Another way to evaluate the effectiveness of influencers and content creators ​

Hi, I'm Ivan, I've been working in Influencer Marketing for 6 years in different roles. I designed and implemented an influencer marketplace architecture and built processes for an agency. I learned how to run 3,000 placements a month and communicate with hundreds of content creators a day.

I have come across that there is a lot of "magic" in Influencer Marketing that my engineering mindset doesn't allow me to understand. Let's try to structure this "magic".

Advertising with influencers brings good results, but in this text I will not consider when this advertising should be used and when not. We will analyze the situation when you are already working with influencers or are ready to start.

How influencers and content creators are usually chosen ​

You can find tons of articles on the topic of Influencer Marketing on the internet, where you will learn what you need to pay attention to when choosing an influencer or content creator:

Select the right content niche. Of course, the creator's content should match the interests of your audience.

Check the price. You should not choose the most expensive influencers, and you can leave the celebrities for later.

Check the price per view. You can compare influencers with each other, obviously the cheaper the view, the cheaper the lead. A cheap lead is what we need.

Check influencers for fake subscribers. Of course, we only want real views, so we'll only hire influencers without fake followers.

Make sure the influencer's audience is engaged, i.e. the account has a high ER. More engaged subscribers means more sales.

Match the creator's audience with your target audience. Naturally, if your customers are women, then the audience should be female. The influencer's audience should match your target audience.

The usual algorithm of actions ​

Let's skip the crucial stage of setting the goals of an advertising campaign and drawing up a client portrait. We assume that you have already done all this and fully know what you want to get as a result. And, of course, you should not forget about this stage.

When you have determined the criteria and goals of the advertising campaign, let's try to make an algorithm of actions:

1. Choose an influencer who covers the topics we are interested in;

2. See that the price is not very high. Let's say we decide that one placement should not cost more than €300. We want to make several placements. So we will only take a creator if she costs less than that.

3. Let's calculate CPV. Let's look at the last post from the influencer, or preferably several, divide the price by the number of views and realize how much a view will cost us, how it compares to market prices;

4. Let's check for fake subscribers in some service. We won't work with the influencer if the percentage of fake subscribers is high.

5. Let's calculate ER and take into work accounts only with good, let's say, with ER above 3.5%;

6. Find out the statistics from the influencer or use any influencer analysis service. We will work with a creator only if she has more than 50% of the audience of the gender we need.

Looks logical, but there seem to be a few problems. ​

1. A lot of work. Find an influencer, check ER, write to them and find out the prices, calculate CPV, check the gender of the audience, check that the account has a real audience.

   Each creator has to be analyzed individually, although in fact half of the parameters could well be calculated automatically.

2. We take into account all the parameters separately, but in fact — they work together.

3. There is no way to compare two influencers. An influencer is either suitable or not.

Example ​

Let's take, for example, a dating service we want to promote with content creators. Let's say we need a female audience aged 18–35.
Interests... What are girls who are looking for new acquaintances interested in? We need to experiment. The goal is to increase the recognition of the service and get registrations of new users.

Let's pretend you've found the following creators to advertise with:

empty	Creator A.	Creator B.
Price	€300	€200
Average Views	30,000	36,000
ER	4%	2%
Women	51%	80%
18-35	30%	60%
Fake subscribers	15%	40%

Which creator should you order advertising from? The first one seems to fit, but the second one not so much. The second one has a lot of fake subscribers and low ER, most likely they will not be recommended for placement by influencer analysis services.

But is there another view of how things really are?

How to choose differently, how to take everything into account? ​

Target Audience (TA) ​

Our task is to choose the influencer who will have the highest percentage of our target audience and at the same time the lowest price.

Let's try to calculate the percentage of the target audience. In our case, it can be calculated as follows:
%TA (percentage of target audience) = percentage of women * percentage of audience aged 18-35 * number of "live" subscribers.

%TA(Creator A) = 0.51 * 0.3 * (1 - 0.15) = 13%.
%TA(Creator B) = 0.8 * 0.6 * (1 - 0.4) = 28.8%.

