Gas Prices and Measuring Internet Sensitivity

Bill Tancer, the general manager of Hitwise and author of Click: What Millions of People Are Doing Online and Why it Matters, blogged here earlier this week about Internet trends and data. This is the second of three posts on the subject.

One of the things that I’ve learned from monitoring the Hitwise aggregate search behavior of 10 million Internet users in the U.S. is that, as a group, we have a remarkably short attention span.

Take any big newsworthy event, if gore or nudity is not at play in the story; expect the spike in Internet searches and resulting site visits on that event to last no more than one week. Rising gas prices are no exception — until recently.

If we chart the visits to sites that provide gas station comparison shopping (sites like against average U.S. gas prices (all grades), we find examples of short-lived hyper sensitivity in September 2005 when retail prices crossed the $3.00 per gallon line, and then again in May 2006 as we approached that same price. It’s interesting to note that subsequent surges past $3.00 per gallon have been met with ever declining visits to gas comparison sites.


Even though gas prices maintained their high levels for several weeks, as Internet searchers we were on to other topics beyond finding the cheapest place to fuel up.

Since this Spring however, gas comparison shopping is showing steady increases indicating that as we hover around $4.00 per gallon, we may have reached a point where getting the most fuel for our money becomes a more integral part of our lives.

To understand this trend further, I analyzed the demographics of visitors to the most visited gas comparison site,, over the last four weeks. Visitors to the site were predominantly older (49.57 percent of site visitors were 55 years old or older) with fixed or lower incomes and 47.34 percent were from households earning under $60,000 per year).

This is a dramatic departure from the demographics of that same site in June 2007. At that time, the largest income segment of visitors to the site was households earning over $150,000 per year (25.69 percent). Today the upper income segment accounts for only 9.6 percent of Gasbuddy’s site visitors.

Historically — adding validation to the adage that you have to save money to make money — households earning over $150,000 per year were the most likely to visit comparison shopping sites like, Shopzilla, and, but in the case of surging gas prices, Internet behavioral data indicates that finding the best price is becoming more of a necessity for those on lower or fixed incomes.

Mark in NY

I gave up on the sites when it became obvious that certain stations were using them to engage in e-bait-and-switch.

let them eat gollum balls

wow, popular freako thread here. "Once I’ve found the cheapest station, i don’t really need to go back."

exactly. occasional checkup to see if a new seller has been added to the lowprice end of the reports list. the spikes in interest are from ppl who are *new* to gas price sites.

re:Fuelwatch. yep. that and "private" gasprice watch sites are (potentially) efficient resources to help sellers fix prices.

semi-ot, i've noticed that regional prius prices seem to be standardized, presumably because the car has no comparables (competitors)


Seems like the search hits is related to the derivative of gas prices, not the gas prices themselves.

When gas increases, people look for cheap gas. When gas price is stable, they don't look as much.


Also in response to Black Political Analysis. Right now 71% of Americans have internet access. Pretty well verified.


The primary reason why hits to these sites have not increased recently is because new visitors when they put in "gas prices" into a search engine are directed to mapquest or google maps which integrate the data from gasbuddy. I know more people have been using online gas pricing information, but they don't use gasbuddy directly.

Thomas Brownback

While the sites may be strong in some communities, they are definitely weak near my home. "Witty" nailed the reason I visited once (after gas hit a psychological milestone) but never went back. Gas Buddy had prices for two gas stations in my area, the prices were four days old, making them completely inaccurate.

If you could get data on where the site gives strong results and where it's weak, you could apply some basic GIS to see how much the "weak results" hypothesis accounted for one time visits.

I imagine it's a combo of that, the "commit to cheaper stations" hypothesis (comment 6), and the original hypothesis that people are more sensitive to gas prices near psychological milestones. I'd be anxious to see the stats on the contribution of all three of these factors.

S. Copeland

Where did the Tancer get the demographics on members? From his Hitwise business? He did not fully explain. How accurate can the demographics be? I certainly don't give out my personal salary information. site helps me to easily stay in touch with the latest energy news, with local fuel pricing, & share consumer tips with others. I agree with the comments that multiple factors effect "hit" data: publicity of site, hits becoming routine ( site), individual needs (2 stations in area vs. multiple & also individual's income--a few cents over time becomes significant), etc. I think this article oversimplifies & does not accurately address the issues it raised.

