Post on 11-Apr-2017
Once upon a time...
“The conclusion here is quite obvious: SEO forecasting is unreliable by nature of the data. Intelligent estimates can be produced, though doing so is immensely time consuming and a poor allocation of resource.”
Sam Crocker, 2010
Then we started doing a lot of this...
Search volume / Projected rank / Average CTR= potential traffic / CR = potential revenue
78% forecast their activity
75% say their forecasts are wrong
It doesn’t have to be this
But first, some considerations...
The different forecasting psychologiesOptimist
It's a big opportunity, let's smash it
Low ballerforecast low and
exceed expectation
Statistician Best statistical method possible
Sales & CSAnalystsSEOsMe
The reason Worf & Guinan don’t get along is because actually, they’re doing different things
Forecasting vs Opportunity analysis- not the same data!
Opportunity Analysis Identifying the size of the market opportunity. What can happen.
Forecasting Estimating or predicting what will
happen
Is it so bad to be wrong?!
When good is good enough
Historical data is essentialhttps://www.distilled.net/forecaster/
As a baseline
If you have outliers in your data you can also use this...
http://kirstyhulse.com/forecasting-for-seo/
We need to use a search volume/CTR model to add specific SEO activity/page level impact to the baseline
I googled ‘ctr studies’ and in the top 10 results a first place position captures about...
31.24%
17.16%
18.20%40.20%
21.24%
35.12%
Big brand
Big brand Who?!
Brand size
Search Engine
DesktopMobile
Search Volume
Work out your own CTR
Calculate CTR with GWT
https://www.branded3.com/blog/calculate-true-ctr-wmt/
Test with SERP Turkey
http://www.tomanthony.co.uk/tools/serp-turkey/create_new.php
Take your baseline and add in proposed traffic increases from SEO activity
Forecasting is NOT the first thing you do
● Based on investment
● Based on proposed activity
● Based on sign off● Based on
implementation
Forecasting adding new pages
Forecasting link campaigns
Down with link targets!
Page level forecasts against proposed activity = tangible page level reporting against proposed activity = proving value = JOY
Thank you!