MTS MARKETING

SERVICES

ADDENDUM

Appendix 7 – MTS Marketing Services Addendum

The Partner and any Operator (hereinafter referred to as “Customer” or “Operator”) agree to be bound by the terms and conditions contained in this Appendix that shall apply solely to the MTS Marketing Services. The Parties acknowledge and agree that the terms contained herein as applied solely to the MTS Marketing Services shall prevail in the event of a conflict between the terms of this Appendix and any other part of this Agreement (including, but not limited to, appendices, exhibits, amendments, and addenda).

1. GENERAL SERVICES PER MODULE

• Initial full dev & dev support
• Full ops & maintenance of service in Sportradar cloud
• Life time support:

o ongoing optimization of the AI
o for Personalization: if agreed in scope, widget UI based on agreed scope including small/reasonable new feature requests

• Configurable AI (filtering, grouping & boosting)
• for Personalization: personalized & pre-login support, “cold start problem” support
• AI analytics dashboard
• 8×5 DevOps & Dev support
• 24×7 escalation handling SLA
• Supporting all-you-can-eat API requests/widget includes, up to 15 million API requests/ week and 20 million transactions/day
• ISO 27001 & GDPR compliant

2. INTEGRATION

2.1 PERSONALIZATION

2.1.1 Data integration: Sportradar AI will receive:
o Betting/wagering transactions, from Sportradar or other sources
o Sports: offering via Unified Odds Feed and other sources if required
o Casino: operator game offering, by brand
o User journey impressions and click tracking, via Sportradar, Operator or any other source.

2.1.2 Frontend integration: Customer will be using the Sportradar API, documentation communicated by Sportradar, or widgets built for specific use cases defined in the project scope, or any CRM system agreed by the Parties.

2.2 PLAYER INSIGHTS

2.2.1 Data integration: Sportradar will receive:
o Betting/wagering transactions, from Sportradar or other sources
o Sports: offering via Unified Odds Feed and other sources if required
o Casino: operator game offering, by brand
o Financial data via Operator or Platform feed, or other sources
o Sign up Data and Customer Information via Operator or Platform feed, or other sources
o Bonus Data via Operator or Platform feed, or other sources
o VIP Loyalty Program via Operator or Platform feed, or other sources

2.2.2 Data output: Sportradar will send agreed output based on packages subscribed as per contract and the MTS Marketing Services SOW.

2.3 Data input requirements

It is acknowledged that while the onus is on Sportradar to implement the Service, that:

• the Customer and its sub-contractors provide to Sportradar the data, historic one-time & recurring feeds as agreed between the parties defined in the Schedules to this agreement, and
• this Data and its sub-contractors will be available to Sportradar for use and shall be treated as Confidential Information as set out in this Agreement, and
• Sportradar will simply be processing information provided by the Customer, and
• Sportradar shall have no responsibility for verifying or checking the accuracy or completeness of such information, and
• the timely, correct and complete provision of the required data, access to certain Customer staff and other inputs identified by Sportradar or the Customer (‘Data’), is critical to the viability and success of the Service; and
• any issues in the data feed shall be monitored, raised and handled as set out in the SLA in Schedule 1, and
• the quality of the Service provided is reliant upon the quality and quantity of the Customer Data. In the case of issues Sportradar cannot be made liable for interruptions or quality reductions, and
• the Customer provides to Sportradar data as agreed between the Parties during the project
• Sportradar may make further data requests from time to time.

3. CUSTOMERS MAY CHOOSE FROM THE FOLLOWING PACKAGES:

a. Sports – Recommender Model
Sportradar will provide its Customers with most relevant sports, leagues, events, participants, and markets, personalized per user, to allow providing sports content by relevance to the user. Recommendations are pre-match & live, personalized or pre-login.
This includes Popular & Trending recommendations across the sportsbook. Prematch & live, personalized & pre-login sports, leagues and events.

b. Sports – Similarity Model
Events: Relevant related events & bets, to users’ single or multiple bet selections. After placing a bet, or as suggestion when browsing events. As well as Leagues & Sports: Suggesting related leagues and sports to the one users are already interested in.

c. Sports – Personalized Accas, BetBuilder
Pre-populating Acca or BetBuilder features with events & bets the user already likes, or help users build their bet with relevant recommendations.

d. Sports -Personalized Scheduled Push
Pushing personalized bet offers to players via API. The AI service directly integrates with an agreed operator or sports platform UI, app push notification, or any mail or SMS dispatch service.

e. Sports – Player Value Prediction
Al-driven predictions for Future Player Value and Wager Amount based on an operator’s internal VIP Loyalty Program.

