Two simple graphs provide useful information for planning of promotional campaigns, system maintenance and hardware upgrades.
Our web search engine is Ultraseek Server 3.1, last upgraded in 2000. Every day, just before midnight, it emails me a text file containing the day’s activity log:
- date and time of search
- collection searched
- number of results
- search term
This week, I’ve been collating the number of searches per day, month and year. These are now summarised into two graphs.
In the first graph, the solid blue line shows actual searches from September 2002 to September 2007. We’ve gone from 700,000 searches in 2002 to almost 3 million (projected) for this year. In the graph, the dotted orange line extrapolates this trend over the next three years: in 2010, we should expect our web search engine to handle about 5 million searches.
This information can be used for planning:
- upgrades for the search engine’s software and licence
- hardware and software upgrades for our web servers
It can also contribute to forming an idea about what kind of capacity our campus IT network will require in the future.
The second graph is based on monthly data averaged over the years 2002 to 2007 inclusive. The solid blue line shows three significant peaks in search engine usage during the calendar year:
- the largest peak occurs in March, at the start of first semester
- the second-largest peak occurs in July, at the start of second semester
- a third peak occurs in May, towards the end of first semester
This information can be used for planning:
- routine periodic maintenance of the web search engine by IT staff
- routine periodic fine-tuning of search results and the best-bets list by the search engine’s business owner
- promotional campaigns via the ‘best bets’ list generated by the search engine, and via non-organic links in the results pages
- promotional campaigns via the organisation’s home page and other key pages on our own web sites
Post a Comment