Can Google Data Be Used To Predict Stock Values

Est. Reading Time: 8 minutes

Can Google’s data be used to predict stock values?

People have been asking for some time now if Google can predict the stock market. I recall seeing in an interview somewhere that Google said that it most likely could but wouldn’t feel comfortable with putting that into the hands of the public.   Kidding aside, there have been some studies out there that show that Google trending correlates to stocks changing hands but doesn’t exactly provide insight into the value of the stock.  Unfortunately, that’s a big chunk of the equation.

I wanted to investigate this a little further and see if I could make a case for using a series of keywords that replicate patterns or stages in the buying cycle of industry leaders.  I’ve used these same tactics for search marketing to help clients understand shifts in behavior in their industry.  Now my goal is to see if I can use that data to understand shifts and swings in the consumer behavior and ultimately market share which can be used to create understand the future directional value of the company’s stock.

I wanted to start with big brands so the data sets would be large enough to show swings when I dug deeper into keywords that more represented brand demand and favor.  I started with the automotive industry and took a look at Ford, Chevy, Honda, Toyota, and Kia.

My first attempt at segmenting product demand.

I first started my search for various types of vehicles.  I tried the to select an equivalent vehicle from all the various brands I was analyzing.   My first search gave some surprising results.  I didn’t expect for the Ford Focus to dwarf the others like this.  I thought the Kia Optima would have stronger search demand so that was a little surprising as well.

 My second, more refined product demand Segment.

I refined the search queries and added the Toyota camry, Honda accord, and the Audi A4.   That provided a little better result with closer search volume.  The Ford focus still seemed to see the largest increase.  Honda accord actually declined and the Camry stayed about the same.  The newcomer to the party was the Ford Fusion with what looked like the largest gains.

What was driving this shift in product demand within the auto industry?

It was fairly clear that Ford was grabbing market share for queries that represented product demand.  But why was this? Then I remembered a study that was done back in 2008.   The study investigated people’s likes and dislikes towards a vehicle’s “face”.   What the study found was that people generally liked cars that had meaner, angrier, more aggressive looking “faces”.  By looking at the design changes from 2008, it’s apparent Ford took notice.  You can see the newer model has downward slanted headlights, bigger cooling vents, and a stronger “jaw” that gives its face a little more bite.  That change has also caught the attention of the car buying public.

2008 Ford Fusion                                               2013 Ford Fusion







Does this shift in searches translate to purchases?

So my next question was how does this shift in product demand based on keyword search data impact a buyer’s inclination to actually purchase this vehicle?   I decided to run a comparison of keywords that would indicate a prospect’s likelihood to go to the dealership.  That can be represented by the Brand’s name tied to a “dealership” query.  This particular trend seemed to provide the most insight into Ford’s takeover of market share.     We can at the beginning of 2008, both Honda and Toyota both seemed to experience a decline in people looking for their dealerships while Ford seemed to take that market share.    Their was also a bump in both Kia and Chevy but the volume is large enough to do any damage.

At this point, it seemed that Ford was gaining the most and taking market share from both Toyota and Honda.   But I always recall that Honda’s and Toyota’s generally have a very favorable brand perception when it came to reliability and Ford, whether deserving or not, seemed to have that domestic auto quality problem.   People just tend to think that a Honda is going to last longer than a Ford.  Surely this would have some impact on people’s perceptions and buying behaviors?

Predicting shifts in quality across manufacturers:

My next concern was identifying what keywords people might use that could represent the quality of their purchase.   Owning an Audi, I know for one, when the check engine light comes on it most likely means a sensor is bad, or some part is broken, or some major repair is about to hit me.   So I thought I would check the query data for these brands in combination with people searching for “check engine light on”.   This report produced some very interesting insights.    In 2005, Ford was averaging slightly higher than both Honda and Toyota.   From 2005-2008, most all the manufacturers were trending downwards and then started to increase around 2009.

What I found most interesting is that both Honda and Toyota’s growth of search demand for people with check engine lights on outpaced Ford and Ford is now behind Honda and Toyota.  Considering this most likely represents an issue with your car, this indicates that Fords quality has either gone up or Honda and Toyota have gone down.   Either way, it’s good news for Ford.

A big concern would be for Chevy buyers.  The search demand for people researching check engine lights on skyrockets from 2009 to current.   This would give me concern if I was in the market for a Chevrolet.     Brand perceptions for quality and reliability for a product that might own for 5-10 years takes time to develop.   However, these signals could be some of the first indicators that drive such a shift in perception.

Combine all of these findings together and it looks really good for Ford.   People like their products more, they are searching for Ford dealerships more frequently and at much higher growth rates, and now quality related searches point in the direction that Ford is building a better longer lasting product.

How does Ford’s Financial Look?

When looking at longer term trending of the companies worth by evaluating their stock price, we see a longer, more gradual decline than what you typically see when looking at a chart that has the 2008 recession.   The uptick was fairly large after that, at first.   Then it declined for 2 years, which oddly were the highest demand years Ford has seen.  There there is an uptick at the end of 2012 followed by a 6% loss yesterday on earnings reports.


I’m not a true stock analyst and am trying to use Google’s data and my marketing savvy to come up with a unique method for evaluating and selecting a good stock.   So what this looks like to me is that the company’s stock value has been down the last 2 years when they’ve seen the largest growth in the US market.  Their latest dividend was the highest it’s been since 2006 with .74%.

I’m not really a gambling man but that makes Ford look undervalued to me and poised to see a good period of strong growth.  I am going to buy $100 worth of Ford stock today and will update this post with the value of my search data driven stock pick on a quarterly basis.  This will provide actual ROI this data was able to produce.   My guess is I will see a significant gain in less than a year as all this pent up demand for new focuses and fusions is finally released when people get their taxes back.    Over the next 2 years, I think their stock will double in value.

I hope you found this article informational and entertaining.  Google provides a wealth of data and there are plenty of ways we can use it in our daily lives.  This is just one example that has interested me lately.



One Trackback