• Acquisition Marketing: Predict the Future Player Value & use it to Calculate Return on Marketing Invest for any Acquisition Channel/Campaign.

o Player segmentation: Create segments based on player value for targeted CRM Campaigns.
o Standard Output: 30, 60, 90, 180 & 365 days, further outputs configurable

• Wager Amount: Predict the amount a player will wager in the future.

o Player segmentation: Segment/group player into low, medium and high and target them with different campaigns.
o Standard Output 7, 14, 30, 60, 90, 180 & 365, further outputs configurable

f. Sports – Retention Marketing AI
Each player will be scored a percentage chance of churning in the next 7, 14 and 30 days’ time (further outputs configurable). The churn score will be produced for players who were active in the last 30 days.

• Repeat visits: Predict the number of days a player will be active in the future.

o Segmentation: Group players from low frequency to high frequency & combine with Wagering segments to create Value Segments
o Standard Outputs: 7, 14, 30, 60, 90, 180 & 365 days, further outputs configurable

• Churn percentage: Predict the likelihood of a player churning in the next 7, 14 & 30 days & compare the churn score between 7 and 30 days (further outputs configurable).

o Segmentation: Group player.; based on their score and target the high risk of churn and high player value customers first.

g. Sports – VIP predictor
VIP Prediction: Predict which new or existing customers can become a VIP the next 30 days (further outputs configurable) using the operator’s VIP loyalty scheme to classify which tier a customers belongs to.
Standard Output: % probability of achieving any VIP status & then broken down by % by tier.

h. Sports – Personalized Player Bonusing
Recommend the bonus type and suggest a bonus amount for each player and provide AB test analysis.
BONUS TYPE & AMT: Identify type of bonus suited to players (Signup Bonus, Reload Bonus, Cash, Points, Free Bets etc) & recommend an optimum amount to bonus

i. Casino – Recommender model
Personalizing games, game categories, themes, mechanics and game studios. Integrated via API, feed or frontend component across player UI, CRM database or messaging services.
This includes Popular & Trending recommendations. Popular: The most popular games played by users across the site, brand or country. Trending: The hottest trending = fastest growing games across the site, brand or country.

j. Casino – Similarity Model
Suggesting related games, themes and other data points to players, to the ones they already / recently played.

k. Casino – Gamedatabase
Game database of casino games of the operator’s or across the industry, tagged across >80 game categories, themes, mechanics, game studios, wagering-specific and other tags. Updated regularly with new >100 games per month. Accessible via full table download or API endpoint.

l. Casino – Smart Search
Using AI to predict which casino games a player is searching for, can handle searches for themes, suppliers and game mechanics and features. Will also handle spelling mistakes to return relevant results to each player

m. Casino – Player Value Prediction
Al-driven predictions for Future Player Value and Wager Amount based on an operator’s internal VIP Loyalty Program.

• Acquisition Marketing: Predict the Future Player Value & use it to Calculate Return on Marketing Invest for any Acquisition Channel/Campaign.

o Player segmentation: Create segments based on player value for targeted CRM Campaigns.
o Standard Output: 30, 60, 90, 180 & 365 days, further outputs configurable

• Wager Amount: Predict the amount a player will wager in the future.

o Player segmentation: Segment/group player into low, medium and high and target them with different campaigns.
o Standard Output 7, 14, 30, 60, 90, 180 & 365, further outputs configurable

n. Casino – Retention Marketing AI
Each player will be scored a percentage chance of churning in the next 7, 14 and 30 days’ time (further outputs configurable). The churn score will be produced for players who were active in the last 30 days.

