Earlier it was easy to log in to any analytics platform and download every single keyword that drove traffic to the site (or keywords relevant to a section or a page). Today secure search has made it much difficult to determine the keywords you are already getting traffic from; there are plenty of large scale data sets available that can give you the same results in a fraction of the time and cost.
Large keyword indexes like those offered by SEMRush or seoClarity provide the ability to view the keywords you rank for across tens of millions of keywords, complete with information on search volume and rankings for each. However, getting a download of all the keywords you already rank for is just part of the equation and simply shows you where you stand right now. As the following diagram displays, your keyword universe consists of irrelevant keywords, keywords you are relevant for (but not ranking for) and keywords that you rank for (mostly that you are relevant for).
The next step is to find keywords that are relevant to you that you do not rank for. The wisdom of the crowds methodology I covered in a prior column is my favorite tactic to quickly discovering those hidden gems of keywords which I should be targeting but have not yet succeeded with.
Scenarios for a forecast allow you to evaluate a range of possibilities from the best-case to the worst-case. There are two ways to create your growth scenarios:
A “snapshot in time” approach is a simple model that assumes you can get all your keywords to position one and base your traffic estimates on that. In my opinion, this is overly simplistic and ignores the fact that different keywords are most likely in different starting rank positions — meaning they may take different amounts of time to get to the top positions.
A more sophisticated approach involves applying a rank estimate at the keyword level or a percentage improvement in rankings off of the base rank. You can then plot the month-over-month improvements for the number of months that you expect the benefit to accrue until you reach a reasonable point of rankings for each of your keywords.
For example, you may want to build a spreadsheet as follows:
With the hard part out of the way, on to the easier part — estimating traffic!
You’ve probably seen a ton of research studies on the click-through rates for different rank positions in organic search. Take your pick and then apply the CTR for each rank position for each month you have estimated in the prior step to calculate estimated traffic from that keyword for that month.
Sum for each month and what you have is the traffic estimate for each month
Apply your conversion rate and average order size and you have your estimated conversions for each month as your rankings improve.
At this point, you have your forecast!
But wait; that’s not the last step.
Track To Actuals
It’s easy to think that building a forecast is the end of it. However, if you want to maintain any respect and trust within the organization, there’s one more step: track your performance against the original forecast.
Share the results with the team — even if things don’t actually turn out like you expected. An honest, open approach to explaining reasons why something may not have met targets goes a long way toward building trust. This is serious:
Do not undertake forecasting if you don’t plan to follow through on comparing to actual results!