To solve one client’s task, the employees of an advertising agency have to examine thousands of influencers, and assess a mountain of parameters. An average agency employee might need several days to compile a list of blogs for an entry-level advertising campaign.

This routine work can be described as follows: firstly, you manually find suitable blogs, then evaluate content and look at the figures (engagement, number of views, advertising CTR, etc.) in the influencers’ database. There are plenty of those databases on the market, but you have to know what to ask about in each of them. Then the received information needs to be gathered in a media plan and estimated costs, coverage, CPA/CPC should be calculated and sent to the customer for approval, if we are talking about an agency.

It is the development of media plans that we have learned how to automate. An agency sends a customer a link to our terms of reference template (brief), where in a few clicks you can select the desired subject of the blog, as well as geography and socio-demographic characteristics of its audience.

Then the magic happens and the client receives a prepared media plan with a list of bloggers and all the most important campaign figures. We additionally check the automatic release manually, so it takes approximately three hours from the brief to the finished media plan.

We learned that the market really needed ready-made media plans when we were testing other products. We started with the idea of contextual analytics of Youtube videos. After a few iterations it turned into an influencers’ database. Later our business analyst obtained a position in a partner’s advertising agency. A close look at the processes has shown that a mere influencers’ base is not enough, as well as developing a media plan is a complex and time-consuming procedure. What’s more, about 80% of leads go cold having received their plan. And the employees who create these media plans burn out fast and leave in a few months.

Feedback from the market suggests that media planning and marketing software is really needed. When we talk about it, we are always listening attentively. We haven’t launched yet, but have already signed annual contracts with major agencies. We’re the ones who have a comprehensive solution in the niches of influencer marketing — competitors offer simply bloggers’ databases.

...

Monetization and markets

We have provided several monetization schemes for different client groups. There will be a free version where you can get a media plan and select bloggers on your own. At the same time, we make the influencers’ search with all the data about them totally free, also we make forecasts about the cost and the results a company will have if it collaborates with the bloggers chosen.

All filters will be available for a fixed price. This scheme is suitable for customers who want to place ads with influencers by themselves.

For advertising agencies, we can offer a monthly subscription package. The agency will be able to brand the form of its brief and plan. We are also considering the option to add a form of payment and work with a partner for a share of this revenue.

Today, we analyze only English-speaking YouTube bloggers. Thus, we focus on the USA, Canada, Britain, Australia and New Zealand. In the coming months, we will add Europe and the CIS.

...

Let’s look under the hood

Our project is based mainly on the data collection system. We have designed a group of robots, each of which does its part of work and gives its result further along the chain. Consequently, we have detailed analytics for each of the monitored channels, and it is constantly being updated.

We monitor the dynamics of views, subscribers, likes/comments, clicks on advertising links, and a number of other options. On the basis of collected data, we consider efficiency and engagement metrics, predict coverage and advise on a fair cost of the ad placement for a particular influencer. In total, we have more than 20 metrics.

We know what the content of the channel is about as we collect the semantics of speech in videos, clips’ titles and descriptions. We revisit the video we monitor about 6 times on average to track the dynamics of views and get clearer and more transparent data.

We also know how often you click on the advertising links from the description of the video. We have learnt how to recognize bit.ly links, and from them it is possible to find out statistics of clicks. In addition, we have our own link shortener for influencer marketing, through which we also collect data.

We count part of the metrics on our own and regularly update the results. That is, the number of views, involvement etc. is different according to the day of the video placement: on the first day there will be one figure, on the third day — another one, on the seventh, fourteenth and thirtieth days — it will also change. So, our data cleansing tools work without stopping.

Our main technical task at the start is to create a data collection system that would spend a minimum of resources and would be easily scalable. Our team of four engineers and a technical director spent four months to find a solution. Now everything is working properly and doesn’t need attention.

Platforms usually don’t share data, so we get part of the information from the API and some info we have to parse. It’s legal, because the data is public, however we had to rack our brains trying to find a way to collect data without attracting platforms’ attention.

Now we have a cloud computing infrastructure that enables to handle millions of channels across multiple sites. Today, such work costs us about $3–5 thousand a month, and that’s not much for this scale of tasks.

...

The team

There is a team of 6 people at our office in Minsk who are currently working full time: 4 engineers, a business analyst and I as a product manager and CEO. Besides, we are assisted by a technical director from Silicon Valley, who is a specialist with extensive experience and with whom we have already done several projects.

Additionally, a marketer, a copywriter and a designer work on the project remotely. Now we are also looking for a smart salesman who can work with LinkedIn, as well as data specialists. We have a fertile ground for the development of AI, and we have collected datasets, which no one else has yet.

...

Prospects & Investments

The launch of Easy Bloggers platform is scheduled for mid-March. Right after that, we will provide data on English-speaking Youtube Bloggers. Within two or three months we will add bloggers from EU and CIS. By June, we are planning to involve Instagram-bloggers in these regions.

We aren’t planning in the years ahead, because the market is very dynamic and we need to constantly adjust our actions. But we have a common vision, we create a platform that automates all the work with influencers: from search to payment. Our aim is to develop a system for influencers’ booking.

Developing a media plan is only the first part of Easy Bloggers platform. Next, we will add communication with influencers about advertising placement, signing contracts, control of execution, payment and secure transactions.

Now Rocket DAO helps us raise investments at the Pre-Seed round. We need the funds to enter new markets and expand our usefulness.

We have already attracted funds from one of the Belarusian business angels as well as from the LaunchMe accelerator. In addition to money, the accelerator helped us build the team and gave us an impulse to start. Once our idea was chosen for participation, then somebody needs it. That’s how it turned out later.