Why are statistical sports models similar to marketing models?
Our team is not worse in developing the odds for sports events than the bookmakers specialists. Therefore, we can confidently say that this is a very related concepts. When you develop a mathematical model for sports, then you must form the parameters that affect the outcome of the match. Different options doesn’t have the same effect at the end result. After all, you have to substitute your ratings of the strength of the teams that you defined from the parameters, substitute into the formula for the probability distribution or other algorithms. For football, the Poisson distribution is well suited. At the output, you get the probability that you will need to test at historical data.
As you have already understood, the marketing model at the output will have such results depending on the parameters. For the basic marketing model, we have defined the following parameters that are close to our situation:
- Photo content
- Budget for facebook targeting
- Selected ads channels
- CTR site or account
- Selected algorithm for calculating expected sales
The first stage of the model is the creation of content. It is important what composition be on your photo. Pictures with blue, white colors and beautiful people will get 70% more interaction than other photos. Therefore, carefully plan your content, for this you need to create a marketing strategy.
The budget for advertising may be different. The main thing is to understand how to use the budget effectively. For this example, we took facebook targeting. But advertising is not just targeting. There are plenty of channels where you can effectively distribute your content. With facebook you can choose the audience. Each change in the target audience attributes will affect the end result.
The same affect the results on which channels to advertise. You can do it automatically, but you can manually: instagram, facebook, facebook messenger and so on. The more we share the parameters, the better we will be able to understand our audience.
When you select the characteristics of the target audience and where to place the advertisement, then Facebook predicts the approximate reach and the number of clicks. For example: with a budget of $ 10 per day and your target audience, expected clicks = 100 clicks
Just those 100 clicks will exactly your CTR. In this case, the CTR is a referral to the site. If the coverage was 20,000 and 100 clicks, then CTR = 0.5%. Your niche affects to the CTR. In most cases, it is above of our value.
The best sites in the world have a CTR of about 30%. In the beginning you will not have such indicators. Knowing the average figures for a niche, we can calculate the expected number of buyers and the average profit from the product. For example: CTR in niche clothing is 12%, considering our budget for ads and content, the CTR facebook will get the following – 100 clicks x 0.2 = 20 clients. 20 clients are approximate, the distribution of customers is quite wide.
Poisson distribution in the digital marketing model
We have approximated the value of 20 clients. To find out the probabilities for 17, 18, 19, 20, 21 … buyers, we substitute the average value (20) at the Poisson distribution. The highest result was obtained in 18, 19, 20 clients, at 8.88% for each one. To know all other values, go to this site and paste the value into the top of the page.
This model is just an example. These models can be unlimited. But it can serve as a good starting point for you.
Knowing the approximate amount of product and sales revenue you can calculate ROMI.