Google has just announced an update to its Insight for Search tool which now includes predictions for future search volumes.
I thought I’d give it a quick test on a few relevant terms. First up, online PR.
According to Google, search volumes for “online PR” in the UK will reach a peak in September of this year. They will tail off dramatically in October, rise again a bit in November, fall off again at Christmas – and then reach another spike in March 2010 – but not as high as Sept 2009 (interesting to note that the graph has shown a recurring pattern over the last few years – what happens in September to spike interest in online PR? AdTech?)
As noted at the Google Research blog: “Having predictable trends for a search query or for a group of queries could have interesting ramifications. One could forecast the trends into the future, and use it as a “best guess” for various business decisions such as budget planning, marketing campaigns and resource allocations.”
For example, anyone in the business of selling online PR services may want to focus their efforts on September 2009 and March 2010.
There are also ramifications for PR content development. You might plan to develop specific content for the future based around predicted keyword search volumes.
Of course, as ever, this all hinges on the accuracy of the predictions. And as Google are quick to point out, they are only extrapolating from previous data. Still, better than nothing.
In Front Of Your Nose 
Andrew, good post, this is interesting. Google is predicting how much traffic is actually going to be searching for a particular term, in a particular month, but I can’t see how that can be accurate.
For instance, imagine if an influential article is published highlighting online PR as the next step forward in marketing or if an award winning online pr campaign occurs in October and is covered everywhere. This could make more people interested in it and in-turn search for it. I would be most interested to hear how Google predicts this but I doubt it will share that.
I suggest you do a follow up post in September and October and see how accurate it was.
LikeLike
Hi Chris – thanks for the comment – if you read the original Google blog post, you’ll see they aren’t making claims for 100pc accuracy – or indeed for all search terms. What they are saying is that for certain types of search term, they have enough historical data and a recurring pattern that would suggest an outcome with a small variation of error. As they say, certain categories are much more difficult to call eg social media and online communities. And of course, “black swan” events are impossible to predict – and can impact any prediction. But you can’t deny that as a way of helping to shape future action, this is better than pure guesswork.
LikeLike