Jon Barsanti Jr Hillsborough, NC

More people are sensitive to price upswings than downswings ... Activity Spikes as Gas Prices Spike ... I participate on and report prices normally every weekday ...

G Pendergast

Grasp a man's wallet or his testicles and his heart and mind will follow

Jason B

Just eyeballing the chart, I'd say the rate of change in gas prices versus the gas shopping site traffic might be an interesting relationship to examine.

Drew Lessard

I find this pretty interesting; however, a far more interesting question to answer with data is: How much of the increases in e-commerce can be attributed to spikes in gas prices? And, therefore, are shoppers turning their shopping carts in for "shopping carts"?

Simon H from Downunder

In the State of Western Australia we have a government mandated scheme called Fuelwatch. Under the scheme, retailers are required to post their prices for the following day and are not allowed to change prices.

The newly elected Federal Government is now pushing to imlement the scheme on a Federal level amidst considerable debate on whether the scheme will force retailers to set higher prices than would have occured without the scheme.

The question that Australians face is: Does the provision of information outweigh any unintended increases in prices, ie,the price set buy retailers varies by up to 12 cents per litre (about 40 US cents per gallon).

not so witty

@Witty Nickname-

Do you have any idea how difficult it is to get up to the minute, accurate price data from the more than 200K gas stations all around the country? They represent thousands of separate companies and most are not connected to the internet. so they're doing pretty well. and you get the info for free, so build a better mousetrap or shut up.


In order to conjecture more I find myself wondering if the increase in lower income visitors can be explained away in part due to steadily high gas prices on fixed incomes, as mentioned before, but also on increasing food, clothing, energy, and other essential good costs.

Would the inflation of all of these goods combined be enough to push individuals who would otherwise be doing fine on a fixed income into the red?

Gene S

I'm a NH Gas Buddy. We have one of the most active GB sites in the country -- compare our 203 members with over 35K GB "Points" with any other state on a per-capita basis.

The lowest price in our area varies. When supplies are tight and prices are going up, the regionals and independents are squeezed into higher prices than the big multinationals. When prices are dropping, as they are now, the independents (such as Z1-Xpress) and regionals (like Gulf and Hess) lead the way down. I have been active in GB in hopes of forcing prices down in our area, but I have not been encouraged by the results.


Could it be at some point of time people just gave up? Kind of took rising prices in their by now jaded stride?


Nathan and A.S. both arrived at what is probably the biggest driving force between the relation to Gasbuddy hits over time - when prices first soared a huge spike occurred as savvy buyers found the closest, cheapest stations across their route.

Fido comes up with an equally valid point about direct media mention of specific sites, although I feel that this would require more specific analysis of the number of times Gasbuddy, specifically, was mentioned in various news media over the same time periods.

The article's main interest is the relationship in average income to site visits. While I find it difficult to explain the June 07' 150k+ percentage I feel that Fido's theory best fits the data. The current percentage of lower income visits most likely represents the increase in how gas prices are affecting those with fixed incomes as they stay at elevated levels over time. Even fixed incomes are rarely completely fixed. Most fixed income individuals have at least a reasonable amount of floating money in checking or bank savings accounts they can easily withdraw. As gas prices stay elevated and this money evaporates many people will only then realize the effect and cost gas is having on them. Still, that is pure conjecture :)



Like most people I visit these sites only when they're linked to in an online article. The fluctuations are likely caused entirely by the different media that link to them. These sites were a novelty in June 07, and upmarket publications discussed them at that time. Newsweek online linked to them on 5/15, that's where the prole traffic comes from.

Bob History

Have you paid any attention to the results of any search on GasBuddy over time? What I find is that, with great consistency, the same 1 or 2 stations in every place I visit, always have the lowest prices. Once I've found the cheapest station, i don't really need to go back.


Unit shipment of long tail ecommerce widgets, be it books or a pint-sized medal from your fav5 ebay junk seller is more dependent on gas prices than the article suggests -- in fact, ecommerce is gas! Sure Amazon etc will gain in the short term as price contracts and service level agreements run their course with the supply chain vendors and shipping providers. In the medium term, a lot of assumptions driving a 10 buck delivery run to rural Idaho will just perish like the low cost flights to middle-of-nowhere.