ensure measurability across thousands of ads. Supply partners have brand protection clauses, anti competition ad-preferences, floor pricing, and house ads. The ad-server honors and automates all. Ranking: Advertisers want acquisitions, the publisher wants eCPM and fill, while the network wants to make profits from all interactions. The ad-server has a unique ranking algorithm that automates optimization.
  • 16.
    K-Bidder - Architecture Ad-server Exchange - 1 Exchange - 2 Exchange - n Algorithms Adapter Integratio ns Data Summaries Campaign Budget Campaign Details Post Impression Ad-serve Logs Conversion DMP Win notification Supply Forecasting Price Landscape Offline Bidding CTR Prediction Online Bid Adjustment Risk Managemen t
  • 17.
    Key Features Highlights: • RTB bidding across any number of exchanges • Offline bid feedback provided to the online bidding agent to minimize server costs • Single point management across hundreds of advertisers, thousands of campaigns, thousands of ads, and multiple exchanges. Keeping ROI and profits positive Key features: • Profit Maximization: Advertiser enters the max bid he is willing to pay for targeted traffic, while the exchange bid request keeps modifying over time. The Bidder component ensures it bids a value lower than the max bid, and for which traffic is available. This provides scale to and ensures profits for the network • Supply Forecasting: Forecasts traffic against targeting granularity and combinations. Learning from interactions over time. • Price Landscape: Predicts the price of the inventory, learning from interactions. • CTR Prediction: Our bidder allows for your advertisers to bid on CPC while the bidder predicts CTR and bids on CPM on the exchanges. • Risk management: Supply through exchanges is ever changing, interactions once positive could go into heavy losses in a matter of seconds. Kritter’s unique financial risk module, monitors and mitigates this.
  • 18.
    K-DP - Architecture Ad-server Supply Products Demand Products Advertiser Publisher Transport Raw Data Store (Distributed Data System) Transport Processing Fan Outs HDF S MySQL Propri ety Hive Reporting API Reporting Analytics Feedback DMPs, Data Enricher, Data Augmenters Machine Learning
  • 19.
    Key Features Highlights: • Integrate with multiple logical boxes, like ad-serving, bidder, buy side, and sell side products. To transport and store data of various dimensions and granularity • Fan outs/APIs: Pipe this data in to various front end products, like analytics, interface based query system, customer reporting. Pipe processed information into machine learning systems to empower various algorithms. • Monitoring and alerting: Create various rules and processes to alert business and technology teams on anomalies and fluctuations. • Processing: Create custom rules to process the raw data, integrate external DMPs to enrich data. The processing will keep in mind the varied outcome of the data needed for the business and technology Key features: • Big Data: This platform transitions with your business scale with seamless big data/distributed data infrastructure. Works on RDBMS, and when your data hits inflection point, can se ported into a Hadoop infrastructure without any coding • DMPs: Integrate external or home grown DMPs like bluekai, lotame, factual, etc to enrich ad-serving and bidding algorithms • Customize: The DP can be customized to fit your business models like, hyper local, lat long, retargeting, UDID based serving, cookie technology
  • 20.
    Kritter Pricing ProductPricing: One-time price for a life-time license • Business training and handover • Technical training and handover - Documentation 30 days: Acceptance period – Free 30 days: Warranty period - Free Services: Charged on time spent • White labeling the product • Integrating the product with existing technology • Customizing client’s business model into the product Maintenance: • Reactive Maintenance: Send an email to Kritter support and pay hourly charges for debugging (In case of client having an existing technology team, wanting to manage tech operations) • Proactive Maintenance: Monthly fixed charges to manage 24/7 tech operations and debugging of the software (In case of client having no technology team, wanting to outsource full tech operations Enhancements: • Enhance patch: Kritter will keep releasing patches to enhance its software and will notify the customer of the new features. Customer can pay a one time patch update price to acquire new features • Enhancer retainer: Post delivery if the customer is looking at regular enhancement of the software. A pod will be built within Kritter to manage the road map of all developments, having a product manager, project manager, and engineers (based on work load). Patches included
  • 21.
    Kritter Product- Infrastructurecost 150mn Requests/day
  • 22.
    Kritter Product- Infrastructurecost 1.5bn Requests/day
  • 23.
    Kritter Vs. Appnexus(Rough Estimate)
  • 24.
    Pricing Models ProductAcquire: -- License and customization Price split by deployment stages -- Engineering handover -- Acceptance and warranty period -- Maintenance as required by client Who? - Clients with Engineering team - Clients with existing tech - Clients looking to extend business and tech Product Hire & Acquire: -- License and customization Price split into monthly installments -- Once installments are completely paid, the license is handed over -- Engineering handover -- Acceptance and warranty period -- Maintenance fee incorporated into installments Who? - Clients with no engg team - Clients with no tech - Clients looking to move business from third party to their own - Clients looking to change business model through technology - Clients trying tests with tech and business model Specific tech module Acquire: -- Part payment for Integration and deployment into stack -- Engineering handover -- Acceptance and warranty period -- Rest of the payment -- Maintenance as required by client Who? - Client’s looking to acquire specific modules, like, RTB, Fraud det, Billing, Algorithms, etc… - Clients with Engineering team - Clients with existing tech - Clients looking to extend business and tech
  • 25.
    Thank You www.kritter.in Enabling technology for advertising business 12-12-2014