In the world of marketing, ROI or return on investment is a major talking point in planning meetings. Often times, the ability to garner the right resources comes from painting an accurate picture of what will happen given the proper strategy and execution of said strategy.
Here are 5 factors to cover in order to properly forecast and consistently deliver results you expect.
#1: Historical Trends
When forecasting, history is your first friend and the more historical data you can gather, the better. When looking at historical data, look at your outliers, as in, those spikes up and down, and be sure to re-adjust in order to establish a proper base forecast. When in doubt as to whether you should account for a spike answer the following questions:
- Why did this spike occur?
- Will it happen again?
To further improve our baseline information we’ll be looking at the data and account for a few more items, the first being growth.
Growth is good. Unaccounted for growth or growth that can’t be duplicated, not as good. Look at all your growth and match it up with the following potential outputs:
- New product launches
- Ongoing SEO efforts
As you index your growth year-over-year, also be sure to break it down month-over-month and see if trends in search (specifically, in your industry) align with what you are observing in your adjusted historical data. Gather data for keywords driving organic traffic at that time and you can properly forecast for the next season.
Different than general growth and seasonal spikes are events, which may results in specific increase (or decreases) that you can account for. Examples of this may be:
- A yearly sale
- Or a big piece of published content (like an interactive)
Again, based on keyword research, determine your expected increase and add to corresponding months.
Now, if you really know your baseline and forecast accurately, you may need to get reacquainted with some not-so-basic math.
To set a base forecast you’ll need to:
- Calculate the mean and standard deviation
- And then set the rule by calculating the coefficient of variation or COV
As a general rule, if the COV is less than 1, the variance of your data is low and there is a good probability that you don’t need to adjust any data points.
If you’d like a tutorial on how to calculate the mean and standard deviation and then set a rule based on the coefficient of variation, give this video a thumbs up or leave a comment below.
That’s it, 5 factors to cover in order to properly forecast. Until next time!