Each Peer parses those keywords that are connected to the search engine in its region. The rest of the Peers in its regi
Whitepaper
Search Engine Reputation Management Tools Based on Machine Learning and Big Data
Table of Contents 1. How it all started
1-3
8. PRE-ICO conditions
25-27
2. Market description
4-6
9. Road map
28
3. Product application
7-8
10. Budget allocation
29
4. How to use blockchain for this project?
9-13
11. Our team
30
5. Growth possibilities
14
6. Product description
15-23
7. Product technical features
24
1
1. How It All Started Due to constantly growing demand for reputation management services in search engines, our task was to organize e!cient department, based on the system analytic of existing situation and results. When we started working with an important client (Fortune 500 company), we realized that the information we had provided is not su!cient and we needed to present it as concisely and clearly as possible. In this way the idea of SERMLY service was born. After that, we actively began to develop the product and realized how indispensable it was for almost everyone who runs his own online business and cares about its reputation. Our product was conceived as an individual solution for one of our clients in order to improve the workflow and processes, reports and the general situation understanding.
Necessity
Idea
SERMLY
2
1. How It All Started
Parsers
How did We Manage to Make a Closed Beta Project so Quickly? - we have vast experience in creating parsers for search engines (we created project Rankinity, 20k users right now) - we have vast experience in creating our projects for internal tasks - we have well-coordinated team that operates in Hackaton mode
Creating projects
That’s why we had a number of draft projects and, most important, a clear understanding of how this should look and work. As a result, we created a closed beta version, which is being actively tested by a number of our clients, meanwhile we are finalizing their features and properties on request.
Team
3
1. How It All Started If We Have Already Created a Product, Why do We Need Money? Everything is Very Simple: - in order to accelerate the process of completing algorithms, which automatically determine the negative and the positive, we need more developers and testers - for our maximum growth and to reach our operational payback, we need a marketing budget All processes will be developed by our own team, as close to the product as possible, and not by means of the outsource. Since we have enough information about every process and know how to use and optimize them e!ectively. Main Directions Inside the Product 4:
Product
Marketing
In our road map, we show in detail all the stages of this product launch.
Sales
Support
4
2. Market Description
The SERM market is quite large. But it is impossible to evaluate it with due
SERM Market
precision, because in this area everything occurs strictly under the NDA. The SEO market is estimated at about $ 65 billion per year, SERM holds about 15% of this turnover. That is, approximately, $ 10 billion per year
$ 10B per Year
5
2. Market Description
The Market is Distributed in This Way:
11% Other 42% ORM Agencies
- ORM agencies (full cycle of online reputation management)
24% Specialized SERM agencies
- In-house departments - Private specialists - Specialized SERM agencies - Other Such services are not advertised very much and this is the particularity of the service.
Search Engine Reputation Management requires silence
8% Private specialists
15% In-house departments
2. Market Description
Main Clients are: - Medium Enterprise and Corporate business - Politicians - Artists - Famous personalities - Popular products / services / projects - Investors - Banks / insurance companies The bigger the company is, the more image related risks are involved in search, they should be monitored and e!ectively eliminated.
6
3. Product Application Practice
Investor The investors want to invest in a number of projects in other regions. They look for interesting and promising projects, but sometimes possible deals fall through because of a shady past. The project owner would google any name and surname and finds out not the most favorable information about the subject, which definitely influences the opinion and the deal itself
7
3. Product Application Practice Company and Sales Conversion The project may become popular very quickly, and jumps far ahead of the main competitors. To interfere with this, the competitors can make a number of indecent actions, which impact the reputation, giving a bad name to the company. For example, if you enter the name of a company with a prefix reviews or scam in Google, you get a number of results, where the project is presented in an absolutely negative way. As a result, the conversion of visitors into buyers falls dramatically. According to our statistics, sometimes it can be of 50%. The buyer in most cases googles about company or product before making a purchase. If there is nothing suspicious in the first 5 or 10 results, the conversion is good. If there is, it can seriously damage any business and sales.