• Repeat visits: Predict the number of days a player will be active in the future.

o Segmentation: Group players from low frequency to high frequency & combine with Wagering segments to create Value Segments
o Standard Outputs: 7, 14, 30, 60, 90, 180 & 365 days, further outputs configurable

• Churn percentage: Predict the likelihood of a player churning in the next 7, 14 & 30 days & compare the churn score between 7 and 30 days (further outputs configurable).

o Segmentation: Group player.; based on their score and target the high risk of churn and high player value customers first.

o. Casino – VIP predictor
VIP Prediction: Predict which new or existing customers can become a VIP the next 30 days (further outputs configurable) using the operator’s VIP loyalty scheme to classify which tier a customers belongs to.
Standard Output: % probability of achieving any VIP status & then broken down by % by tier.

p. Casino – Personalized Player Bonusing
Recommend the bonus type and suggest a bonus amount for each player and provide AB test analysis.
BONUS TYPE & AMT: Identify type of bonus suited to players (Signup Bonus, Reload Bonus, Cash, Points, Free Bets etc) & recommend an optimum amount to bonus

q. Bingo – Basic
Bingo CH Manager enables operators to learn, adapt and react to insights in player chat data. Enables effective management of staff in order to increase wagering & player sessions through engagement.

r. Bingo – Add-on: Casino Wagers
This add-on incorporates players’ Casino wagers that occur during their Bingo gameplay.

s. Bingo – Add-on: Add. Language Pack
Basic package includes English by default, but any language can be added with this add-on.

t. Bingo – Add-on: Real Time API
Basic package includes daily upload of all data, this can be modified to work in Real-Time through API with this add-on.

4. DETAILED SCOPE OF THE PROJECT

Detailed scope of the Services, project details and product configurations are documented in a statement of work (“MTS Marketing Services SOW”) communicated to the Customer by Sportradar and agreed between the parties in electronic form.
The MTS Marketing Services SOW shall include, but is not limited to, the following decision points:

• Brands
• Countries / Regions / Locale
• Languages
• Channel: mobile web, mobile ative app, desktop web, retail, …
• Users: realmoney, playmoney, freemium, other
• Sports: sports types, including racing
• Casino: game types, game tags/categories/mechanics/other labels
• Player Insights AI: value segment definition, time range, other
• Features launched: selected from above core offering, and adjusted in the MTS Marketing Services SOWs
• QA & Rollout: defined and run by Customer, support by Sportradar
• AI Analytics, KPIs and dashboards agreed

5. SERVICE LEVEL AGREEMENT

• Normal operations:
Normal operation hours are 9AM-5PM UK time, Monday through Friday, apart from UK bank holidays.

• Production systems:
Production systems are any components deployed by Sportradar to operate the features fulfilling its obligations under this agreement.

• Access and use :
Sportradar shall make the Service available to the Operator twenty-four (24) hours a day, seven days a week, excluding the necessary maintenance breaks.
Sportradar strives to achieve the highest possible availability of the Service and guarantees a 99.8% calendar monthly uptime of the Service.
The uptime shall be determined calendar monthly and the following formula shall be used: Service Hours rounded to full minutes – (less) length of break rounded to full minutes / (divided by) Service Hours rounded to full minutes x (multiplied by) 100.
The time used for maintenance work shall not be taken into consideration to the detriment of Sportradar when determining uptime.

• Planned Maintenance, Updates, and Change Management:

o Sportradar undertakes to coordinate, within reason, any maintenance with the Customer in order to minimize impact on Customer Operations. Such Components are:

• Production systems with intrusive maintenance updates:

• API and inference module
• database, daily or live data import, db update

• Production systems with non-intrusive maintenance updates

• training module(s), data import for training
• any admin interfaces or other UI in operational use by the Operator

o Any planned maintenance or updates affecting production systems (both intrusive and non-intrusive) will be communicated in advance, 2 working days in advance for intrusive updates, at least 3 hours for non-intrusive updates.

o Maintenance frequency

• Given the interest of both Parties in regular updates to the platform in order to facilitate regular improvements, both Parties understand and agree, that Maintenance Frequency may be difficult to predict during the first year of operations.
• Non-intrusive updates will be carried out as often as reasonably required. Intrusive updates (planned and notified downtime) is to be done as a max 1-2 times per week, unless exceptional circumstances arise. See below for more details on unplanned downtime.

o Maintenance Timing

• Intrusive updates: VAIX will work with the Operator to ensure that Maintenance is performed during low-traffic times, such as 6am GMT on weekdays
• Non-intrusive updates: not during peak time, a guideline is between 6am – 3pm GMT on weekdays.

o Communication plan for planned updates

• Unplanned downtimes, defects and escalation management:

o Operational contacts

o Classification of Defects and Solution

Defects are categorized as severity 1 and 2 in the table below while non-material Defects are categorized as severity 3, 4 and 5. Either Party is responsible for solving or provision of temporary work-around within the time intervals specified below.