How to find trends that can’t be found with Google trends

Information overload: Everyone tries to get meaning, detect trends and spot
key events.

We have ran a few experiments showing that focusing on influencers within
a niche community and using eCairn’s conversation application, it is possible to :

  • learn faster than using Google Trends or any bottom up monitoring application
  • learn things that are invisible to Google Trends, although they are critical for a given community

Experiment 1: Android

First, on a hot topic that has hit main street not so long ago: Android,
Google’s mobile platform aimed at competing with windows mobile, iphone’s platform, symbian and so on.

They’re 4 top device manufacturers in the open handset alliance: HTC, Motorola, LG and Samsung.
HTC has released the G1 in September and got tons of buzz but everybody did at this time as a rising tide lifts all boats.

Since mid September, leaks have pumped up the buzz on Motorola’s Android plan though their device isn’t yet on the market, not much info exists
(from a mobile blogger’s knowledge) on LG or Samsung plans yet.

The charts coming from Conversation(tm) are based on a community
of +400 mobile bloggers.

Here’re the charts (and by the way, we put together, as a comparison, trends from our application and google trends for HTC and Motorola – it’s not super pretty as we had to make the Google trends charts narrower):

htc_chart-google-trend1moto_chart_google-trend

samsung-android-combined1
The Conversation(tm) and Google trends look the same for HTC
which is normal because they announced the G1 in September (i.e “the niche community of mobile bloggers and the public had the same information available to them). Looking at the Motorola charts, we can see that there was quite a few conversations in September, visible in the spike
on the Conversation(tm) trends but not on the google trend.
We also noticed that the second spike is super sharp in google though it’s not in Conversation(tm); as soon as mid October, the second buzz starts in Conversation(tm) though it’s only when Motorola confirmed they have an Android plan in the making that Google trends start to spike.
Within the mobile blogger community, when HTC generated tons
of conversations, there was at the same time quite a bit of speculation on Motorola; Based on our reading, several bloggers found some information making them think a big team was being put together at Motorola. While the conversation slowed down a bit, it restarted as soon as the rumor came back this time because some bloggers spotted interesting information on the web.

If one can draw a conclusion from this example, listening to a niche community enables an observer to understand inter-related stories as they unfold
and to anticipate upcoming buzz
. The more relevant your community
to what you’re dealing with, the more of these effects you get…tools that gather data from the broader web will somewhat have a tendency to be reactive
and tell stories after the fact.

Experiment 2: The Age of Conversation II

We ran the same trends on a very niche topic very relevant to social media marketing ‘The Age of Conversation II’. Unfortunately, google trends says
that there isn’t enough search volume activity to display a chart.

But Conversation(tm) trends, ran this time against our list of ~1000 social media marketing blog, shows that the buzz has been growing lately on this topic :

the-age-of-conv

By the way, if you’re interested to read what they say about
‘The Age of Conversation II’, here’s an RSS feed from Conversation(tm)

Conclusion:

One of the major payback of closely following influencers is that you can

– See the story unfold,
– See a story that you wouldn’t see otherwise.

This is a critical know-how for companies that not only want to measure what just happened but make decisions relative to what’s about
to happen and influence in return if they can.

One thought on “How to find trends that can’t be found with Google trends

Leave a Reply

Your email address will not be published. Required fields are marked *