8
4. How to Use the Blockchain for This Project? There are 2 Main Problems: 1. Search Engines Parsing For search engines parsing (for instance Google) you need a proxy and an anti captcha. Search engines are constantly struggling against this, as parsing creates an additional load on their servers. In general, there are no problems with a proxy. But sometimes it is necessary to change the scheme for obtaining a proxy. For Example: - you can use third-party services (these are often so called bad proxies, as upon the first request from the search, they disable your access), and besides it's not cheap - you can create your own mini-servers (instances) to enable the process We have been through both experiences: the first and the second. Successful and applicable. Nevertheless and considering the following:
9
4. How to Use the Blockchain for This Project? - that many miners have appeared recently - the fact that they do not use the entire ferm source e!ectively - they can help to parse the output - they are earning more with this There are a lot of miners and they are located in di!erent regions, so by means of this source you can: - parse the output as accurately as possible - in unlimited quantities - parse it cheaper than by the above mentioned methods - parse without a break We are designing a special system that can be connected by miners to mine our SERM coin. Secure data storage Due to the large amount of data, we can distribute it to servers. The number of parameters will show us whether it is bad, good or neutral data. That will allow us to learn our system and algorithms much more faster and e"ciently. User data will not be used with the blockchain technology, as it is extremely confidential.
10
11
WHAT? But where is the blockchain?
4. How to Use the Blockchain for This Project? HASH 0000034FA35 MINER
BLOCK
MINER
NOUNCE UA
USA
MINER
MINER
MINER MINER
S D R
O
YW
KE
Ledger Keyword 1 USER
USER
Keyword 2 Keyword 3 Keyword 4
12
4. How to Use the Blockchain for This Project? Here is How It Works: Each Peer parses those keywords that are connected to the search engine in its region. The rest of the Peers in its region do the same. Thus, they hammer the block with keywords, subunit block, find hash, and give the remaining members of the network a valid block. But a vital moment here is that the peers could get these keywords, for instance: 1. Peers from the US take away keywords that you need to check in "Google.com". 2. They perform search queries, thus forming a semblance of "transactions". 3. Blocks are formed from the issuance to make sure that each peer of the Earth has the same base (blockhouse), peers from the US (for example), subunit this block, when there is a suitable hash in the block, they send it to the remaining participants of the system, who is the first to get a hash will get a reward. 4. The rest of the system members check that the block is successfully completed and add it to their chain.
13
14
5. Growth Possibilities Our goal is to get 15% of Global SERM Market in 3-5 years
85% Other Market
15% SERMLY
15
6. Product Description There are many tools that monitor mentions in social networks. But there doesn’t exist any service that is based on search reputation analysis Let's go on viewing the detailed video about the product itself
SERMLY - 3 Main Stages of Development:
Algorithms of automatic detection of negative and positive
Stage 1
Automatic system
Automatic system
of reputation
of trends in search
management
reputation
Stage 2
Stage 3
16
6. Product Description How it works now:
1
Create project
How often do you want to check?
Where will you check?
What keywords will you check?
17
6. Product Description How it works now:
2
Select regions
How often do you want to check?
Where will you ch
What keywords will
18
6. Product Description How it works now:
3
Add keywords
How often do you want to check?
Where will you check?
What keywords will you check?
6. Product Description How it works now:
4
Scan TOP 100 search engine results
19
6. Product Description How it works now:
5
Determine the level of negative references
20
21
6. Product Description
How Will It Work in the Future? 4 main functions: - Analytics - Reputation Management
Analytics Reputation Management SERM Trends
- SERM Trends - Blockchain Parsing System
Blockchain Parsing System
All in all, everything is grounded on the work
Machine Learning and Big Data
on search results for the queries you need in order to obtain positive search results.
22
6. Product Description
Analyze
How Will It Work in the Future? A) Analytics - Create a project - Specify desired regions
Result
- Enter keywords - Service analyzes TOP 100 results for each keyword - Automatically detects positive / negative / neutral - Visualizes your current position - Automatically gives recommendations on how to improve it Recommendations
23
6. Product Description
Bad Reputation
How Will It Work in the Future? B) Automatic Reputation Management - Service analyzes TOP 100 results - Detects negative / positive / neutral - By internal parameters determines the strategy for promoting positive
SERMLY
results above - It works to ensure that positive results promptly squeeze out negative - Automatically supports the required results in the search It's like SEO, but here you are only dealing with branded queries and at least the first 20 search results, that is much more complicated.