It seems that now Creator B is not so bad, they have more than twice the share of our target audience than Creator A.

And how much is that in absolute numbers?
TA = %TA * number of views;

TA(Creator A) = 0.13 * 30,000 = 3,901.
TA(Creator B) = 0.288 * 36,000 = 10,368.

The gap is crushing, up to this point, one would have thought that Creator B is not worth choosing, but it turns out that they are 3 times better than Creator A.

Cost per view of your target audience (CPV(TA)) ​

The result is not bad, but we didn't take into account the cost of advertising. I suggest using the price per view per representative of your target audience. You need to keep that in mind as well. How to calculate it?

CPV(TA) = ad price / TA

CPV(TA) (Creator A) = €300 / 3,901 = €0.076
CPV(TA) (Creator B) = €200 / 10,368 = €0.019

The calculations clearly show that now the difference is 4 times, Creator B has become even more effective.

Active target audience ​

We did not want to take Creator B also because they have a low ER. This indicator was not taken into account at all. So what should we do in this case? In fact, at least partially, we have already taken it into account when we counted only real subscribers.

But let's make more precise calculations with the ER index. We'll get a slightly synthetic index, which we don't need for anything except for comparing influencers with each other, but it helps us to take everything into account at once and compare influencers with each other more objectively.

Let's find out how many people from our target audience will rate a post or write a comment.

ATA (active target audience) = TA * ER

ATA (Creator A) = 3,901 * 0.04 = 156
ATA (Creator B) = 10,368 * 0.02 = 207

Creator B wins, but the gap has narrowed.

Cost per view of your active target audience (CPV(ATA)) ​

Let's multiply everything already and find out the general indicator that will help to choose the best influencer.

CPV(ATA) = CPV(TA) / ER = ad price / ATA

CPV(ATA) (Creator A) = €300 / 156 = €1.92
CPV(ATA) (Creator B) = €200 / 207 = €0.96

It turns out that Creator B is much more efficient. Of the two given options, it is better to choose them, despite the fake subscribers that may have been at the beginning of the creator's career and the consequent low ER.

Can it be done better? ​

In the end, we settled on choosing Creator B, which has a target audience share (%TA) = 28.8%

This is an indicator that can often be calculated without laborious communication with the influencer. But if it's a good result, how can you tell?

The smaller your target audience is, the smaller share you can expect from the influencer, obviously. Which means you can only understand if it's a good result by comparing %TA from different creators. You just need to choose the best one according to this indicator.

We constantly update influencer statistics with our analysis tools and social media data.
To select the best influencers, we sort all of them by the share of the desired target audience and start working with the best ones.

Thus, the advantage of using specialized tools to compare the %TA parameter is that you can quickly find influencers who will be more effective — in our case, a creator with the same target-audience requirements would be nowhere near the top 100.

Why are there so many formulas? ​

It might be worth limiting ourselves to one metric and evaluating everything against it. For example CPV(ATA), but, unfortunately, we don't always have all the data.

At the first stage, %TA is a good way to cut off those influencers who are not worth cooperating with. And proceed to the labor-intensive stage of communication only with those creators who fit the external analytical data.

Conclusion ​

I hope I was able to clearly show that an influencer with fake subscribers is not always a bad thing. It is not always effective to throw out subjectively “bad” influencers without calculations. An analytical approach can greatly reduce labor costs and improve ROMI.

A good influencer is the one who best suits your target audience. Their indicators are relative qualities; you cannot evaluate a creator without taking into account the tasks that you set for them and without comparison with other influencers.

Selecting an influencer or content creator is only the first stage, but using this approach allows you to save a lot of money and time.

Thank you for attention.

2TA.IO - here is a beta version of the service that will allow you to try our approach. It has a free trial.

Hire Creators with EPICSTARS - here you can order an effective advertising campaign with influencers and content creators.

2TA Facebook group - Join our Facebook group with project updates and news.

Telegram by Ivan Kobzev - my Telegram, where you can always write.

Telegram channel 2TA - here we publish service news and useful articles about influencer marketing.