1. Short answer: We measure weekly popularity of a vehicle as a percentile rank in terms of the number of unique people who saw various vehicle in a week. The percentile rank of a vehicle is the percentage of unique views in its frequency distribution that are the same or lower than it. For example, a popularity score (or popularity index) that is greater than or equal to 75% of the total unique views is said to be at the 75th percentile rank.

Some of the common terms that you may have come across :

Popularity Index: Simply put, a Popularity index of 57% indicates that out of 100 vehicles (of different make/model) that were viewed, there were 43 (100-57) other vehicles that were viewed more times than the vehicle in question. Higher values indicate greater popularity.
Spread: Number of Unique Viewers (or distinct individuals) who saw a particular vehicle in a week.
Views: Total number of times a vehicles was viewed in last one week. This includes multiple views by the same person.

The primary motivation to offer this statistics is to keep you abreast with what your peers are considering, so that you are better informed about various options available (while you decide upon buying your new commercial vehicle).

2. Okay, then what is the Long Answer?

Apart from the usual percentile rank that is available to the general viewers we crunch additional weekly statistics further, along with Bayesian average of votes (out of 5) cast by you for each vehicle. Bayesian average is a method of estimating the mean of a sample space (in our case various ratings) consistent with Bayesian interpretation, where instead of estimating the mean scrore strictly from the available data set, other existing information related to that data set is also be incorporated into the calculation in order to minimize the impact of large deviations, or to assert a default value when the data set is small.

Coupled with various specifications of  a vehicle (eg.GVW, Price, Power etc.), and weekly popularity statistics we are able to map beautiful correlations and predictions based on heuristics, free from cognitive biases - that provide us unique insights into a vehicle's demand across layers of geo-economic strata. Everytime you view a vehicle, our servers crunch these enormous streams of data continuously in realtime.

3. I see the word realtime  being used everywhere, is this 'really' realtime? Also what if these results are biased?

We suggest you open two tabs (or two separate browser windows). browse to the homepage on the first tab. on the second tab open a vehicle's details specification page. As the second page loads keep an eye on the first page. You can see the Live trend box updating instantaneously. Click on the "More" button of that vehicle in the Live Trend, you can see the statistics (eg. popularity index, spreads, etc)  changing with each view.

4. This sounds fascinating, what softwares/technologies do you use?
While the frontend data assimilation is handled by Google famous V8 Javascript engine and Redis, the realtime computations and co-relations are piped through continuous Node streams and are handled through Apache Storm and Hadoop.

5. I am a techie, can i see your codebase?

Absolutely. We will gradually release our custom adapters and API consoles that will help you interact directly with our servers. Infact, you can follow all our code developments at Github/truckaurbus.


6. Do you collect personally indentifiable information?

No. At Truckaurbus we value your privacy and never collect any personally identifiable information. All streams of data are anonymised before they are fed into our data systems, whether its obfuscation of last 2 bytes of your IP address or Opt-Out cookies, or local 'Do not track' browser settings - we honour your privacy.


7. I have additional questions, how may i reach you?
Feel free to reach us through any of the following channels:

Generic contact: contact[at]truckaurbus[dot]com
Support & help: support[at]truckaurbus[dot]com
Development team: developers[at]truckaurbus[dot]com
Data & insights: insights[at]truckaurbus[dot]com