Good Reputation
24
7. Product Technical Features Technically, our product looks like this:
SPA
Front End Vue.js
Single Page Application (SPA)
AJAX
PHP, Laravel Framework
MySQL
Redis
LSTM
Word 2 Vec
Other
LSTM
MySQL
Other options are also possible
Back End
Machine Learning
(stands for long-short term memory)
8. PRE-ICO Conditions for Token Buyers
The sale of tokens at PRE-ICO begins on December 21 Join us in our massive presale and get 25% bonus for SERMLY tokens.
25
8. PRE-ICO Conditions for Token Buyers
Token Sale Name of the token: SERM - Utility Token What represents: inside currency for service payments - utility token Token price 0.5 $
0.5 $
Discount for PRE ICO
25%
Hard Cap for PRE ICO
$ 500,000
Period PRE ICO
December 21 - December 31, 2017
!
Important notice: If we collect the maximum cap, we’ll close funds collection.
26
8. PRE-ICO Conditions for Token Buyers 1. Limitation on the Amount Hard cap for PRE ICO: $ 500,000 When we raise $500K, we close the PRE ICO and begin active preparations for the ICO.
2. Mandatory Condition to Join the Telegram Chat To participate in ICO you have to join the o!cial SERMLY chat in Telegram: RU - https://t.me/joinchat/BgSzlxFVHLyt8UDkgxthew EN - https://t.me/joinchat/BgSzl03whdVMuifkwq_qew
3. Token Freeze Tokens that have been purchased at PRE ICO will be frozen for the first month after the ICO.
27
28
9. Road Map 1st Quarter of 2018 release SERMLY ICO finalize algorithms and start beta in public access develop an automated reputation management system set up marketing through content and expert media build a marketing department that will advance the product
2nd Quarter of 2018 add Yandex search Russian localization
develop basic processes before localization add YouTube search add Bing search
French localization German localization Italian localization Portuguese localization Spanish localization project algorithms finalization capacity increasing automatic system testing
3rd Quarter of 2018 add Baidu search Chinese localization Japanese localization pour out an automatic reputation management system into beta project algorithms finalization
4th Quarter of 2018
content project launch on reputation management
make project’s White-Label give API project algorithms finalization automatic reputation management system complete launch add Naver search Korean localization
1st Quarter of 2019 search reputation trend tool development project algorithms finalization parsing and continuity systems’ improvement through blockchain and miners search reputation trend tool launch
29
10. Budget allocation Budget distribution: 50% marketing 25% product 20% infrastructure
50% marketing
25% product
5% administrative expenses Key goals: - current team strengthening - accelerated work on the product - active product marketing - product active sales and its localization to key markets Token distribution: 1,5% bounty 10% PRE-ICO 88,5% ICO
20% infrastructure
5% administrative expenses
30
11. Team
Victor Karpenko
Yakov Kluev
Yuri Pokrovsky
Dina Shurkhina
CEO
DEV Team Lead
UX Team Lead
Front end dev
- 12 years of SEO experience
- 8+ product development
-
10+ years of experience as a designer
-
Front end dev
- 10 years of Western SEO experience
- Parsing and monitoring systems
-
5+ years of interfaces creating experience
-
1.5+ site layout experience
- 10 years of active work in SERM direction
- Designing and implementing
-
Working with large amounts of data
custom analysis systems
-
Designing systems for the largest
- Head of SeoProfy - Projects creator: Rankinity, Link Checker
adaptive and responsive sites creation -
BEM methodology using practice in projects
auditing company in the world
- NaZapad conference organizer - SEO experience in the most competitive niches
Mariya Kanifolska
Oleg Moroz
Dmytriy Voznyi
Pavel Lyashenko
Project Manager
Senior UX Architect
SERM Expert
Research Data Scientist
- 10+years of local and international
- 5+ years of interface design for web and mobile
-
12 years of SEO experience
- Research Data Scientist
-
11 years of Western SEO experience
- Collection, parsing, processing, analysis of
- Designing systems for the largest
-
8 years of active work in SERM direction
auditing company in the world
-
SEO experience in the most competitive niches
projects management - Working with large amounts of data - Projects’ internationalization
Working with large amounts of data
large amounts of data - Algorithms creation and implementation - Information searching - Research and experiments on data search on public resources
31
Join SERMLY
Join SERMLY and become a part of global product based on machine learning, blockchain and big data.
tokensale.sermly.com