|keywords for niche technology site|
how to find long tail searches in low search volume topic?
| 9:15 am on Jul 11, 2008 (gmt 0)|
I'm trying to find out what sort of long tail searches people use in a high-price low-volume tech industry. I can find a few general searches (related to technology) but not the sort of long tails that convert to customers.
I was given a large list of keywords that the employees thought would be relevant and I have analysed and ran them through various tools but no long tails came out - I guess the search volume is just too low.
My thoughts are that they should naturally appear when we start to create content on the applications of the said technology i.e. hit the problems that it solves. My concern is you are doing it blind and have no idea where to start in terms of search volume and prioritising.
Any ideas on where else to look for the additional phrases that people might search for in a low volume market?
| 7:54 pm on Jul 11, 2008 (gmt 0)|
What tools have you used thus far and in what context?
| 9:04 pm on Jul 11, 2008 (gmt 0)|
word tracker, adwords and overture
| 5:52 pm on Jul 13, 2008 (gmt 0)|
|...a high-price low-volume tech industry... |
HoHum - I've had a fair amount of experience in this kind of market, and identifying search terms in these areas, particular new tech areas, is a very difficult problem. The new Google AdWords tool probably offers the best promise.
Of the tools you mention, WordTracker is probably the worst for this kind of targeting (or in fact for any infrequently searched terms), because of its very low sample size and its demographic, which I believe is very wrong for hi-tech, big-money search.
Trellian, which has become one of my favorite tools, nevertheless falls down in this particular kind of market, in part because I believe that its numbers are extracted from toolbar info (and perhaps also IP data). Depending on the technology, potential customers for the kinds of products you're describing (niche high-priced low volume tech) generally tend to be large corporate customers, which often don't go through IPs, and which probably wouldn't allow the kinds of toolbars that collect such data on internal computers. Also, the Trellian demographic is a big unknown.
While I don't have direct experience with Hitwise (a very expensive service that's very highly regarded), its data is also toolbar data, and, as such, is likely not to reflect many B2B searches.
The old Overture tool was often a good starting point, because it had numbers and a semi-decent sample size, though it had its own kinds of built-in skewing as well. The tool appears to be no longer operational, though, so further discussion about it is moot.
The Google AdWords Tool has just starting showing numbers, and there's discussion about this in this thread in the Google Search forum...
Google is now showing numbers in its keyword tool
I note in particular on that thread...
|What's really good to see... are numbers for infrequently searched but very important phrases in B2B markets like big-ticket, large company hi-tech widgets, where the buying cycle might be a year or two or three. The data simply falls through the cracks on other engines and tools. |
I am now able to download csv files, but the problem I note elsewhere on the thread, that the tool seems be completely dropping some known high-volume phrases. Nevertheless, Google offers the best demographics and the largest sample size.
AdWords campaigns as a test can be difficult to unproductive, I've found, in low-volume tech targeting, because of the length of time over which you need to maintain ad presence to get impression data (let alone conversion data).
|My thoughts are that they should naturally appear when we start to create content on the applications of the said technology i.e. hit the problems that it solves. |
This has generally been been my experience too.
Often, though, you need a starting point, and in new technologies this sometimes needs to be a series of intelligent guesses. I'd take employee lists with a grain of salt. Without knowing individuals... I'd trust the tech people more than the marketing people, though I'd also look for evidence of search-literacy in their suggestions. Talking with an engineer can often be more helpful than looking at a list that someone else has gathered for you.
While I also try to study the literature in a field, I would not always trust the language that competitors use to target their sites. In general, I find that people search for nouns, gerunds, and participles more than they search for verbs, and, as you note, they often search for solutions to problems.
If there's some history of prior products in your area, I recommend running a great many test searches, using all kinds of syntaxes, wild-card searches, tilde-operator synonym searches, etc, to see what might turn up to give you ideas.
[edited by: Robert_Charlton at 5:54 pm (utc) on July 13, 2008]
| 6:28 pm on Jul 13, 2008 (gmt 0)|
Forums, blogs, dead-tree pubs, etc., that relate to the niche can also be a source to farm as a starting point.
| 8:24 pm on Jul 13, 2008 (gmt 0)|
jimbeetle - What are dead-tree pubs?
| 9:50 pm on Jul 13, 2008 (gmt 0)|
He, he, yeah, I guess you could look at that a couple of ways :-)
Magazines, trade journals and the such, especially if they have a "letters" component or something similar.
| 11:30 am on Jul 14, 2008 (gmt 0)|
Thank you very much for the responses.
Wordtracker gave me nothing at all. I find it can be misleading at best for low volume searches as the statistics are too easily swayed by an individuals 'odd' searches.
I'll see about doing some more advanced searches to see what comes out, plus having a look through the trade publications seems like a good idea.
I'll also arrange a few interviews with individual employees to get them to tell me in their own words what it does and what type of person would be searching for it. So far most of the keywords came from marketing (with some technical input) and TBH they don't show search literacy.
The competition currently doesn't appear to be targeting their sites at all and are competing on authority, domain name and relevance for the few obvious searches.
I've also forewarned the company that this could take time :)