Innovation > Innovation

MARKETING OPTIMIZATION PLATFORM

LG CNS, Seoul / MARKETING OPTIMIZATION PLATFORM / 2023

CampaignCampaignLayout(opens in a new tab)
Case Film
Presentation Image
1 of 0 items

Overview

Credits

OVERVIEW

Why is this work relevant for Innovation?

Nowadays, digital marketing faces a situation such as real-time data analysis, numerous targetting, and infinite inventory. And this is far beyond human ability that requires to perform lots of calculations to find a way of cost-efficient operation.

We adapted AI and mathematical optimization to digital marketing to achieve the best possible performance and we finally realize data-oriented marketing by developing a high-performance solution MOP.

MOP has been proven that it is a technology that innovates the marketing performance and process by confirming that advertising performance could be improved by more than 10% and replaced repetitive tasks by more than 50%.

Background

Digital marketers are seeking a breakthrough due to repetitive tasks, such as exposure status checks, daily reports, and real-time ad adjustments, as well as difficulties caused by different operating methods for each media, including hundreds of thousands of creatives.

Although a number of automation tools are introduced and a large number of manpower is invested, it is hard to meet the needs of current market requirement data-based optimal decision-making considering various indicators, and there are limitations in terms of maintaining consistency of results in handling multiple advertising.

To overcome this, MOP was developed so that all marketers can easily use advanced marketing optimization technology. By providing the complimentary MOP to the market, we aim to overcome the limitations of the existing operation method, meet the needs of the market, and contribute to creating a high-quality digital marketing industry.

Describe the idea

We adapted machine learning and mathematical optimization, which utilized and approved in various industries such as spaceship, airplane, logistics and etc, for cost-efficiency and productivity, to digital marketing successfully, and the name of the solution is Marketing Optimization Platform (MOP)

MOP mainly performs a bid optimization for search advertising and a budget optimization for display advertising to maximize performance by discovering areas that humans have not yet discovered. In other words, it derives an appropriate budget and bid price in real time that are difficult for humans to judge intuitively, and automatically manages bids and budget by reflecting the market conditions to secure the best performance. It maximizes usability by making it available for major media such as Google, Meta, Naver, and Kakao. Also, MOP could replace repetitive tasks by more than 50% so that marketers can focus on planning and analysis without doing simple repeated jobs.

What were the key dates in the development process?

9 /2019 Keyword bid optimization R&D

11/2019 Keyword bid optimization algorithm derived

6 /2020 Pilot system built

12/2020 Pilot system trial application

9 /2021 Multi-channel atribution model derived

10/2021 Multi-channel budget allocation optimization algorithm derived

5 /2022 ML prediction algorithm for Keyword bid optimization derived

6 /2022 Alpha system built

8 /2022 Alpha system trial application

10/2022 ML prediction algorithm for budget allocation optimization derived

11/2022 Beta system built

12/2022 Closed beta service start (https://mop.co.kr)

12/2022 The paper published on the Journal of the Korean Institute of Industrial Engineers (https://doi.org/10.7232/JKIIE.2022.48.6.584)

Describe the innovation / technology

MOP mazimizes advertising performance with 1)machin learning algorithm that analyzes data and predicts advertising performances based on budget and bid price, 2)mathmatical optimization algorithm that establishes a bid price/budget plan to maximize advertising performance, and 3)an anomaly detection algorithm that monitors and responds to competitions in real-time.

[Optimization Process]

1) Machine learning algorithm to predict advertising performance

Using advertising operation data, MOP learns industry nature and competition status, creates dozens to hundreds of custom machine learning models for each advertising, upgrades models automatically according to the current situation, and improves the accuracy of advertising performance prediction.

2) Mathmatical optimization algorithm for bid/budget planning

Using advanced mathematical optimization technology, MOP calculates enormous cases of possibilities for every time and every creative(or keywords), so that automatically derive optimal budget and bidding plans by hour/minute for best performance. MOP has a unique advanced technology that achieves the desired performance simultaneously with its multi-goal optimization algorithm and it operates 24 hours a day.

3) Anomaly Detection Algorithm

MOP monitors the current status of advertising, detects the abnormal patterns such as competitors, peak season, under/over budget, and automatically responds to achieve the best performance according to previously established advertising plan.

Describe the expectations / outcome

MOP is validated by operating real-world instances. The results show that the platform improves advertising performance by average 10% and reduce repetitive operation by 50% than before.

1) LG electronics aims to increase brand awareness by leading customers to the website and exposing information on the search result. Therefore, achieving target of both impressions and clicks while securing cost effectiveness is an important goal. LG electronics set up an campaign for each product and applied MOP with the goal of simultaneously maximizing both impressions and clicks. As a result, the number of impressions and clicks increased by 15.3% and 11.5% respectively, and CPC decreased by 12.9%.

2) Siwonschool aimed to achieve cost-efficiency reflecting increased competitors of online education market during the Covid-19. Siwonschool applies MOP, and as a result, CPA decreased by 30.3% and ROAS increased by 45.3% without a large number of people monitoring and managing advertising thoroughly before.

More Entries from Innovative Technology in Innovation

24 items

Grand Prix Cannes Lions
MOUTHPAD^

Early Stage Technology

MOUTHPAD^

AUGMENTAL, WUNDERMAN THOMPSON

(opens in a new tab)

More Entries from LG CNS

6 items

CHOOSE COCA COLA @ COOLER DOOR (TRANSPARANT LCD MEDIA)

Fast Moving Consumer Goods

CHOOSE COCA COLA @ COOLER DOOR (TRANSPARANT LCD MEDIA)

COCA-COLA, LG CNS

(opens in a new tab)