It is a common question. You are doing some planning with the client. The client would like to know what trends are in play and expected to be hot for next year. What are some of the topics and technologies that are growing in importance?
As I blogged yesterday, having an awareness of emerging trends can be important in helping to shape PR tactics and content such as pitches and messaging. Looking beyond the world of PR, it is important to understand what is up-and-coming to make sure your business stays on the leading edge and is ready to deal with emerging competitive threats or take advantage of opportunities.
In response to the above question, most PR teams would do any number of things. They’d mine editorial calendars to see what their core publication targets plan to cover. They’d chat with their PR colleagues. The pros are well read and constantly looking for trends that seem to be gaining steam. The more ambitious (or better funded) programs do formal surveys, taking their cues from so-called domain experts – people who work in these industries, and report on them and analyze them.
All of the above can be hit or miss, and there are good reasons for this. None of us have a crystal ball and it is impossible to predict with certainty what will be invented next or which of the nascent trends or technologies that are already out there will gain a head of steam and possibly even cross the chasm to mainstream acceptance.
Now that much chatter and content winds up online, one would hope that there is a more automated way to at least determine the topics that are increasingly being buzzed about. It is an interesting search and database indexing problem, and you would think that it would be solved by now.
Unless I am missing something, it has not been solved, and there are some good reasons for this. In general, search engines are designed to interpret a search query and then provide results pages of the most likely Web page matches based on your query and its index.
They are simply not designed to wade through the corpus (Wikipedia definition) that is the Web and flag new words or phrases associated with up-and-coming technologies and trends. Trying to adapt search and content management technologies for such a purpose would be no small task.
(For more about the challenges of mining text and tracking words, see the WSJ piece Making Every Word Count).
Looking at the world of storage networking, for example, such a system would need to know that SSD = solid state drives. It would need to parse each word in a sentence and “understand” the difference between the exact same word combinations in two different sentences:
“Solid state drives are replacing disk drives at an astonishing rate” and ” Technologies that are solid state drive power, size and cost savings when vacuum tubes are replaced with chips.”
In other, words they’d need to understand that “drives” in the first sentence is a noun and part of a phrase and hot trend. In the second sentence, the same word is a verb.
Ideally, such a system would be able to cluster similar concepts such as solid state drives, NAND flash storage, chip-based storage, etc.
As you can see it is no easy task. This shows the limits of machine intelligence when compared with human intelligence. The AI tools are just not there yet.
Buzz tracking Sites and Indexes
They say that search engines are the ultimate database of intentions. By exploring the index you can see who is concerned about Flu symptoms, correlate this data and hence get an early warning about regions where the flu may well be breaking out (see NY Times story Aches, A Sneeze, a Google Search).
By exploring search records you can see who is looking for good places to bury bodies and ways to kill people (or themselves), as has been made famous in news headlines, TV reports and crime shows.
According to the above NY Times article:
Researchers have long said that the material published on the Web amounts to a “collective intelligence” that can be used to spot trends and make predictions.
So why not use these search tools and buzz indexes to get a glimpse of what new technologies people are looking at for their data centers, for example?
Said another way, if you can use Google to get early warning of Flu outbreaks, why not use it to track spontaneous outbreaks of newfangled jargon?
In the case of Google Flu Trends, they simply have hard wired the system to track a few pretty predictable search phrases: flu symptoms, influenza, muscle aches etc.
Looking beyond the world of flu symptoms, in most cases search engine buzz indexes report on the most popular terms people are searching on. By virtue of how they work and people search, they are weighted in favor of already popular items that people are trying to learn about or buy. They don’t in most cases report on terms that are rising in relation to other searches.
There are exceptions and I will be writing more on this later this week.
Interesting post Bob. Beyond the buzz gauges of search engines, obviously a number of social media monitoring and measurement tools are useful for trendspotting. But as you mentioned, more traditional sources like editorial calendars are more useful for finding out which trends journalists see as the leaders for the coming year. Tech magazines and sites have already released calendars containing thousands of topics they plan to cover in 2009.
Thanks, Jay. We know you guys have that well covered (ed cal databases, that is) at Cision, we use MediaSource at Fusion PR.